If you have issues viewing or accessing this file contact us at NCJRS.gov.
\ " ooT"HS,805-864 ..... :. r. !i "~
• " DEVELOPMENTAND FIELD TEST OF PSYCHOPHYSI
V. Tharp M. Burns H. Mosko~tz
Southern California Research Institute i, 6305 Arizona I~r~ Los Anples, CaJifornia 90045
Contract NOo ~3T-HS-8.01970 Contract Amto $205,579
MARCH 1981 ! FINAL REPORT
This document is available to the U.S. public through the National Technical Information Service, • Springfield, Virginia 22161
, Prepor~ for U.So DEPARTMENTOF TRANSPORTATION j.... NalJonal Highway Traffic Safety Adminisb'alion Washington, D.C, 20590 o- /. :a"
.~~,:~.~,.~-.~: _,-~. !~ ~
..... ~.~ ....:
• ~ ~ ~ L ~
-~ I~,~ ~I 8~
~D ... ,, ~ ~ • .~'
~ I~l~ ~.~ .~I~
____3~._• ......
•~ ,-~ .~ ~..~ ~ ~ ~ | -i|-
~ ~ ~ .~ ~ ~.-'~.~ TECHNICAl REPORT STAND+,,,D 4, .Title end $ublitle + $. Repo,t Dote DEVELOPMENT AND FIELD TEST OF PSYCHOPHYSICAL MA~Ch 1 qR1 ! TESTS FOR DWI ARREST 6. Performing O,0o~ia~tion Cede 7. A.tKo,(,) ~. Por/4)¢~ng Ol@on|Beflon I~epo,, i'de. V. Tharp, M. Burns, and He Moskowitz SCRI-TR81- .+ i 9. Pedo,ml, 0 O+g~llotio, N~o ~d Ad~,oc+ 10. We,& U,~t Me. ~ "Southern California Research Institute • I !. Co,t,oct o, G,ent No. 6305 Arizona Place +~,~ ; • LOS Angeles California 90045 <~< DOT-HS-8-01970 If 13. Type of Repo,t end Pe,iod Co~,er*d ]2. Sponsorlng AgencyName end Add,osl '~i '*'[/.~ Final Repoz~h • I Department of Transportation /i ;Aug._~,1978 -March 1981 National Highway Traffic Safety Admin'istratior l • 14, $ponso,i~gAgoncy Code " . '+ Washington, D. C. 20590 I IS..~,upplementayy Nolo0 none i 16. Abnt,oct " •Administration and scoring procedures Were standardized.for ' i a sobriety test battery consisting of the walk-and-turn test,**.the .+I~ one leg stand test, and horizontal gaze nystagmus. The effectiveness of the standardized battery was then evaluated in the laborat0rY.and, to a limited extent, in the field. Ten police officers administered the tests in the laboratory to 297 drinking volunteers with blood alcohol concentrations (BACs) ranging from 0 to 0o18%o The officers were able to classify 81% •Of` these volunteers, on the basis of their test scores, with respect to whether their BACs were above or below 0o10%. Officer est:imates of the BACs of people they tested differed by 0.03% on the average from the actual BAC. Interrater and test-retest reliabilities for the test battery ranged from 0°60 to 0.800 - ° In a limited field evaluation police officers filled out 3128 data forms, each represented a driver stopped during a three month ' period. Police officers, after training on the administration and scoring procedures for the test battery, tended to increase their arrest rates and appeared to be more effective in estimating BACs of stopped drivers than they were before training. Anonymous breath ~~' 1 testing of released drivers who were stopped indicated that many of the drinking drivers were never given a sobriety test. ' ++ : i 17. Key ~o,d, sobriety test one leg stand This document is available to. the standardization horizontal gaze- U°S. public through the National o alcohol nystagmus Technical Information Service, I police Springfield, Virginia+22161 . Walk and turn "+*++vc°+"+h++uncla++***e onclas+i *+ l" + • + :ors DOT F 17g0.7 co-69) ', ..+ ,. i I - + .'... •.r PREFACE This study involved the participation of three different police agencles whose cooperation and support was essential. We are especially grateful for the exceptional contributions to•the : S project of th e administrative•and patrol personnel of these agencies. The•agencies, along with our principal point of contact, are listed below in alphabetical order. California Highway Patrol, Southern Division Cpt. Kenneth Rude LosAngeles County• Sheriff's Department Sgt. Harry Douglas Los Angeles Police Department, Central Traffic Sgt. Richard Studda~d • • ~ We also wish to acknowledge the contribution Of two research ~ assistants, Melinda Baille and Leslie Rosdol, whose contribution was exceptional. The Contract Technical Managers for t~is project were Stephen Benson, Pamela Anikeeff, and Robert Schweitz. We appreciate their assistance and support. 4 lli TABLE OF CONTENTS PAGE I o INTRODUCTION ooooooo•ooooo °°e°°oeo•o•e•oe•o•ooo•.o.oo 1 Ao PRELIMINARY ASSESSMENT OF FIELD SOBRIETY TESTS.... B. PILOT TESTS WITH THE SELECTED TEST BATTERY '5 ,, 2 1. •Standardization . . q 3 2, Field Observation°::: ::::::::::::::::::::::: 3" 3 • Pilot Subjects ooooooo•ooo•••eo~.•e•oooee..,o " 4 ao Walk and Turn Test ...... 6 4 b, One Leg Stand Test ...... 5 ...... c. Gaze Nystagmus Test 7 If. LABORATORY EVALUATION OF THE TEST BATTERY 13 / A. LABORATORY PROCEDURES ...... •33 i• • Police Officers, Observers, and Laboratory • ParticiDants "••°o•eooeoe••••o••o•ooe.e~oo, 13 2. Training Precedures for P01ice 0fficers ..... 16 3• Testing Procedures oo••e•oo••o•eoe•oe 16 a Participants Oo'•e•o•ee~•e&o•e••" ...... 16 bo Officers• and Observers • ...... ooeo • ,... 18 .j • ' B. TEST BATTERY VALIDITY e°e•°"oooooe•oo~oe•ooeoe.o'e. "'. i9 ~ . 1% . 1. BAC Estimates ...... 2o Impairment and Arrest Decisions .::':: ...... 19 3. Ability to Classify Subjects with Respect • ~20: to 0.10% BAC ...... -- 20 4. Nystagmus Criteria ...... ". 2•5 a• BAC Versus Angle Onset ...... 25 b. Rater Estimate Versus Machine Estimate of Onset ...... 30 C. RELIABILITY --Qoooooeelooooo • 30 lo Interaction Reliability ...... " '~ " " ...... : " " - 3!0 J 2. Test-Retest Reliability " - 34 : I II. FIELD EVALUATION PRECEDURES ...... 37 , " :. .~ , " . , ~- ' . . , A le POLICE AGENCY @ o • o • • • • o o • o o o e o o e e o "'" ...... 37 _ .....7 L -" Y i Ac~u~s~7~o~s ' • ',." ~ , ,, , . - ,. ,- PAGE 38 ~ ~£ation c 3~ 38 °•°°°°°++&°o*bo+~+%+o'ee +~o°®e+°=.* 38 ~ sTuDY DESIGN 39 ++ +~+++tNG POLICE OFFICERS °°°~'&oomo.mgoo~eg.,oomo o 41 +°D+ +++LmC+tOm ++ . . l++lO~++OOmlOOO++.~'O~+m++++.O+Ir++lm 42 t.+'.~ ba+a F~-ms +ooooooo+ooemb.++&oo~oa+o%.++++.+ ~2 2+ Ridealong Data ° .... +°+o+o+o+ .... +o.°°°+ .... r 45 .!++9+ +~O~TtS OP THE FXmLD STUDY eo • +oLeo•••eeL. ~0 +~NATOREQF T~+ STOPEE POPULA+T+ON +o + +° " "om%e D~ o• o 5O i Age • o oeooo~-ooloooe.osoea~oooo.o m ~2 2~• Sex ...... "'+" + ma+c+ +'+'''r"',++'~+++''°++~ +++'.'"+'0++" 52 + °+°+°°°°°'°° ++°o +*•+•'co+ + 52 +B•% ++m++~+ BAKERY EFFECTIVENESS .... °oooo~o..oeo+oo%,o.0.e.~o. Qo 54 ~I+ Wil*l the Percentages of ~he S%opees ~Pn+ ,Are Arrested Increase after the Training on +the Test Battery? 6°ooee +2'~ Will +Police Officers Make More Accurate 55 •Decision with Respect to a +~AC of 0,o10% after iBeing Trained on the Test Battery? .o 57 +a. ~xposure to Drinking DrriVers +,.oo+.oo.+ 57 b. AcCuracy of Decisions o.~..o..+ .... .+o...o 57 3 ~, [Wi+l~ the Mean BAC of Arrested Drfvers Be +Reduced after hhe Test aattery +Is ~nt+rbduced? o o . +..+ooo+o+ °++. + +~60 + o o *+62 Refusals. o0oo°'ooo0ooooo~oo~°o~ oo- eo+•oo 62 +~. Will ~Police Officers More [accurately Est'~ma+te the BAC Levels of st opees M~+t~r Being Trained on the Test ~attery?, °~°+°°•°°°°°m°°o~o~oo+o+ • °oooemeoeoe°.oe' ~62 + Few Stop ~re ~eS~ o ...... o+.°° a+ ees 162 b. Most Officers'BAC~Est.ima~eS Were Inva lid +.o++ooo++++.+•...... ++. Co ~lood and Urine Da~a+Were 0bta+~ned on a Biased Sample of Arzes~ees ...+ 163 ~d. Given ~hese Problems ~ the ~Accuracy of the Officers ~ BAC Estimates Tended To Be More Accurate after Training' • ° ° e e • o • ° • ~ o'~'o o o" • • - ~L •L+ P J PAG~ 5. BAC Distribution of Police Stopees ...... 65 a. Driver's BAC Versus His Arrest Probability.. 65 b• BAC During Different Phases of the Study.... 65 1 c. BAC Versus Type of Driving Error ...... 68 C. FEASIBILITY ...... , ...... 68 i. Police Attitude Toward DWI Arrests ...... 68 2. Police Acceptance of Standard Administration and Scoring Procedures QIe°''o•o•e•eeoeeooe•oee• 3 Set BAC Levels 69 md • • oeeoeooeooomooeooeoooooeeeeoo.eee 70. V. CONCLUSIONS AND RECOMMENDATIONS • °eoo 0oooo. • oooooo° oeoo... . 72 A. CONCLUSIONS O'°Ooo@°°oeooooooooeoooooeeoooo@eeeaooeeo 72 B. RECOMMENDATIONS ...... - 73 1. Police Attitude and Motivation 73 2. Adequate Time Frame for Data Collection ...... 73 3. Other Considerations .... "''" °°°e'eO•eeooee.ooee•eeeeeeeo 74. IV. REFERENCES •O@°"••@gOQtg•qleeemOoQoeme~oooo~oe4eooeeeeoeee 75 APPENDIX A: LITERATURE REVIEW •°eeDeeeooe~eeooeeeeeeeee.e 79 A. ALCOHOL AND NYSTAGMUS ...... 79' l• Vestibular Mechanism ...... -.-o- 79 .2° .... Neural Mechanism °•°'••°•e-oeooeoeeoeuooe..°ele.' 81 3. Gaze Nystagmus ...... • ...... 82 B o ALCOHOL AND BALANCE °eeeeeeeoeee.eee.oeeooool o, ! 83 1. Walk the Line ...... 84 2. One Leg Stand °l°°eeoooeooooooeeoooooteoeoo~eooo, dp • 8•4 APPENDIX B: INSTRUCTIONS AND SCORING SHEET • ooeooeeoooe • . 85 .J 1• vii i • C. • " ).: " , •. .... o LIST "OF TABLES TABLE PAGE i. Backround of Officers Who Scored and Administered the Field Sobriety Test Battery.. 12 2~ Age and Sex Comparison of Laboratory participants with Stopees from the Field Evaluation ...... 14 3° Mean Absolute Value of the Difference Between the Actual BAC and the Estimated BAC of Each Rater ...... 21 p 4. Percentage of Subjects Classified as "Arrested" or "Impaired"' at Each Alcohol Dose .o...... 22 . Rater's Criterion for the Arrest/No Arrest Decision °°°~"~"'°°6°e~oo~ooooo~eooeooeoeoo~o ~3 . Rater's Criterion for the Impaired/N~t Impaired Decision ...... • ...... 24 . Classification Percentages with ResPect to a BAC of °10% for Individual Raters ...... 26 Decision Matrix for Police Officers ...... 27 9. Decision Matrix for Observers ...... 28 I0. Correlation between Machine Nystagmus Readings and Blood Alcohol Concentration ..... 29 Correlation between Machine Angle of Nystagmus Onset and Individual Rater Estimates of Onset, ~the Rank of these Correlations, and the Rank of Each Rater's Classification Ability o. 31 .21o I~errater Reliability on Each Session ...... 32 !*nterrater Reliability: Individual o~ficer- Observer CQrrelations ...... 33 Test~Retest Reliabilities for Decision and ~Test Scores °c~a°o°~'oo~aI~ooeeoloooooooooooeo 3~ i vliL ~GuRE m , P~GE A~~ent @f Participants to Cells Ac~qrding .... and 2~. oeeeo eo 15 5, Three Phase Design 4Q ~e~i~e fo~ Obtaining An~m~ous ~ea~h S.~pl~ 46 "L /. TABLE PAGE 15. Analyses of Variance for Between Session Rater. .. ~, . BAC Estimates for Officer-Same, Officer- : ~ Different, Observer-Same, Observer-Different ...... 36 ..... 16. Pre e Training Data Form ...... ~..... ~ .. '~< ~'~~ ° °o'e°oooo• o.m eeooooe.ee.oee 43 17. Post Training Data Form °''------.-.o.o..oo...... : 44 18. Data Obtained from Stopees during Ridealongs . ~... 51 19. Age and Sex Distribution of Four Groups of Stopees during the Field Evaluation ...... 53 e ° e e 20. Stops and Arrests Made during the Field Evaluation as a Function of Officer Grouplng ' and Study Phase ...... " ..... " ...... 56 21. BACs of Released Stopees as a Function of Officer Grouping and Phase of the Study ...... 58 22. Before Training Decision Matrix • z ...... 59 23. Arrests, Available BAC's P and Mean ~AC as a function of Officer Grouping and i Study Phase Q ° 0 0 D " e ° 0 0 0 0 O " " 0 0 0 " " ° ! ° 0 ° " " 0• ° e " ~ 0 e e • e 61 24. Mean Absolute Value Difference Between Estimated BACs and Actual BACs of arrestees giving blood or • ~ Urine Tests .o. 64' 25. Distribution of Stopees According to BAC and Arrest Probability Before and After Training ..... ~ as a Function of BAC "''''''------o---.--...°.... 66 26. Most Common Reasons for Stopping a Driver • during Field Evaluation ' •~i:~ °'''--o-°----...... ~67 LIST OF FIGURES " . ~J ~'.2q j FIGURE t• 1. Nystagmus Device °°'°''°°°oo-oe-eoo..o.omol.oeo..o 6 2. Regression of Angle of Onset on BAC for the Right and Left Eyes ...... •8! . Angle of Onset as a Function of Time of Day for the Right and Left Eyes under Two Alcohol Conditions Q'oloo'--ooo,m,o,,,,eo,.gooe,,,ooo,Qo,, 10 ix .... - ' " " - " ~ .' ~ ...... ~" ~~ ~z ~ R~R~ . "~" :5"2"; " .... ~ :; ,:. .... ~,. J{ ~v~a~. ~- ~-'~R- e "~ ' ~ -' ,t~" "i ...... ~ _=~,.. , ~nd th e ~I~ ~he li~e t~e~ pzov.~ ~o be. ~~:e~ o~ t~e level of in~-~ -~-'~-'- -- ~. - +--~-,- ~,,~'~.ve ~ml:u+...... -I,n,~oe~ Zealand ~ ~edical eznmin~iQu is given in cs~es ~,h~c e~!n~t zo,n ©o.n ~i~t i~,~ entirely " of eye ~n~ o~ ~I coho ~el:a~io+a~:hi~ to, i,nCo~ic~ion. B~ed upo,~ Pre,:l:im!u~s~y pi~o~ ~o~k,, ~n.~l~y.iee~.. T,h,e three nb.e~t0, teet~ ~::ere (1) the one-,les,-~.t~nd ; (2;) ~he!.. '. '~iailk--an,d-turn; ~nd (3) slcohol, game ~.Y~&,~>gmu.s. T.hi:~- ~°~R~men, d,~J: teat,; b..at.te~y,.Could corxectly Claa~irf:y ~O~e ~.h.an 83,X o~ ~he e~a'1,uat,lon otud.y P~rtici.pa.nts ~ith. z.e;s.pec~ to, ~he~he.~. they ~e~e bqve?. '~ • o~.~ : be:lo.~,: . .... ~. BAC"...... of OolOZ. :he ~urpose of Phe~e I o.~ ~his. contract. (DOT-~.S-8-1.9,70,). ~,~ c " ,..... ~:.0 ~om.ple:.t~e, t h,~e l:ab,o!~tory development and va~li~da:tio,n. of. t:he oobriety ~,~ !~Jbat"t.ery i,~.entif.~e~ b,y Burns a~d ~o~ko~its (.~.~-77)° ~i,re,t, the mt.~ard£sin~ the ~eoring, and ~dm£ni~t~tiou, procedu~ev i~ incXu~m~ • ... , +. ' 1 "¢ Alcohol use contribu~ee ¢o a large propo~tion of the fatel ~ and injury relsted ~cciden~e n~tionwide. Curren~lyo ~ttempts todeter the drink£n~ driver consist of informing the public of the hagards of driving while ~spaired (D~I).~nd of the threat ~nd consequences of being ~rre~¢edo Unfortunately, the perceived ri~k by the. publl~ is quite low, e~nce the combined probability.of having an accident or of being arrested for one DWI trip is e~timated ~o be .0.00089 (Summers and ~srris~ ~78) or less than one in 1000. '- " One reason for the low probability of being arrested in a DgI trip. i¢ that large deficiencie¢ eziet in the detection and arrest.of drivers ~ith blood ~lcohol concentrations (BACs) over .... 0.lO~, Drivero on the ro~d, ~s estimated . by Beiltel, Sharp, end Gl~uz (1975)~ are three ~ine~ ~s likely...to have s BAC in the range of, 0ol0Z tO 0.~4Z ~ in the 0.1$Z to 0.19Z range. ~n contrast, the probability of en ~rro~ted d~iver havin$ • SAC in the 0.10Z to 0olAZ r~nSe is b~f ~ great ae that of having a BAC ~n the. 0,15Z to O.l~X r~n~eo ~h~ deficit may be directly. ~ttrihuted to .the police officer in the field, who must detect end arrest the alcohol impaired driver. :~ , ...i.~.:~. The dt¢crepancy between the distribution of BACs among drivere end the d£str ~bution of BAts among arrestees results from the- following: (1) the high SAC driver make~ Bore frequent driving errors which are detected by ~he police; (2) decisions tO .errs.st ~re e~sier to ~ake with the highly intoxicated stopee;, end ,'(3.) ~aany police officers are not ~otivated to arrest d~unk drivers, especially those with lower 3ACso These discrepancies may be at least partially offset by training police offi'cers te discriminate B~C level~ ~ore effectively by using a standardized f';eld .sobriety test batteryo ~e~xMzeA~,~ ~ oF ~ SOBRZzTY ~Bs~s / Much of the available llterature on sobriety testing comes from countries in which a medical e~aminstion by a physician is required to determine intozicatiOno For enample, Finland has no ststutory, blood alcohol limits for driving, but the courts give severe penslties for driving under the influence of alcoholo Pencil.l-a, Tenhu~ and Kataja (1971) ezamined the sobriety test performa~nce o"f. 6839 Finnish driver~ suopected of driving under the influence.-.of alcohol. Xn this studyp the teQt b~ttery included: walking, 8ait in turn~ng~ b~lance, fingor-ta-nooe, picking up matches, counting back~a~ds~ time ~n~ place o~ientetion, end observations ofspeech and general behavior. The three mast sensitive tests were counting ba©kw~rds by threes from 102, balancing with the eyes op:en,..and walhiaS down e corridor with eyes clo~edo The counting :teot~,: however~ ~ p~rticul~rly difficult fo~ people of lo~ ~ocioeconomic background. J. Standardization Cronbach (|970) de~ines a standardized test as being "one in ~ ~hich the procedures, apparatus, .and scoring have been filed so that precisely the same testing procedures can be followed at different times and places." The process of gathering normative data is,,a!so called "stendardization, " but this process is not very profitable• .. until the procedures and scoring have been standardized. • r~. ~'. The first step in standardizing a field sobriety test battery istO determine what aspects of the test bat.tery make the tests particularly sensitive to alcohol intoxication. That is, thefirst step is to fine-tune the tests to beet discriminate between .the intoxicated person and the sober person. These variables' most sensitive to alcohol intoxication are discussed in Appendix A;-: Testing i8 a social relationship in which the interactions between the tester and the testes are very important. These interactions .between stopee and police officer will be impossible to standardize. For example, we have found during police rldealongs that most stopees are fairly calm ~bout getting a ticket, although 30Z to 40X will argue with the officer. About 5Z of the st opees can be very hostile, however~ displaying behavior rangin~ from temper tantrums to hysterxce Zntoxicatedstopees, who are generally the ones given sobriety resets, are much more li.kely t:o, display these behavior extremes. Hostile behavior, or the~ polilce reactions to it, ~ is impossible to duplicate in the laboratory situation for purposes of standardization. o The police officer, in scoring the field sobriety resets, ..is~ interested both in how well the suspect can perform (i.e., is :the individual impaired?) and how well the stopee's perfor.msnce compares with that expected from drivers at various BACs. The primary reason that ~ field sobriety battery is given ~ (i.e., in~tead of using a portable breath analyzer) is to show that the driver's performance is impaired. In this sense, the field sobriety tests must be cpntent referenced, sO that the Police: officer can observe what the suspect can do. However, the police officer in some areas also may know from experience that no matter how impaired the suspect's performance is, the suspect will not .be convicted of driving while • intoxicated unless the individual's~BAC is above 0.15Z or convicted of reckless driving unless the individual's BAC is above Ool0Z. Thus, the police officer;is a]:so interested in a norm-referenced test so that he can estimate . the~. suspect's BAC .... . 2. ~Field Observation A.critical phase of our pilot testing involved observing a ~,'highly:. efficient traffic team working out of the Los Angeles Central, Police Facility which sp®clalizes in arresting intoxicated drivers,. These officers were all using nystagmus in their sobriety t!esting.~ We noticed from observing their arrestees that the angle of !onset of the nystagmusD which occurs as they follow a moving object ~to . the side with their eyes0 occurs with fewer degrees of lateral • . , ~ . , ,. deviation (i,e## ~#£th less lateral ~oVeme~t) as the ~AG i~Cf~.~i@~ .Zn ~dditio~p the ~agn~tude o~ the ~yst~g~u~ ~t eRtrebe i~ter~i ..de~iations is much l~rger ~ith. i~cressinS ~AC~ (i~ ~ ~h~ j~vti~ m~vement i~ Isrger)o ' ° S eic ond ~ we learned that a divided sSt6ntlom ~k ~SU ].d ~e imcbi'porated i.nto the walk-and-furs test ~y h~vi~g ~he s~p~c~ Stand ~eel-to-toe on the line Nhile the dlr~ction:~ o~ ~he t~t e.re being~e~plained. ~ ~ntozicated person can typ~c~11y e~ther li~,en ,to the in~truction~ or ~eep his,baiance~ bu~ cannot d~ both~ ":. • ~enty-fi~e subject~ ~ere given alcohoi and run a~ pilot ~ubje~t~ in the lfiboratoryo Initially, three ~u~j~ct~ ~ere used to rult~ ~ut many of the unimportant v~r~ables in the three teSt.So F~f~n ~ub~ects ~ere then run to determine the ~ffectiveneS~ of ~.he more ~mportant vari~ble~ and to s~d in determih£ng hb~ the ~est battbry • hOuld be scored. Five subjectu ~ere t~Sted" houriy foi" ]l 8 hOUr s ~ ~ both:.sobez and at a BAC of 00~0% -- to ~eterm~n~ the Comb£ndd influence o~ alcohol ~nd ~atigue, FineIly~ ~e at"so tested ~2 ~ober ~ubjects for nystagmu~ in order to determine the effects Of a~e, Visual acuity, and alcoholism history on the iuCidemce of nystag~uS ~n - sober SUbject~. ~he r~alts of ~h@~e pilot vtud~e~ ~,~,re Sum'mariged below a~ they relate to each of the three tests in the ~obriety test 5artery. a~. _ .Malk-aud,Turn_ ~. The ~uspect i~ asked tO assume ~. h eei'to-~oe position on a designated 1in'e, with hie/her arms et ~e sides.; ~hile the remainder of the instructionS are given. ~e st slie -i.s ~hen told to make nine heel-to-~Oe ~ps on the line, to. turn-.around keeping one foot on the line~ and to return in nine heel-.to-toe steps. The ~uspect is requested to WatCh his/her feet ~t all times, ~aking sure that every ~tep i~ heel-to-the and thvt "~h~ ~tep-s are t~ken i~ ~ ~trsight line. ~sking the suspect to balance heel-to-toe while ~isteu£ng to the re=t of the tbsk instruct ions effectively creates a divided attention task in this test. We found that this addition greatly improved the sea,it ivity of the test ~o alcohol, into~icated subjects either keep their balance, ~hile ignoring the subsequent instructi'ons~ Or ~.re unable to keep their balance ~hile listening to the instructions. The sensitivity o£ this addition to the task Sh.pport~ 'the contention of ~oskowita (1973) that divi•dcd s•ttention tasks ~re very sensitive to alcohol intoaicatiob. ,! g'equesting t~at people "watch their feet" while performing this t"e'st • 'also incre~se~ its ~ensiti-¢ity to alcohol~ but makes the ~a~k diffi-cult for p~ople ~ith monocular vision (J.e,~ poor depth perc'eption), Per~ormin~ the walk-and-turn task ~ith .the e ye~ op'en wlth.,'enough light to see ~ome frame of reference is essential if s6b:~r,-, individuale @re to ~erform the test Without difficulty. Finally~ ~e found that the time taken to ~alk the line and the number of ~teps taken ~ere ~.elatively u~mp~rt~nt. varisbles in te'rhO bf altr~ring the sensitivity of the-test t~ mlcoho.l 4 r . .o .4 ~ertaln Individuals have difficulty ~ith this test when sober, including: people over 65 years of age; people with back, l:eg, 'or middle-ear problems; and people ~ith hlgh-beeled shoes (over two inches). Ne recommend that only the nystag~us test be used with the first four categorie~ of ~topees, while people with high-heeled shoes should be asked to remove them. Standardizing this test for every possible road condition was beyond the ,cope of this project~ so we recommend that the , walk-and-turn test be performed on a dry, hard, level, non slippery surface and.under relatively s~f~ conditons. If these requirements • cannot be met at roadslde, we recommend that the suspect be asked to perform the test elsewhere or that only the nystagmus' test be used. The test also requires a line which the police officer can manufacture. Finally, the police offlcor ~nd the suspect should be able to communlcat,@ fluently. Perfo=mance of this test was not o Q worse under the combination of alcohol and fatigue in the 24 hour pilot study of clrcsdlan effects, than under alcohol alone. b_.~. One,Le~ Stand ~eSCo The suspect is asked to stand with his/her heels together, feet at a slight angle and arms at the sides,~ I He or she is then asked to raise one leg about 8ix inches /off the ground (i.e., ~ith both legs kept straight) and to ~old that position while counting r~pidly from 1001 to 10300 Either leg may~ be raised. ~ Cenerally, few variables alter the sensitivity of the one-leg sts.nd test. The most sensitive variable was time. We found~that ~a suspect at a BAC of 0.10Z might easily keep ~£s/her balance for~ 20-25 seconds, but would likely falter after that timeperlod" Consequently, the officer must ask the stopee to count aloud- "from I00] to 1030 in order to estimate the passage df 30 seconds~ ...... ~ Two other important variables are that: (l) the suspect .must • ~,~. able to see in order to orient himself or herself; and (2) the police officer must stand back from the suspect in order not to provide an artifical reference frame which could distract the suspect. Cenerally, if the stopee cannot see or orient with respect to a perpendicular frame of reference, then this test will be difficult to perform even if sober. ~ Certain individuals will have •difficulty performing this test:, und:¢~r sober conditions, including: people over 65 years of age;, p~ople' with leg, back, or middle ear problems; people who are overweight by 50 or more pounds. These individuals should only be giVen the nystsgmus test. Suspects who are wearing over two-inch heels should remove them before performing the test. The one-leg stand test should be performed only on a hardil, , dry,, level, nonslippery surface under relatively safe conditions.i When these requirements are not met at roadside, then the stopee should be asked to perform the test elsewhere or only the nystagmus test should be used. Performance on the one-leg stand test was no'"worse than alcohol alone under the combination of alcohol and fatigde~in the 24 hour circadian pilot study. ,• .J, -% / g , -FIGURE 1 NYSTAGMUS DEVICE. ANGLES ARE PRINTED ON THE FRONT OE THE DEVICE FOR EASIER READING° / 6 va~e ~.y_~ Test. Gaze nystsgmus is a jerking movement of the eyes that ~ometimes can be seen when the eyes are deviated to their lateral extremes (Toglia, 1976). The Jerklng has a si'ov ~nd fast phase~ with the fast phase being in the direction of the gaze (Goldbergp 1963), Gaze nystagmus is considered to be pathological when it occurs at a less extreme lateral gaze (Toglia, 1976); such as with brain damage or depressant drugs. We checked for nystngmus in 42 sober iudividuals,, including 27, former alcoholics and 25 staff members. Approximately half of the people tested ohowed a slight nystagmus in at least one eye ~hen their eyes were deviated maximally. The occurrence of nystagmus in these sober individuals was not related to (I) age, (2) visual acuity, or (3) a history of alcoholism. We did notice Zhat the maximal angle of deviation, measured twice by each of two observers using the device shown in Figure I was 3.03 degrees larger in the O left eye than in the right eye (t, 40p - 5.8, p .001). This • occurred in 28 of the 42 subjects and was not related to hsndedness. We saw no tendency for nystagmus to occur more often in one eye than the other. ~ ' A strong correlation exists between the BAG and the angle of onselt " of the nystagmus. Regression lines for the right and left eyes are" illustrated in Figure 2. The correlation between the angl~e " of onset and the BAG was -0.78 for the left eye end -0o74~for the ~ right eye. In every pilot subject, the angle of onset decreased a:s "the BAG increased and vice versa. Both correlations obtained were quite close to the -0.788 correlation reported by Lehti (1976)' between the BAG and the angle of onset (measured in five degree increments) for 56 arrestees at the time of arrest, We found tha't • "i at a BAG of 0.I0% nystagmus onset occurs at about 41 degrees of: lateral deviation...... ' In our initial pilot work with gaze nystagmus in intoxicate~d 'j. subjects~ we were able to rule out a number of unimportant variables. These variables include: (I) stimulus brightness; (,2.) room brightness; (3) fination distance; (4) velocity~of 'the stimulus movement; (5) monocular versus binocular fixation; (6), instructions to inhibit nystagmus; and (7) the vertical positioning of the eyes. Some of these variables, however, ,are important in siding an observer to record the occurrence of ~ nystsgmus. As a result, we recommend the follow~ng administration procedure: ..!... ~.. First, corrective lenses should be removed. The stimulus should be placed above the eyes in order to elevate them and ~ reduce squinting. At night~ if the street lighting is inadequate, /s penlight must be used as the stimulus or a flashlight is requi-red to illuminate-the face. In looking for the onset of .nystsgmu's, .~'~.e recommend that the stimulus be moved fairly slowly (i.e., at about 10 degrees per second)p but not too slowly, otherwise normal oscillation of the eyeball may be mistaken for nystagmus. The suspec t should keep his/her head still. The officer°s freehand makes a good chin rest for suspects who persist in moving h~i~s/her, head. The officer should move the stimulus twice to the left and r 7 t 'RE.~R~'S~S~ON :~ dF: ~N-G~Ei:..OF:- ONS~;T: ~ B~C:: ' .- t '." I . -] t,JY 0,. ~ ¢,! -:. I~ Z ~ ,t . ., [k. 2b 0 ~ " •e2~ "~:- ' ~'~; - ,^.~,U~ " ,,:~" ',, , • " ' .. O, .1@ • B~Ld'oD A~LCeHOL, ~"- '- .... ~" .... .- • - ~ J . I i" twXca to the rlghtp looking at the eye on the slde of the he~d t0 which he L~ movlng th~ ~ti~ulu~. On the first ~ovement, ,the officer should observe whether or not the onoe~ of the nystagmus occurs before ~5 de,ream with . at le~t IOZ of the conjunctiva (i.e., the white of the eye) ~howingo The ~5 degree angle is easy to eetLmate a~ it eplLt~ the angle connecting the tip of the nose and the center of the ear with the middle of the head,* Some Lnd~viduale cannot deviate their eye s ~ore than ~5 degrees, -so atl least 10Z of the vhLte of the eye must show to escertaln t h'at nystmgmus ie not occurring at the most extreme deviation for that ind£vidu~lo The second movement in each direction should be faster (about 20 degrees per second) and the observer should note whether or not ~ the suspect can follow smoothly and how distinct the nyetagmus is at the maximum lateral deviation. The breakdown of the smooth pursuit • and greeter amplitude nystagmus at maximum deviation are also good signs of • BAC over 0.10Z. Thus, the police offler has three eye signs to look for: (i) onset of nyatagmus before 45 degrees; (2) the distinctness of the nymtagmus at the maximum lateral deviation; amd (3) the breakdown of smooth pursuit eye movements .... ~ The gage nystegmus test may not be applicable to ind'ividumis ~ wearxng contact lenses0 sxnce hard contacts may prevent extreme lateral eye movements. About 3Z of the population wili show early-onset nystagmus, and impaired balance, with no alcobol in their gyatem. This ny~tagmus could be the result of drugs ' other than alcohol (e.g.~ barbiturates or phencyclidine), the result 0f brain damage, of illness (e.g., ~orsakoff,,~s syndrome),, or of unknown etiology. Since police officers often arrest intoxicated persons after midnight0 possible effects of fatigue or circadian rhythms on gaze nystagmus could be signlficant. Five subjects were individually checked for nystagmus each hour between 9 ~om. and 4 p.m, . and between 5 p.m. and 4 a.m., at a SAC of 0.I0~ and without alcoho,l: Thuo, oubjeete ceue to the laboratory four times: (1) between 9 a.mo and 5 pom. wLth no alcohol; (2) between 9 aom. and 5 p.m: at a ~alntaLned BAC of 0.10Z; (3) b, etween 5 p.m. and 4 a.m when sober; and (4) between 5 p.m. and 4 aom. at a maintained sac of O.IOZ. , , ~ , Figure 3 illustrates the angle of onset plotted against time ~f0r all four conditions. Under sober conditions when no nystagmus,was • eenp the manimum lateral deviation was recorded. These data •. were divided into four-hour segments and analyzed with a fully repeated ANOVA~ with the factors being alcohol and time. There-~ was a ~fgnfficant alcohol effect on angle of onset with the .drug decreasing the angle of onset by about 15 degrees. There was also a significant interaction between the effects of alcohol and time in that the alcohol dose decreased the angle of onset~, by 'an addLtonal 5 degrees (i.e., by 20 degrees) after midnight. In all cases the eagle of onset had returned to the baseline level~ at about 9 ~.m. the following morning at which time the SAC was 0.02X or less and the subject had slept 5 hours. The average. ,SAC x-',l (;.u RE ,3 ~Angle .of :Onset as ,a ,Function ,of "Time .o.'f.fD~y ~foc the"~Rig'ht':and Left ;EY,ie; bnd'er~,-~o Alcohoi,Conditdons .@ . 5 5 o. & :~ 50 ° - 4J z ¢ O ¢: 0 "~o (0.o ,~ ...... Right ~ye - SAC=0 ...... ~ Left eye - BAC=O ...... Righ~ :eye- BAC=O.IO~ ..... _~ Left.'.eye ~ BAC~0.10~ 8 [?M-No(,J Noon-4 PH ! F 4 PH-8 PM ~ PH-HJd Hid-4 AM I fluctuation between test periods under alcohol wss less than O.OIZ. When the observed BAC wao introduced as a .coverlets, only• the interaction between the effects O~ the aru8 and time rualn@a'" slgn£ficent. D O 11 L~ TABLE I BACKGRQ .UNQ OF. OFFICERS WHO SCORED AND ADMINISTERED 'THE FIELD .SOBRIETY TEST BATTERy 2 !' YEARS OFFICER FORCE DWI DW: EXPERIENCE STOPPEES SUBJECTS ARREST TESTED # 1 9 7,00:0 1,750 46 # 2 ~D 13 8,000 2,400 48 # 3, ~CSD 1 4 42 ~csD~ 350 25O 4,0' # 5 ¸ : ~HP. 13 3,000 300: 43 # 6. CHP 7 3,500 90O #.7 4:2 CH~ 13 300 2~40 45 #8': LA.CSD 4 25 8 43 # 9 ¸ ~PD< 9, 5,000 750 47 #,1o 19 i0,000 3, 000 45 : LAPD - Los. Angeles Police Department I~CSD.~- Los:: Angeles Ceunty Sheriff' ...... ""::...... ' :...... ,~ ...... • s Department CHP"," ::~.CaLifornia Highwa.y Pat.to1 12 ' ~ '1~ ~ EVALUATION OF ~ ~ ~ATTE~¥! Once the 8corin 8 and sd~inistr~tion procedures had been standardized, s l~boz~tory ~tudy ~ae conducted to evaluate the validity and reliability of the s~ndardised test battery. Ten police off~cers adm£niste~ed end scored the tests. They also made judgements ~,~ to ~hether the subject (ioe,, the testee) yes too impaired to drive, ~hether the testes should be arrested, and estimated the person0s BAC. Each police officer • tested 8pproximateiy 30 people ~ith BACs ranging from zero to 0.~8Z. The Performance of each testee ~as also scored and evaluated by a trained observer eo that interreter reliebKlities could be e~sesesed. Zn addition, half of the subjects ~eturned to the laboratory and ~ere retested under ~n identical alcohoi dose. . Thus. teat-retest rellabilities ~ere al~o assessed. This chapter details the procedure~ involved in the laboratory evaluation ~nd Presents conclusions r.egerdlng the validity end ~eiiebility of the test bmttery. Ten police officers ~ere recruited to administer the test battery.. The officers crime from •various police agencies in the Los Angeles~. • res ~nd varied considerably in experience as i,ndicated in Te.ble I.•.: Yr, trained research assiltants served as observers. : A total of 297 individuals Participated in the study, including~202 males and ~ 95 females. One of the 95 females, dosed to 0.05Z.~ va•s ~unable to participate in the evaluation due to illness. One hundred forty five of the 296 first-time participants return~ed ~for a second session. ' ,: Table 2 compares the age and sea of the 296 participants with the,i:- age end sex of the 3128 stopees from the field evaluation (see Chapters 3 6rid 4) end the 384 stopees who vere suspected of being under the influence of alcohol by the police in the field evaluation. The distributions are quite close, except that fewer people suspected of being under the influence of alcohol were female in the field. In eddition~ individuals under 21p who :could not be given alcohol in the laboratory, represented 23.8X of eli 8topees end 14o2Z of the at.peas suspected of drinking. ~ ~ The experience of the SCRI staff in ~dministering alcohol to people! . vith different drinking histories indicates that dosing limits must : be Get according to drinking history to avoid overdosing subjects. : Volunteers with ~ "heavy" drinking history, as determined:by.the,,~ Q-F-V questionnaire of Cahalan, Cisin, end Crossley (1969), can be~ dosed to a m~ximum BAC of 0.15Z; those eith e "moderate" drinking hi~tory can be dosed to a maximum of O.llZ; and those with a ight" ~rlnRing history can be dosed to a maxlmum of 0.05Z. In order to include light D moderate, and heavy drinkers in the . 13 i! ~C0HOL SUSPECT S~EES STOPEES N~B~R 3128 384 MALE 89,.! 32°o !0,9 o~q.% 14,2 ~ : %,3-,:.%% % 3:2,.. 1% 18...2 %. 9:,4:: % 13o2 % 2.4,, %. 3~.2 !..S~ %: oi.8,~ % ,- ?.;'. ? / ,) -, .. FIGURE 4 DRINKING HISTORY i ~g BAC LIGHT MODERATE HEAVY 0~ n=30 n=32 n=35 (n=18) n=97 .:,~ (n=16) (n=16) (n=50) .05% n=33 n=33 n~36 n=102 (n=15) (n=16) t (n=lT) (n=48) .11% n=30 n=34 n=64 (n=15) (n=14) (n=29) ~/~/ '~ " ./i .15% n=33 n=33 (n=18) (n-18) n=63 n=95 n=138 (n=33) (n=47) (n=65) +; FIGURE 4 Assignment of participants to cells according to ~--a~--drlnking history on session 1 & 2 (in parentheses). 15 4,1, " ~DOrat:ory ~valuatiozi0 together ~ith a ~i~e ran~ of (1,e.D pla~b~ ~O.().~SZ), the design iilu~t~a~l £u ~Ie0~oi ~e~ fi~~d ~ " ~il ©eli sbOuid contein ~pproxi~te19 33 Figure 6 ~a~ :se~erai ~ayb p~i6r to ~es~i~g psr~iclp~n~s. ~icti '0ff~eer ~ given -~t i~h~ training SeSsion hel~ e~ SC~I, ~he P'~OJ¢¢~ Virector ~hen w~n~. thorough she ~anu~l ~e by p~ge with e~ch offic¢~ ~l~rify~ng -:~if~iefilti~ a~d empb~sigin~ i~portant i~e~ • ~he eye~ Using the ~evice i llu~rate~ ~n F~ure I. Th~ ~oc•~Ur'~ ~bueted eo Covering the ~ar~ing~ on the 4e~ice en'd ~i'fi'g ~fie officers to estimate 30 ~nd 45 de,~e¢~ of late~al d~iati6h On ~he ¢yee of v~rioue ~aff ~embere. ~or thi~ ~rainin•8 ~i~' £y,pi;cail~ Ubed one ~taff ~ember ~ho~e ey¢~ ~'ould only ~evia~e tO ,,~.bofit ~3 8e/~re'e~ °~nd another uhose eyen ~ould ~ev~ate ae much ,,~ ~.~ ,d.:_., egree.e. c~ .~-.. The 'o~ficers ~ere ~iven immediate feedback on ~h~ir n J-£ they had trouble other people ~ere brocght ih •lfiOr testing until they could estimate th'fi ansle~ within thre'e degrees O'f the reading on the device ~hree con'~ecutive ~imeSo '-F~ifi~liy~ t~O to feur people~ several of ~hom had been drinkingb -~er:e 'teb~t~e.d ~ith ~,he entire .field ~obriety ~est b~tteryo One staff imem~er ~ith nb vision in hie left eye and a bad lef~ inner-ear ,a;1,~, ~ ~iit ~ab:ilit 7 to balance ~hen ~ober e~ ~a~keely i~air~d) ~a~ (x"e'i i'h~c~'iu~de'da~ong the people te~ted, Thi~ preliminary testing %ii0~%d th'~ F~%jece Director to observe each o~fiee; ~d~iniStering ~he ;t'~'~t b:at't%ryo Thle craining procedure b~ou~ht ~il o ffice~s to crxZerxou IbiS1 of Performance in test battery ~dministratiOno ~" ~ Subjects were required to egree not to consume any bl~bhol 'f6r 24 hours prior to arr~ving at BCIII and not to ~oh'~u~e• ~hy food for mt least f • .... -o A,~ .rb.k~i...... ~,. ,~__ , -_. our hour~ Przor re the~r ~rrlvalo ~i~re. .~ seedy. V)Z to V/;~ of the Voluuteer~ ¢0m~plied ~th these ~e~ete~ - :'Three p~eople arrived at SC~X BAC of ~ith ~ 0,05~ or %gC~,z'.~:~ih~~r "b'hd 12 people admitted eating prior tO their mrri'val a~ iVo'l~t~"r~ ~ere ~l'~o a~ked not to consume an"y ~ruS~ fo~ 24 hours ~rxo.r i~o t h:eir ~e~tiugo ~e vere especially conc'eria~ed abOUt dru~s ~hi'Ch~nigh~ pr6-d'uee additive effects ~ith alcohol~ so each subject 16 , was teated for nystsguus using the device pictured in Figure 1 pr£or to being Siren alcohol. Individuals showing moderate Jr. strong nyatsguus at their maximum devistio~ were 8ires ~ p'lac~bo dose. These people ~ere high. flake for being "false p0eitive" classifications (i.e., the police officers would classify them as being over O.IOZ when they were not) when teated. Yhus...the •placebo group was actually biased so that actual roadside decisions might be better than the labor~tozy decision~ depending upon the unknown factor of the incidence of drug use among police stopeeJ. SClI chose to increase the pxob~bility of a false positive classification with these people rather than risk that they had consumed drugs which might cause them ~o become seriously ill if they also consumed alcohol in the laboratory. Rovever, only i3 such individuals ware found representing 4.4Z of our Subjects~ Although actually at a zero BAC, only one of these individuals was estimated to be over 0olOZ by the officers and four of them • were estimated to be over 0.10Z by the SCRI observers. Participants were scheduled on weekend days between May 6, 1979, end July I. 1979. During each of these sessions, two subjects were~ asked to arrive at SCRI at the same time at prescheduled 1~ ".minute~ intervals between 7:30 a.m. and noon. Thirty. eight time slots per ~aY thus were allowed for subjects~ estimating that approximately- u people would actually come to the laboratory. ~ i.., Subjects were each given three drinks containing ormnge jui~ceimixed~ with Vodka according to their assigned dose level. Eac;h of the/ three drinks wa~ to be consumed in a half hour. The importance o/f dr£nking all three drinks for the study was stressed, but subjects were also advised to atop drinking if they thought that con t~inuing~ might make them ill. Eight people (2.9Z) failed to consume all three drinks. These subjects, except for the f~male who became il, F mud was never tested, were reclassified into a lower alcohbl dose group. . , • . .I.. ~ One half hour after finishing the lest drink, a subjectUs BAC, as measured by analysis of breach samples by an Xntoximeter, and angle of onset for nystagmus~ measured with the device measured in Yigu~.e 1, were determined by a trained research assistant. This. information van withheld from the participants~ who were then Bh0~n Co a room where an officer end an observer were located for..testing purposes. After the testing hmd been completed, a second BAC i,was taken on the Intoximeter and the subject was told the approximate time he or 8he could leave the laboratory. No subject ~as .allowed to leave until his or her BAC fell below 0.03Zo Subjects were then given lunch (also dinner for those staying long enough) .... EaCh... participant, prior to leaving, wa~ asked whether or not he.or, she wished to participate a second time. Returnees were then selected. by ~he Project Director from a l£st of tho~e desiring to return.. Those who fit the needs of the study in terms of dose ,(i.e ,~.~ subjects vere given the same dose on the return session) and drink£ng history were asked to return. No subject desiring to. return was given feedback about, his or her performance or dose level until the coupletion of the second ~eseiOno .+ ~+ ~.~ice~ ~ ~ Officer~ ~n~ ~bservers r~ported so the ~l-obo~atory about 9 e.m. ¢o ~et ,them~elve~ :up tn .the ~.eetiv~g ~ooms+ g~eh officer-observer pair ~ao iaoia~ed from contact ~ith ~h~ +flr++~.~emm~U~ ray of+ice+-ob+erve: pairm rmmmime~ to~ethm~ ~m end+re dry+ On the mecend +estin+ day the m+O ebme+ve+s e++m+hed :se~'~on)~ ++the t~o ob~e~ver~ ~i~ched plsc~m ~fter te~tin~ ~bout + ~en-~ar~c~pmu~m. O++£©er+ end ob+erver+ m+'e col~e+ctively cmlle, d ~+'r~te+r+m~or ~tester¢ ~ in ~he remainder of ~his report. P~Tt!c£p~nts were te~ted mt ~5 minute ~n~erval~ between 9 a.mo ~nd 2 -pomo ~hen ~ mubjecc reported for ~e~tinC, he or ~he ~s~ qu~mmed by the OffiGer (1) o_n how much alcohol b~d been consumed; (~) -.on -how++~+ ~ntoxicm~'ed he or ~he felt; and (3) on ~ny medical problems ~h~xch might contribute to poor perform~ceo The officere ~lso ~8~ed. the participant to b!o~ into h~e hand to determine if en o~or of_ ~:lcohol +~a~ pre~ent. Appendix ~ contains the enti~e list of queer ~on~, asked by ~he officer, to~ether ~i~h ~he te~t £nstructiona~ the scoring .~heet ~nd the ~ecie~on ~heet. ~ ob~erve~ ~ene~lly ~eke~ ~hmtever q~e~tious the of~ce~ might h~v~ ~ok!pp, ed o+r forgotten. ~k~,number of the p~rti¢ip~n.te~ despite bei~ mdvised to behave m~ t+hely..... wO+u!d i'f they bad been +Copped ~ ro~e~de by ~ police ~f:£!cer~ promptly ~nformed the temters th~ they were much to~ ~runk to dr~ve a car. Th~s information w ~ er often ~ ~ r ~. m £ ~ 1 e ~ ~ ~.n ~ , becaus.e the placebo effect for liSht ~z++inkere in this study ~ ~e+r-ystrong. ~eevy drinkere~ onthe other hsn~ ~end.ed to. day ~hmt they Would have no trouble drivins even ~hen they h~d been do~e+d to .O.15Z.., All particlpant~ were given three dr~n~s, regardless o.f the e+lcohcl dose, so they 8enerally informed the p61ice o~f£ce.r thm~ th.ey-+++,hsd co.n,.~umed three dr£nk~o The testers were not able to ~et mu+ch m~o.re ~nfo~+mation f~om questioning the I~boratory partlcipsnts. ~+.h,~n. they would from questioning roadside ~topee~o 8ome of the re_sponse~ to ~he of~icer~ ~ que~tion~ may have. been qu~te unus-ua~ for ~o.Sd,s!d~e+ ¢co+pee~o .s+nce our +ubjects were not af+ai+d of beins ~r~e~+~+ed. and a_ ~tr.on~ placebo effect ~s not liRely to occur a,t z..o,+.a,d,side. Wh+en qua.crOoned about the co~+te,n~ of the ~r£nk~, the mne,+~.ers £nclude, d the f-ollowins: ~or~nge Juice;-" ~th.ey ~eZe about -l~i.ke you. wO.uld get at a bar e* (this woe m ple,cebo subject); "the £+!~s~~. t~o+. jumt t~ted l~Re wster, but +red tip the bartender for the loot o+ne. ~ Af,fte~r, q ueeti.onin.g, ~he p~rt~c~pant~ the of+f.ice,r administered the f~,iel~d~ eo+bz,iety t.es..t~ described in Chapter I u,Sin.~ the instructions ~i+~!eu ~++n,~Append-ijx B o, Finally~ after the participant left the te+m+t~n+:~,, ro.om~, the off,car and the observer ~ndependently (1) '.± ~+c.i,ded+ wh~thez they would e,zrest the indiv+~dus.l+ ++ that person +a.~. to+~, ~mpmire+d to dr~ve; end (3) attempted + t,o eerie+eta the+ ~AC o~~"~+ ~+he. £.n~.~d.ual to within 0.01~+ Pot the trotter two judgements -the+y also included+, a confidence r~tlng~ con m~stin+ of e number from. one to, t.en ~!~th ten bein s the most confident. Dec£sion c~iter~a~ .+ ~e+oe~d+-on.... . the+. .... pilot teats for the project~ were included . on the " 18 / +~+ . +. i..: ,. +' -w~zmxon sheet (also given in Appendix B) but were not necessarily followed by the testers. After the participants left the roomp.the observer was allowed to comment upon the officeres administration of the test battery if ~uch comments seemed warranted. 3.. ~ ~ VALIpITY Validity refers to the degree to which s test measures whac it is designed to measure0 which in the case of field sobriety tests,• is the impairment produced by alcohol. .The primary criterion by Which the test battery was evaluated, the Intoximeter reading, presents a problem because no absolute impairment threshold exists- .for alcohol. Individuals very in alcohol tolerance. An in!frequent drinker may be severely impaired at a SAC of 0o0§Z, whereas a heavy drinker may show only minimal impairment at this level, Experienced traffic officers in Los Angeles claim they do not use BAC as an arrest criterion and only arrest when they feel that a driver ia too impaired to drive. Their only concern for SAC. is that a conviction may not be obtained, regardless of the amount Of. impairment, if the BAC is too low. This is a common prob~lem. in. states that do not have ~ ~e laws I!.e., automatic when the BAC is above a psrtxcular level conviction The average SAC of those arrested for DWI across the United States is 0.171 (IIHTSA~ 1972). The primary goals of a standardized field sobriety test battery are to lower the ~verage BAC of the arrestees~ to give police officers a more oensitive £nde x of impairment, and to give police officers more consistent evidence: for court use. Because of the problems mentioned above, ~ these ' goals are not ~ynonymous. Thus, the criteria for determining t h e. validity of the test battery are not straight forvard. The lntoximeter reading, the most objective criterion available,, is used in thls r@port. " . ~ Since both police officers and observers estimated the BAC of: ,. each participant, one measure of the validity o£ the test battery is t,o compare the estimated BAC with the actual SAC. The mean dif.'f,erence between these two measures indicates whether or not their errors: of. estimation were unbiased (i.e., were consistently overestimated or underestimated). The mean abmolute difference between these two measures indicates the average amount of error. ~ : The ~ean BAC estimate of the officer~ differed from the actua~l BA'C re~dings by O.O005Z. l~one of the officer's estimates .wer.e. significantly different from the actual BAC reading° That. .is,~ ,. overestimates and underestimates cancelled each other, indicating.. that the errors were unbiased. One observer, however, conslstent:ly overQstlmated the B£C by an average of 0 0126Z (t 221=4.67,., p~ .001). • . • : , . ~. The means for the absolv~ value of the differences between the~ 19 e~tt~ated ~AC and ~he ec~ual BAC fo~ eech off~ce~ ~n~ e~¢h observer between t~e officor ~ti~atee ~ the ec~ ~Ce ~e~e~e~ ~@~ F T~-off~cerc and ob~erverm ~ere aloe ~mke4 to ~ec~e ~he~e~ or ~o'~ ~-~iv~=~i ~ho~l~ be a~ze~Zedo ~e ~Zere ~reed ~ t~ey ~o~l~ "~r~" par~ic~pan~e e~i~a~ed ~n ~he r~n~e o~ 0,06~ to 0.0~ ~ho 9ere.-obv~ously ~pai~ed. ~e~ performance, U~ ~he c~eria ..'. ~:, ~o-officer ever arrested a person ~ha~ he ~i~ ~o~ aleo rate ~ be~n~, ~ inpaire4~ Couversely~ fe~ p~r~icip~n~ ~e~e raCed ae be~n~ impaired ~ho ~ere eo~ else ~azre~e~o" The ~hree o~ficer~ fro~ ~he • oe An~eleo County ~beriff0e Depar~en~ ~ho ~enerally had the lee~:field e~per~en©e~ ~ere ezcep~ion~ end ouIy "er~e~ted" 60~ ~o ,75~ Of ~heee they co~dere~ ~o be ~oo ~mpe~e~ to ~~eo The ~ indicate ~ha~ ~hen an off,car eade a "~o a~eo~" or a "not i~paired" deci~ion~ hi~ eoti~a~ed BAC on ~be a~er~ge ~s ~e~ ~ha~ the accrual ~ACo On the o~he~ hand~ ~hen an officer ~4e a 4ec~s~on ~o ~aZre~Z" or decided thaC the participant ~aa "imPaired e~0 ~hen hie e~timate of the ~AC ~s generally b~ghez ~h~n t~e ac~u~l BAC~ ~h~ :~ tzend i~ probably e~en ~ore pro~eunc-e~ i~ the f~e].~ e~lu~tion.o .Table~ give~ ~he percentage of ~ubjec~o ~Z each ~o~e ~evel ~hO ~ere Warranted" or considered "iupaired'. ~heee ~a~ clearly indicate rhac the officers used more conservative criteria Chert the observers o Consequent ly ~ obeer~er~ "hi~ ~ virtually all parc£c~paue~ ~iven higher do~e~ of alcohol, bu~ a~ the cost of Far~eet~g" ~ore 1o~ 4o~e ~.bjects. The in~i~idual rater~ =arrest" and "i~pai~ed" c~iteria ~e~e calculated by determining ~he estimated BAC a~ ~hich these ~ecibione were made. Table~ 5 and 6 pre~en~ each ra~eg's "~rre~" a~d "impairs4" c~eria~ reepeczivelyo 5ome officers ~eze not consi~eteu~ ~i~h ~hei~ cri~erie, so Zhe value ~s ~a~en ~o be ~he e~zim~ted ~AC for ~hich ~ore "arrest" (oz ~i~palred") ~eci~ious Were Cede Zhan "uouarrest n (or nnonimpaired ") decis~oms. Overall. the Officer~ ° a~re~t crleerion eae O.OS~o ~o~ever~ e fe~ placebo ~ubjecto ~ere ~arrested" because ~heir performance indicated ~ubrecanCial impair~enZ. In many c~ee~ ~he~e were genuine placebo ef~ecZso X~ ~he eels cri~erion nee~ by an officer for arreeZin~ a driver Unde~ the influence of alcohol ~ere a BAC of OolO~. ~hen ho~ accur~tely could BACs be ~udaed u~in~ the ~e~ battery ~coree? In coh~rac~ ~0Y'~-5-0|242. office~ ~ere able to correctly classify 76Z of Zhe par~icip~nCowi~h re~ard ~o e BAC of O.IOZ~ u~ing ~he • 20 m~ TABLE 3 MEAN ABSOLUTE VALUE OF THE DIFFERENCE BETWEEN THE ACTUAL BAC AND THE ESTIMATED BAC OF EACH RATER O #CASES DIFFERENCE s.m° OBSER~R #I 222 .0328 .0263 #2 219 .0278 .0261 OFFICER #1 45 .0278 .0251 #2 48 .0230 .0185 #3 42 .0331 .0237 #4 40 .0379 .0286 ~. #5 43 .0324 .0343 #6 42 ,0237 .0211 #7 45 .0265 .0250 #8 43 .0319 .0272: #9 47 .0344 .0259 #I0 45 .0325 .0304 & 21 . b 15 . '~ " .!. 'OFF~ICERS .6BS~'RVERS OFFICERS oB:SE'RVERS "21% OFFI~ER~ 'uBS E,R%ZER~ °~9% " ~'93% QFF~I~ERS OBSERVERS 9~ % '22 k TABLE 5 RATER'S CRITERION * FOR THE ARREST/NO ARREST DECISION CRITERION RANGE-ARREST RANGE'NO ARREST OBSERVER 1 .085% (.05%-.165%) (0-.10%) OBSERVER 2 .075%" (.00%-.180%) (0-.10%) .08% OFFICER 1 .07% (.07-.19%) (0-.07%) OFFICER 2 .07% (.07-.17%) (0-. 07%) OFFICER 3 .07% (.07-.17%) (0-.14%) OFFICER 4 .08% (.05-.16%) (0-. 11%) OFFICER 5 .10% (.10-.18%) (0-.o9%) OFFICER 6 .10% (.10-.16%) (o-.o9%) OFFICER 7 .09% (.06-.16%) (0-.10%) OFFICER 8 .09% (.085-.14%) (o-.o9%) OFFICER 9 .07% (.05-.14%) (0-. 06%) OFFICER i0 .O8% (.08-.15%) (0-.06%) .082% * ESTIMATED BAC FOR WHICH MORE ARREST THAN NO ARREST WERE MADE DEC IS IONS 23 TABLE 6 PASTER'S CRITERZON* FOR THE xv~AX~D/NOT ~AI~D ~D~C~Si0N CRITERION RJhNGE-NOT XMPAIRED .08% (.05-o165%) (0-o18%) OBSerVeR .08% (0"'.18%) (0-.11%) 08% .05% (. 05-~. 19%) (0 TM . 05%) ,07% (.03-.17%) (0-o07%) OFFICER 3 -, ~.,. .. ,:,,,-o~..- • .0B% (.05-.17%) (O-. 08 %) ,06% • (. 05-. 16%) (0-. 08 %) oF.r.zC.r- s 09% (. 09-~!8%) (0-.,07 %) :!0% (. 10-. 16%) (0-. 09% ) .09% (.06-.16%) ( 0-. 10%) OFFIC,F.R 8 .07% ( .06-. 14'%) (O-. 07 % ) .07% (. o1-.14~) (0-. 0 6 ~ ) 'OF.,F~C.ER i0. -° o8~ (. o8-. i5%) (o-, o6%) • 073% ~ !'E-S~:i~•.~~_ TE:D~ ~BAC~ FQR WH. !ell MORE IMPAIRED THAN NOT IMPAIRED % 24 ,!, * > same sobriety tests. Burns sad NosRo~its (1977), using, a discrimiuant analysis program, predicted that the officers could correctly classify 83X of the subjects by making the best .possible use of the ~nfozms~ion in the test b~tteryo The .dlscriminant analysis essentially finds the best linear combination of s~cores, in .order to cl~aslfy c~ees into groups based upon some criterion . score, i.e.-, in this case based upon en actual BAC of O.10Z.: "~ Table 7 presents the percentage of correct classifications,, fa!ee positives (i.e., individuals classified as being equal to or" above 0ol0Z who veze below this level), and false hegatives (i.e., individuals who gere classified as being belog 0oI0Z who vere equal to or above this level) for each of the raters. Overall, observers correctly classified participants 82Z of the time, ~hile officers correctly classified 81X of the time. These percentages are quite similar to the value predicted by Burns and Hoskowitz (1977). The officers' claaaiflcations included 9Z false positives and IOZ false negatives. The observer clasifications included 7Z false negatives and IIZ false positives. Decision matrices for officers and observers are given in Tables 8 and 9, respectively. ::. :. Both the police-scored data and the observer-scored data ,were analyzed with a discriminant enal~sis This statistical proc:edur.e 'was not able to improve upon the classification of subjects with respect to 0.|OZ for either the officers or the observers. The discriminant enalysia ~as able to correctly cIassify 82~ of the cases with respect to an actual BAC of 0.10Z for the officer-scored data (i.e.o as opposed to 81Z correctly classified by the officers.) and 83Z of the cases using the observer- ~cored data (£.e.,~ as opposed to 82~ correctly classified by the observers). The .~ .fact. that the discriminant analysis cannot classify puch better than the officers suggests that they did an excellent job of interpreting the test scores. ' • 4. ~ Criteria ~". Since the angle of onset of gaze nystagmus was measured, on ~all participants vith the nystagmus device both before and after they ,consumed their drink~, a number of tests of the validity of this measurement can be made. a BAC ~ ~ ~ ~ For both eyes a regression "equation yes calculated for the angle of onset after drinking versus-lthe:-BAC. and the 0ol0X intercept was determined. In addition, e.quations ~ were calculated for the change in angle of onset versus the BAC for each eye. All four equation8 are given in Table I00 Clearly., angle of onset i~ as good a predictor as the change in the~.~angle:.of on~et. The expected angle of onset for e BAC of 0.10Z .£s.,A0.2 degrees for the right eye and 40.1 degrees for the left eye,, .These estimates are quite ~Imilar to those calculated in the pilot-.study of A3 and AI degrees for the right and left eyes, respectivel~ (i.e.0 see Chapter I)o If an angle of onset of 45 degree8 so measured by the nystagmus device prior to testing by the .officers is used ~s the sole classification criterion (ioe., how ~any subjects with an onset of 45 degrees or less have a BAC of 0.10Z or 25 , ,!, ,.',: TABLE 7 • ,° • . C~SIFICATION PERCENTAGES WXTH ~SPE~ T6~BAC OF 'i0% FOR INDIVIDU~ ~TE~ t CORRECT O~!C~Rs FALS~ FALSE CLASSIFICATIONS POSITIVES NEGATi~S 85% 7% 9~ # 2 94~ 2~ 4% # 3 77% 7% 21% 80% 8% 13% # 5 79% 12% 9% 88% 10% 2~ .'#7 . 84% 7% 9% # 8 74% 9% 16% 7~% 13% i~: #!0~ ...... 78,% 13% 9% ~ALL OEFICERS 81,.2% 9% ~)BIS,ER~VERS # 1 81o% ! ~% 6 8~4% 8% 8~% ~LI~, QBSER~EIRS 8:2 % ,I i %. 26 "i' TABLE 8 DECISION MATRIX FOR POLICE OFFICERS OFFICER ESTImaTED BAC g >.10% <.I0% % Correct A >. 10% HIT FALSE n=125 • 64% C NEGATIVE T n=80 n=45 U 18% 10% A <.10% FALSE CORRECT n=~16 88% L POSITIVE REJECTION n=38 n=278 B 9% 63% A % Correct n=li8 n=323 81% C 68% 86% I 27. TABLE 9 DECISION MATRIX FOR OBSERVERS OBSERVER ESTIMATED BAC >.18% <.io% % Correct A >:-%0% HIT. ~ALSE n-124 75% NEGATIVE ,T,.~ n=93 n=_31 7% A <-.!CA FALSE COR~qT, n~ 31,5 85 % POS, IT~IVE o REJECT, !O N n-~48 ~ n~267 .. B - 11% 6~i'%. , ...... i~ ~...... A, %~ Co~reqt n~141 n=29 8 8.2:% q 66i% 90% 2~ ~ore; etc~)# Chert 78~ Of the ' P~:t~¢ip~nC~ can be c~reeti,~ oi oU~et i~ ~ut~re~ into a 4iscrininanz a~iy~i~ SSo2Z ¢~ t~ •~r~£cip~tits ~ould be correctly cl~Ifie~ Wi~h ~p~C~ ~0 ~ ~AC o~ O~-l~O~ C16~IFD nystagm~e angle of o~e~t .i,~ sn ~iieut tool ~o~ P~'~'nt~ corzelatibnS between ~h¢ m~chis 0 a~d r~te~ e~im~es o~ ~y~ta~S 6~eto ~n ~dditlon# pb~ice offic'~r~ aSd bbs~rver~ ~r~ ~'~nRcd ].) ~ccordln s to t~i~ir ebil].ty to e~ti~te the an le of o~ee~ /'(i~'. the correlations ~ere ranked) ~n~ ~ r~'k~d 2~ acco~in8 ~.bth,eir ability to cbrrectly classify P~t]Clpants with respect tO ~, BAc kif O,t~Z~ These two ~eto of ranks {al~ .i~t Table ll) were ¢6mp~ed ~ith a Spearman rank Correia~iono This rabR ~orreiatiott Of 0.~58 wa~ sisnifiCent ~uggest~ng ~hat ab:~lity tO estimate augl¢ ,.~f On:set i~ a Cr~ticai factor in making .accurate decisions from the ~~br~ety t'd~t battery performance. The ~"¢iiability Of the field sobrlety testa t~d :meaSUred in ~'~o ~,ay~ FiZSt~ Sn e~perienced re~¢azch ~e ~iSt ant bb~ e i*vred ~nd i~e'Pend~ntly scbred the ~ubject's performance dUrins e~ch ces't 'm'~mi'-nid t r ~t i on o Ob~erver-of~ice~ p~-~ire ~¢re :i'bt~t ed and both ob.serv ers ~OZke d with every of finer. ~hus ~ an ~z r at,e.g. could be calculated for esc.h O,ffi~ce~-obsezVe~ pairlnS, • nd~ 'in ge'heral~ between officer~ and bet@eeh "observers. Second~ h~lf., 0~ Our partlcipants rSturned to be retorted ~t the s~me ,:%.Icobbl dbee. Half of the return'sos ~ere teated by the 0see "6ff._-ieet ~nd the remainder were tested by e different o~icero $i:'mil:a:riy~ h~tlf the zeturnee~ were teeted by the s~me observer S!nd :¢h~ ~emaihd~z ~ere tested by the other-observers Yhus, ~et£~b.ilit'y c~n be calculated for the same t'ester and for different t:e~te~rs on the two sessiOn~o i. I~,errate~ ~'~t:~i?rati~r z'eliability ~as calculated for -each ~ec'ision (i,e., ~a#rebt~ imp aired, and estimated BAC)~ ~or the tbtal test ~core~ and £Or the individual ~coree of each test. Note that ..thes'e items r:~nge ~rbm quite objective observation~ ~such as individual test ~E~rez to deciaio~n~ derived from criteria spplied to the 'test :bco:reS (i.'eo~ the BAC estimate) tO subjective decisions remotely rela'ted to the ceet scores (i.e.~ Whether ~he subject is impaired br ~ho'uid 'be arzeeted). 'T~bie 12 presents the overall officer-obser#¢r correlations for /~e~cf'~:ibn~s and test scores On each 8esaion. Severn1 aspectsof "~hese"data ~t~nd Out: I) interratez reliabilitieB impz0ve on the ~e'~o~ 8e~slon; 2) total test ec0re reiiability iD higher th~n reliability for ~ny decision, reflecti~~ the need to .inte,pret the "~t+O~] L~e'~'t +mc~or¢ ~0 make ~ deci'sion; the i:nterrater reli+bility 'is :h~-gher ~or th¢ decisions, ~uch as the BAC estimate, that are 30 t...,! /. TABLE i0 CORRELATION BETWEEN MACHINE NYSTAGMUS READI~IGS AND BLOOD ALCOHOL CONCENTRATION CORRELATION REGRESSION EQUATION RESIDUAL MEAN SQUARE N RIGHT EYE ONSET -. 710 Y=50.82-100.62 (BAC) 25.19 438 LEFT EYE ONSET -.717 Y=51. 03-109. 44 (BAC) 28.72 439 RIGHT EYE CHANGE .664 Y=.I93+96.377(BAC) 29.98 436 LEFT EYE CHANGE ~D .689 Y=. 224+109.66 (BAC) 33.82 437 t TABLE II CORRELATION BETWEEN MACHINE ANGLE OF NYSTAGMUS ONSET AND INDIVIDUAL RATER ESTIMATES OF ONSET RANK OF RATER CORRELATION CLASSIFICATION r rank i ABILITY OBSERVER 1 .349 8 6 OBSERVER 2 .469 6 5 OFFICER 1 • 719 1 3 OFFICER 2 • 650 2 1 OFFICER 3 • 583 4 12 OFFICER 4 .234 12 , 7 OFFICER 5 .260 ii 8 OFFICER 6 .650 3 2 OFFICER 7 .568 5 4 OFFICER 8 .309 i0 ii OFFICER 9 .432 7 I0 OFFICER ] 0 .346 9 9 SPEARMAN RANK CORRELATION = .580, p<.05 31 " -" ' t "" • ,: .. . T~LE 12 ~~TS~ ~Li.~¢Lr~ZSS ON ~AC~ S~SS~O~ S:EssIoN ~i SESSION # 2 ~,. ~ i i.L ~--" CA~S Z~,[DE~ 29! 143 CASES E×eLUDED, 5 2: ,62 W..~ALK~...... % TURNS: .,.:,; ...... ,~. ,t I ,~ % I" eg~ .74 . 8~6~ e 7~8: 816, • 5.,81 ,5,,9- .5~8: E,S.~,I ~T,E,D DEC I,S ~ONi .72 .80~. 3~:2~ / i • TABLE 13 INTERRATER RELIABILITY: INDIVIDUAL ,. ':,i '• OFFICER-OBSERVER CORRELATIONS ... ,,,: NUMBER OF CASES ESTIMATED BAC TOT~a~ sdom~ OBS.II OSS.|2 Obs.#z oBs.12 oBs.II oss. J2 OFFICER # 1 23 23 .68 .72 .8'6 ' ~.'83 OFFICER # 2 24 23 .81 .80 • 88 76 OFFICER # J 19 23 .81 .77 • 87 ' ''82 OFFICER # 4 20 19 .66 .78 .81 .83 OFFICER # 5 21 22 .86 .87 .84 -86 OFFICER # 6 22 20 .76 .76 .81 •92 • OFFICER 0 7 20 25 .89 .48 .88 .87 OFFICER # 8 24 19 .80 .80 .64 .... 66 OFFICER I 9 25 22 .77 .76 m .93 .80 OFFICER if0 23 22 .64 .72 .89 .'8~7 n-439 r-.75 r=. 80 ., e+:; [ NYSTAGMUS 1-LEG STAND WALK & TURN 0BS.ll OBS.#2 OB$.#I OBS.12 OBS. |i OBS. 12 OFFICER # i .61 .49 .85 .81 • 92 .85~, OFFICER # 2 .64 .60 .86 .79 .68' .64 + OFFICER # 3 .85 .46 .85 .90 .76 .7'1 OFFICER I 4 .48 .57 .76 .88 ~ 21 : +~ OFFICER # 5 .63 .73 -81 .82 .67' 7' .92 OFFICER # 6 .72 .67 .80 .78 .67' .81 OFFICER i 7 .73 .67 .85 .91 .79 ~ .79 I OFFICER # 8 .31 .75 .55 .32 .60' •75 OFFICER % 9 .74 .83 .81 .71 .S5 .66 OFFICER |i0 .67 .59 .76 .87 .95 .89 r-.63 r-.77 r,~.76 ~ :. 'i 33 mo~t directly related to objective cr!teri~ Ouch ~e the BAC estimate; and 4) the interz~ter rel!mbi]ity fbr the nystag~le '~c~re is mot _~a ~igh ae expected, suggesting that the ~ffi~r~ _~ould profit fro~ ffurther training and practice vith nystagmu~. T~e interrate~ reliabilitles are clearly re]ated to the extent to Which the item is objective or objectively based. For esampl~, :test sc0r@~ which are behavioral ratimgs~ ze~lect I) ~he ~ ParticiP ant~s Performance; 2) the rater~ understanding of the behavi0r belng rated (ioe., boy well the rater understands ~at tonstitute~ '~putting one~ foot do~n"); ~nd 3) the rater~s abilit7 ahd motivation to record what happens. DecisiOn ~cores, on th~ 0their hand, ~re ba~ed upon the t~st ~coxes plus a subjective interpretation of the test scores in term~ of some criteria. Thus the results are not surprising , ~bbr observations on the part of ~ever~l indx.viduals could ]o~e~ ~the-: overal~ within-.session correlation between the officer ~nd the °beerver- Thus, correlations ~ere computed for each bFf'icer-obeerver p~iring ~or the individual test ~cores and for the ~'AC e~timateo These correlations are pre~ented in Table 13~ 0vera!l, ~he~e data are quite encouraging. For the e~timated BAC, 80~ of the Pearson correlations are above 0.7 ~ith on17 one belo~ 0~6. For the total te~t scores, 85Z of the correlations are abov~ 0o,8~ and all of them are above 0.6. Since 145 participants returned a second time to. be teated unde~ •the ~ame ~icohol .dome~ a test-retest reliability ~as calculated: .l) for those participants retested by the ~ame ~ffic~r; 2) for those retested by a different officer; 3) for t'hose reteated by :the same observer; and 4) for those rete~ted by a different Observer. Thes~ data ~re given in Table 14 for test scores and for "decision :scores. In ~ddition~ the correlation between the peak BACa. of the two sessions is given to illustrate that the differences in scores are not due to differences in BAC. ffote that 'only about 70Z O f the Participants agreed to re~urn a ~econd time sad returning participants ~ere selected based upon the needs of the ~tu~y. Thu~ the returnees represent .a biased sample. Test-retest reliability for psychomotor teats are typically on the order Of 0.7 (Gu£1~ord and Fruchter,.1978). As can be seen in T~b!e 14, the obtained reliability is of the same order, an ~ccep~able level under these teat-re teat conditions. ~et~een-~ession ~AC esti~ate~ ~ere compared using one-way analyses of ~riance and intraclass correlations~ which ~re given in Table !5. ~hese data indicate that ~AC estimates on the esme individual glive'n the same dose ~ere not significantly differ.eat ~hen made by the ~ame rater on each session or.~hen made by a different rater on e~ach ~se.a~ion . Only t~o of the ten officers had significantly 'different ~AC es,ti~ates ~hen they rated the same subjects a second~ time° Teat-retest reliability, determined by the intraclass correlation, is again on the order of 0.7. 34 . , ..;. ,#~.. .00134 c i,,~3 .Oo06'B • OD i0 2 , 'OBS~"# 1 .o0e76 OBS -~ "#-'2 'i 759 D. 5~2 1 ~,3 ~. ..o,o0,6~ "ovg~L •. '6~B o, ,~ o •:1.,,:7 i ,o,O;o,o-71 :# .~ 3 ~. 7,~ .94 5 "0., ii ,. ,OIU~2;0 % 44 3.b,o &:,c@ '. '0'g~0"9 :I • ~# ,~ ,, • ~4,.2!6 1.40 i., :6 ,. 0'0165 ,;6~:s 1 .~05 :,.1.,, "6 .i000;6!8 : '# '6; . ~%"B i-..~.8 •. 00076 .. "'# :7". 8 ~ 7 0"* ...'Oi30~ 5 ~;e%b 11. '5~6"* • '000'i 6 '# 9 •. ~42 3% 9~ :t/;7 .00031 - 't .'~ 59 0 .'5 0 .... :'0:0201 ~'.~66"5 1 .:60 "~1~ 0 "7"9 . .C0'0081 O..t90 '~1/6 3 .00076 3'6 .. % .j TAI~I,I~; | 4 TEST-RETEST RELIABILITIES FOR DECISION AND TEST SCORES OFFICERS SAME OFFICERS DIFFERENT OFFICERS CASES INCLUDED 77 64 "~ CASES EXCLUDED 3 NYSTAGMUS SCORE .66 .59 WALK & TURN SCORE .72 I-LEG STAND SCORE .34 .61 .60 TOTAL SCORE .77 IMPAIRED DECISION .57 .49 .56 ARREST DECISION .54 ESTIMATED BAC .71 .68 .59 BAC .97 .96 OBSERVERS SAME OBSERVERS IFFERENT oBsERvERs CASES INCLUDED 71 CASES EXCLUDED 72 2 0 NYSTAGMUS SCORE .55 .61 WALK & TURN SCORE .39 I-LEG STAND SCORE .53 .72• .55 TOTAL SCORE .73 IMPAIRED DEC]S]ON .62 .59 .58 ARREST DECISION .58 ESTIMATED BAC .54 .61 .67 BAC .96 .97 35 ._CHAPTER Ill." ~ KV~LUATI,OIq ~ " The primary question ~ddre~ed by ~he f~eld evaluation w~s " ehether police officero, by using the sobriety test buttery, can improve their ~rre~t/relea~e decisiono at roadside. Three types 0£ 'data =ere colle,:ted to answer this quevtlon. Fir~t~ feasibility d~ta were collected by talking to police officers and their superiors about the test battery, observing the te~t battery being administered and scored in the field~ and talking to police officers shout their court experiences. Second, participating officers ~ere asked to complete data forms on every traffic stop they made during the three month study. Third, SCRI staff members rode ~ith each participating officer at least three times during the ~tudy. Breath samples ~ere obtained from released stopees during the ridealongs. A....~. POLICE ~GEHCY Four of the 17 stations of the Los Angeles County sheriff~s Department were selected for participation in the study. The four stations were selected by the traffic division of the Sheriff's Department. ~e were told that the primary selection criteris~ were: (I) n cooperative administration within the ~tation; and (2) ~the availability of traffic cars to be assigned to the project. ~'~ : The Sheriff'~ Department mervices unincorpoFated areas .of Los Angeles County and cities vithin the county that contract ~ith. th.em for police Bervices. Traffic york is only done in contract cities. that request it. The California Highway Patrol provides itraffic~ services to unincorporated county areas. ' , :,i,,- The Sheriff's Department has been providing traffic sets:ices in this banner ~ince 1956. Due to the major emphnsisof the agency on crime and the relatively short amount of time that traffic services have been provided, traffic duty is not highly regarded by most of the deputies. One deputy said that the general attitude is that "the only thing lo~er than a traffic cop is a meter maid." Thus, we were not surprised that ~ost of the better traffic deputies that :we rode with talked about leaving police ~ork as soon as they found something better to do. We believe that the deputies participating: in the study probably still are quite representative of the average traffic officer in the United States, based upon our expe!riences , working ~ith police officerB nationally. ,. The trafflc sergeants we worked ~ith were highly dedicated men vho are concerned ~bout the DWI problem and about traffic enfo~rcement in general. In addition, the Lo~ Angeles County Sheriff's Department ~as the California state agency involved in the ASAP progr~mp which may have contributed to their eagerness to Participate in thi~ progr~. The four ~t~tions a~signed to help SCRI with the field evaluation represented different sections of the Los Angeles Netro.politan 7 / Ares. l_.x ~t_!K!i9~ ~o ~ta~ion ~ serviced an ~pper middl~ c].ass city of -42,000~ The popuRation ~s about 95g Caucasian and shout 5~ ~ispanic. Although the city i~ surrounded by ~e~ropclit~n Los Angel'~, it is quite li~ o znr,.1 ~Id-~erica city. The ~z~ffic lights start to fXash red at 10 p.~. a~d fe~ cars cnn be see~ ex.cept on one of ~he state highways ~hich ru~s through the, city° ~u~h of ~he ~rinking and driving found in the c~ty ~ecults from £~=oxicated people driving a~ay from a ~earby racetrack, g secondary problem re~ult~ from teenage partie~ in which a~ many ~s 'Several hsndred teenager~ frock t~ a house ~here s drinking (drug?) party ,is being held. The police usualRy break =p the~e p~rtie~, ms,king fe~ or no ~zzests~ although ~e e~timate that s majorit~ of the drivers l eavin~ these parties are ]egally into=icated. ~ive traffic officers f~om Station A participated in the field evaluation. Three deputies ~orked shifts from 2 p,m, to ]0 p,m. or from 3 p.m. to ~ p.m. The remaining rue deputies ~orked ]] p.m. 'to 7 a.m. shift@. :~. Station B. We ~orked ~ith three traffic deputie~ from Station B .patrolling a ~orking class city of appro=i~steZy 29~000° The population is about 75Z Caucasian with the other 25Z being composed of various minority groups. A lot of young people~ who ~ould like to. live near the beach but cannot afford besc~ rentals, live ~a this city. Drinking and driving is a common problem in ~his • ecti0n of Los Angeles° The traffic sergeant at this station is very dedicated to keeping statistics on traffic accidents and tickets ~ritten. He has convinced hi~ deputiec that the more tickets they ~rite the feuer accident8 the city rill h~ve. Three traffic deputie~ ~orking thlv city"participated in the field evaluation. They work ~hifts of 2 p m. to I0 p.~., 3 p.~. to l] p.m., snd 4 p.m. ~o ~i.dnight. \ 3.~ S_t~tion C Station C services s heavy i~du~trial community of about I00,000 people. Its population is 40Z middle class white, 40g: middle cl~ss black, ~nd 20Z other minorities. Deputies estimate ~hat the city has well over I00 bars Six traffic deputies p~rticipatedin the program~ excluding one . of the original seven who was eliminated for lack of cooperation. Each of the d~putie~ ~o~ked pom. shift~ ranging from 2 p°m. to ]0 p.mo and 6 p.m~ to 2 a.m. Station C has a well o~:ganized and cooperative traffic administration. 4_~ S~t~t~.0n ~_~ This station services Qeveral contract cities and fiv~ ~ ~r~f~tc cmrs from the entire area p~rticipated in the program •"at'~the begluning. Two cars regularly worked II p.m. ¢o 7 a.m. shifts and specialiged iu arresting intoRicated drivers. The other 38 ::' ,. ' IGURE 5 ~TH~E P~SE DESIGN :E~ERZ~AL Un'tra i ned PHASE !Ill Un~t:r:a i :n,ed [.T~r:ai~n~ed • ' ,~raznea ' "i " throe offieQrg were from crime unit~ but ~ere re~i~ued to traffic cmrs to pmrticipmte in the field ev~lu~tiOno ~hese three deputie~ had 8@De ~e~e~t in m~king drunk d~iving ~rre~t~p ~ut mo int®rest in tasking ~ffi© ~tOpSo AIX of them~ during ridealongs, expressed a deelre ~o ~etnrn ~o trine unit duty. ge received l~ttle cooperation fro~ the traffic mdministration at this ~etionD &rid that ~d~ni~tr~tion changed t~ice during the field evaluation. During tbeCcour~e of ~he ~udy the evening shift deputies filled out very fe~ for~o ~hen ve questioned ChemD they clalmed ~be forms ~ere m~ ho~eo" By the ~i~e ve discovered that the~e deputie~ actually ~ere not filling out for~s, the traffic admini~trstion had been chan~ed. Yhu~, the three p.~. deputies ~ere dropped fro~ the ~tudy for noncooperation. In addition, one of the ~o~o ~hift deputie~ ~topped filling out form~ as soon as he was trained on the ~e~t b~ttery. ~a ~ result~ only one deputy from this stmtion completed the field ev~luation. Ironically, vhile these problem~ were occurring~ three deputies from Station C vere disabled from tvo ~epar~te ~ccidente iuvoIving intoxicated drivers. " / ~he requ'ire~en~ of ~be field evaluation included: (1) obtaining sufficient b~eline dm~ mgains~ ~hich the officers ° performance follo~ing ~rainin$ could be compared; (2) havln K u control -group to aceoun~ for ~uch f~ctors ~s ~he time of year (i°e.~ the Christmas RoIiday~) during ~hich the ~tudy ~o. uuderteken;, and (3~) the need to ~ra~n sll the participating deputie~ ms a revard Co the ~ p~rt£cip~tin~ Ota~OD~ for their cooperation. Thus, a three~ phase design, illu~trmted in ~igure 5, yes undertaken° .:. • ~ ..~ Phase I beg~n betveen December 7th ond 12th of 197~. The different . sturtiu~ ~es ~ere ~ue to the f~ct thet ~t~ff ~embere could only visit one station at a tl~e for stortup instructions. In addition, most stmtions h~d to be visited ~ore than once because all deputies involved ~su~lly ~ere not pre~ent a~ ~he first visit° During,Phase I b~seline info~a~ion v~s colXected by ell deputieSo ~ Phase 11 began between ~nu~ry-12th ~nd lgth of 1980. officers from Station A ~nd Station D ~ere trained on the test battery on the veekend of, J~nusry 12tho Off£cer~ from Station B were trained on the te~t ba~ter~ on ~anu~y 19Ch. One officer from Station A ~ent into the hospital for ~urgery on ~anuary 13th and didnot return ~o du~y until l~te 3~nu~ryo Consequently, he yes trained ~i~h the control group° Since four ~eputie~ fro~ S~ation V vere dropped • from the ~tudy (see discussion ~bove)~ ~ ~otal of eightl officers vere trained at the beginning of Phase 11 ~nd' thes e constituted ~he e~per£ment~l group. Seven officer~ (i.e°, six from Station C plu~ ~he one from Station A) constituted the control group, Phase ZIX begeu on ~ebruary let at vhich time all of the co,ntro"l group ~eput~e~ ~e~e trained. ~he enperiment~l group deputies •" continued f~llin~ out forms end using the test battery during .Phase 39 " • , ,- • ." ~P~'~6T~d the two. balance teS~So The deputies viewed the t~t "md~i~i~t~tio~ a~d perf0r~ance of three ~ubjeet~ ~t a ti~e, acor~.ng .~ae~ ~ parfOr~nce a~ they ~aw it. The ~rojec t Director and th~ depu~i~ th~n discussed the scoring until there wa~~o~e agreement. The ~epe o~ the three canes w~s then repIayed ~o that the d~put~,es Could v~¢ why it ~hoald be ~cored ~he ~y it was, Then~ the 'V~¢Ot~p~ vas t, layed for the ne~t tbr~e ~ubjec~ in ~he ~a=e :~a~uer~ Thi~ pr,,,ce~s wa~ repeated until th~ an~ of the video~:~pe° ~e '~Ound ~hat the m~jority of the da~tes had little proble~ ~t~h ~h@ scoring by the ~£~e ~he la~t ~ection of ~h~ tape ~a~ plsye~, ~ho~e ~t~h p~oble~ generally knew how to score a given subject, bu~ :disagreedon ~peci~ic criteria. 'The two group~ of deputies rev, ersed training ~hen both ~ections had ~inished. Tb~t in, the first group of deputies viewed ~he videotape, ~nd ~he second group of deputies ~racticed eat{~attng angleo With the nyatagmua device. " A.t khe end Of .the a~a~on, all the deputie~ were brought back to a .ce.n~.ra! loc~t~o~ £or q~e~tions and summary state.enid. SCRI staff ~'e~be.ra then ~sde every effort to ride ~th each newly tra~d :off:icer to observe them administrating and scoring ~h,v test battery i:fi.the field. On-the-spot corrections were made ~t this ti~e ~d .all additional questions concerning admtniekration and scoring ~ere mnawe~red. Answers to questions which were .not eo~ered in the origin,al training ee~aio~ were then incorporated into subsequent. training ~eaaiona. 8inca a total of four training aessi.o.s were gi.ven~ during the field evalua.tion~ .very few aueatioas remained by the itime the fourth eea$ion ~as conducted. Durinlig baseline da~a collection (ioe.~ Phase I forths experimental. group . e nd Phase I and I~ for the control group)~ officers filled out the da'ta forms indicated in Table 16. For moat stopees, o-ff'icer:~ were only asked to fiil in basic information contained in t:h:e 'top half of the form. Th.us~ they might check that a atopee :was a 25 yea~ Old, Blsck male, ~ho was stopped at 2235 hour~ o,n e ~:~dne.'~day for ~peed'ing on s re~ident-ial city street. The re~t of the ~orm ~ould be ]eft blank unle~a the officer OU~l,,~cted that th,. b%op'~e had bees drinking or taking drug¢, in ~hich ca~e he wo~uld 'make :the appropriate check ~ark on the fo.rm. If beh~vio.ral te,st,s •w:er.e given~ ~hen She officer would indicate the nature of the t. eists ~:nd "whether or not ~he etopee passed each teat. If the stopee ~as arrested, then the type of chemical ~nalysi~ w,aa indicated, the ~AC ~Qs' z:ecorded~ .and th¢ officer checked, ~hether She suspect ~as re'lea~edior ~booked. ~°f blOOd ar urine was tak'en~ then the fluid was sent So .the $'heriff°s Forensic Crime Laboratory for analysis. Often resn,lts Wj~ul:d: not be availabie for four-to 8ix weeks. Deputies ~ere asked tb' put a file number (ioe , the police cave number) on the fo~ ;f 4,,2 .f., i: I III. Phase Ill ended o,n February 16~h for the six Staltiou • deputies, as a number of them we-re transferred ~o new as~ignm( ~s • t this ~i~e. ~he r~aining ~eputie~ continued to-collect ~t~ until February 2~th. - ...... C. TRAINING P~LICE OFFICERS The deputies were trained in smell groups during half day sessions. gach deputy yes given ~ training manu~l, similar to the one used in the laboratory evaluation. This training ~anual covered the history and purpose of ~ standardized field sobriety test; the meanin~ and importance of the nystagmus test; administrative procedures~ including conditions under ~hich the tests had to be ~dministered to be considered valid; ~coring procedures; and decision criteria. The Project Director reviewed the reasons for a standardi~ test battery quite thoroughly so that the deputies would ~hov as little resistance an posslble to learning and using standardized.~iacorinE and administrative procedures. This review included the fact that: (I) If every officer scored and administered the test battery. :.in the sqme yay n theq every officer should_ get the same score for a gxven ~ntoxxceced nrxver. As a result, the test battery scores would be note meaningful as court evidence and would als0 allow police departments to collect their own data and develop norms. (2) General acceptance of a given test score by the courts as indicative of impairment could also help officers in filing ,drug charges for low BAC caces~ since the test scores would still show that the ~topee was impaired, .: ~ .... The Project Director then reviewed the meaning 'and importance, of the nystagmus test~ covering various signs of intoxication that .~can be seen in the eyes. The officers were informed of theoretica! speculations about the reason that nystagmus occurs under alcohol and the differences between Alcohol Gaze Nyst~gmus, which •appears to be neural in origin, and Positional Alcohol Nystagmus, ~hich'.is vestibular in origin. This information is given in the llterature revie~ in Appendix A of this report. In addition, the"bfficers were informed of other potenti~l causes of gaze, nystagmu~i: (e,g, , drugsp brain damage, etc.). .~. ' The deputies were then informed of ~hat to look for in theeyes,~ in order to determine whether or not to arrest a stopee (see gaze nystagmus section, Chapter I). Half ~he deputie~ present then went to another room ~here they ~ere informed of the importance of e~timating the angle of onset of nystagmus and practiced estima~.ing 33, 40p and 43 degree~ using the device pictured in Figure l. Officer~ working ~,mo ~htft~ ~ere told to u~e 35 deg, ree~ :as .a criterion, ~hile p.m. shift officers ~ere told to use 43,degrees as a criterion. Officers ~ere required to practice on each~ other until they could estimate all three ~ngles on e~ch other ~ithin three degrees on three consecutive occasions ..... : The other ~alf of the deputies viewed a videotape ~n which subjects ! • • ,L 41 " ' i /" / , i ~ r' TABLE 16 PRE-TRAINING DATA FORM DRIVER F M Age Anglo Black : Mex. lumer. Oriental Other D~: M T W Th F S Su .Hour: Type of Duty: Location Reason for Stop City Street: Driving too famt/slow Accident Residential Driving on inappropriate area Weaving/driftinq Business -- Nearly striking car or object-- Wide radius turn OthQr Freeway Stops in lane without caus6_____Looks intoxicated Rural Not in marked lane Zqqipment violation Other Ran stop sign/light Driving too closely Bright lights/no lights Assist other"office~------Other Roadside Station Suspected Alcohol Drugs Chemical Analysis: Behavioral Tests: (Specify) Brsath BAC % Pass Fail Blood -- Urine Pass Fall Re f us e-'~----- Pass Fail Booked Released Estimated BAC % Driver's License# Arrested Released PP-20 11/79 SCRI 43 TAI~LE 17 DRIVEL~ M F Age Anglo Black Mex. Amer Orlental Other Eye Probs Contacts galance Prohs Type of Duty D~: M T W Th F S Su Hour: ~ Rea§on for Sto_~ City Street: Drlvlng too fast/slow Residential Accident Business , ~Drivino on inappropria-~e -~rT~---- . ~vlng./~~_ ~ Nearly~striking car or .Object~ h'e~ Other . . [Vide radius tur~. " " Freeway Stops In lane without cause ~ Looks intoxicated...... - Not In marked lane -. Rural Ran s~op sign/ligh~-'----- . ~ Equipmen~ ~'ic~a-i,~. .... Other -- _ Br±ght lights/no li~-~~ AssistDriving Othertoo ci0sei..--?~ Ic roadslde . _ -~-. - Statlon ...... ~-- . - .... =- Suspected Alcohol Drugs Chemlcal Analymas : Behavioral Test Scores: : Breath BAC % Wal k- and-Turn Blood -- -- ...... Ohm- L@g Stand 'Ur Ine--~ Ny~ t a(|mus (A(;N~_ R~ f'*O~------'-- Estimated BAC ~ Bo()k~d Released Arrested~ Released Drl.v'er's License # .... ~Sheet ~or ~T Batt__~ Walk- and-Turn: Cannot keep balance while listening to instructions Starts before Instruct~ons are finished Stops while walking to s~eady self . _ Does not touch heel-to-toe ' LOses balance while walking (i.e., steps .off line Use~ ar~s ~or balance Loses ~balance while turning I'ncorrect number of steps __--, Cannot or refuses to do test {equal to 9 checkmarks.) ---~- 0he-Leg Stand: .... Swayi.ng while ba]ancl.n,! U"es arm~ to balal,-e Qu.lt~ .unsteady ...... ,. • ,Pu~tm foo.t ,dr~wn ...... ~Can~,ot Or r@f,uaes to do test (~ual 1o 5 checkmark~,} -- ,- ,Aico~ol Gaze Ny.stagmu s (ACN} . --~--- RIGHT EYE LEFT_ .EY'E ~Onset of AGN a~t less than 450 and at least I0% of t.Me white showlng - :Estimated .angle of Onset i'Ey,e~ cannot follow smoothly --..... ;AGN at maximum lateral deviation: ---- ~Absent R L Minimal R L Moderate R L 'Heavy R L .AGN at .~xlmh~n latera~/ devlation is moderate ~)r heavy .- % ~lood or urine was taken so we could obtain the results of the analysis. The data on eeveral arrears during Phase I wer~ not available to us because the depnt ie~ forgot to include this information. Probably more blood sample8 than • normal were taken during the course of this study because the Sheriff's Department switched from using the Intoximeter to the Intoxilyzer at about the same time the field evaluation began. Many deputies were ,~nfamiliar with the operation of the Intoxilyxe,r. After the deputies were trained in the sobriety test battery, :.they were asked to fill out the forms given in Table 17. This form ~is exactly like the previous form except that it includes a scoring sheet for the three test battery. Thus, when giving a field sobriety test~ officers were asked to check the probiems the stopee had with each test and record the number of checkmarks for each test and the total test score. Officers were not required to identify themselves on the data forms before they had been trained on the test battery. Thus, an officer who frequently released drivers he or she suspects to be legal~ly intoxicated would not be inhibited from indicating this on'his/her data forms. After the officers were trained, however, we requi.red them to initial their data forms so that we could determine 'if any* of them were having difficulty scoring the sobriety tests. .*In addition, the officers ° initials enabled us to identify each' officer's pre-trsining data forms. Only one officer seemed inhibited by the need to identify himself, and tended to fill out more forms after we requested that the forms be initialed. • One problem that arose in filling out both data forms was that most deputies waited until the end of their shift to fill out t heilr forms. At this point in time all forms were completed at o.nce;from their police lOgSo We urged the deputies to fill out the for.m s immediately, but our urgings did not help as most of them cou~tinued ,to fill out the forms at the end of the shift. Ne then stressed the importance of filling out forms for suspects given sobriety tests, so that the tests would be properly scored. We doubtthat most officers complied with this request except when observers :were in the car. Two staff members from SCRI rode with the participating de purl, uS, throughout the field evaluation, The two staff members 'included the Project Director and one of the observers f.rom the laboratory evaluation. One staff member rode wi~h each deputy one or t wb times during every phase of the field evaluation. One purpose of the ridealongs was to obtain feasibility data on~thie I •sobriety test battery, including the deputies' attitudes sbo.u~! arresting intoxicated drivers, their ability to administer snd~ score the test battery at roadside, and the reaction *of the'.stoPees~ to the teat battery. Some of the deputies were a little :nervous about having an observer with them at first. But they were t. old to do everything they normally did and pretend that we were not in the 45 , [ 'i i . i, ~ ~•• ....DEVICE FOR OBTAINING ANONYMOUS BREATH SAMPLES 46 car. By the second or third ridealong, none of the deputies se ~d to be influenced by our presence. The second purpose of the ridealongs was to obtain breath sample~ from released at.peas. Various police agencies sere concerned (1) about the legality of the.police officers knowing the BAC of a released at.pea who might be legally intoxicated;, or (2),t,he possibility that a released atopee who was intoxicated a-ight later crash his car and then try to sue the police for not arresting him, Thus, an anonymous breath testing system was designed for- dse in the field evaluation. • The device used is illustrated in Figure 6. It consists of an t ALERT J3 Digital Breathtester, mounted in an enclosed box, with a camera. Openings in the box allow the observer to operate the breath tester and the camera, but both the J3 Digital readout and the camera viewfinder were blocked from view by the locked box. Each time a box was opened or closed, it was ~ealed and the time and date were recorded by a notary public, No information was recorded about any of the stopees by the observer. T:he onl:y: information that was recorded were the first and last numbers of the film each night. • Thus, the only data obtained Were distributions of readings by the J3 Digital for •each deputy ,during each phase of the study. The 33 Digital was chosen because of i*t;s ' smell size, its relative accuracy, and the fact that it ,Sas~not been approved for evidential breath testing in the Sta~e of California (i.e., the manufacturer has not submitted it to ~•the state for approval). ~ • ~, , •:~ : Police officers talked to all stopees before anyone was approached by a 8CRI observer. Once the officer finished writing the citationp he or she ~sked the at.pea to get out'of the car to "sign the citation. The deputy was instructed to inform the at,pea, once the citation had been signed, that an observer ~as in his/her car from Southern California Research Institute who was doing ~research: for the U.S. Department of Transportation. The deputies were then asked to say, "I would like you to talk to the observer, but your cooperation has nothing to do with the ticket you received " Individual officers frequently expanded upon this statement~.by explaining that ~e would require a breath sample and indica-t~ng hQw their cooperation would help the police. Officers were requested only to ask stopees for their cooperation once they were certai'n they ~ere not going to make an arrest. ~ We estimate that police officers asked approximately 77,.5 %'~ofthe ~topees to cooperate (see Table 18, Chapter IV). The remaining 22.5~ consisted of arrestees, people involved in accidents,, : p eople~ the officer forgot to ask or didn't have time to ask because of an emergency call; and people the officer refused to ask (i.e., "Oh' I didn't ask him because I knew he wouldn't cooperate anyway" Or "Oh, he was a police officer just getting off duty, so he didn't have to do it" or "He was a friend of mine, so I didn't ask"). I~ the officer ~sked for the ,topees' cooperation, then the : etop ee usually would talk to the observer. A few notable exceptional refused because they were extremely hostile about ge.tting a. ,'... 47 • L i . '. "i, " , • " " " "', t ~ ~.~,~:~ b~PARTM~'T o~ TRANSPORTA_@iO~ A~ PART~ ~ :'t STOPPEr BY ~sIS 0~ic~R T6~IC~T ~6 ~SOW~ ii~To Z~Z ~oUTB~I~CE oF ~'~is ~oz, As ~OUo ~0~I • ~Ot~ ~SOVZ IT V~ii ~BZ ~H ~.. ~"i~~ti~ 6f .th~ ~opeeo Often ~e ~ould ~i~vd ~O sn~~ add~ha.~ que~t:t6~i; ~6h as: J '~%Why are y6u doin s thi~ research? ~S~wER: ~0 O~AI~ A DiSTRIBU~IoN OF ALcoHOL 'READI~CS O~ PEOPL.~ ~S~T(}PPED TONi~? THAT THE OFFICER HAS DEC~DED ~0T TO ARREST° ,, ~6~,'How ~oe~ tlli~ th~ work? ~eaning the anony~aou8 breath test '~NSWER~ ~, D-~ BLOW I~TO THIS ~OUTSFIECE WHICH OPERATES R POR'TABIL'F. IB~EATB T~E~$~,'R~ LOCA$'EID HEREo AFTER A~OUT 20Ua SECOND'S~ T'H!iS L~ZG~T IT~' ~5-@:EV~.R,~ T~S CA~iE'R~ IS POINTED TOWARD "T'HE RE~AbI~S,~ SO 'I WILL "SU!~T T~k ~. ~ PICTURE 'OF IT o ONCE THE FILH IS DEVELOPED, WE WILL :~:~';6~W'~O'H:kT 'T~:E "RESIDING IS, BUT WILL ~0 LONGER K:NOW WHO YOU ARE. %*~:~R b~ S ¢6ub]e Of dr£n'k.s tonight~ ho~ do ~ know you ~r.e tell£hg ~T~'~ STS-TEM"~D • EHP~ASI~:ED THAT T~E SOX ~A:S SEALED, SO THAT ~E :WOU:LD ZI~DI'C~ATED T~T '~BZ ~RE~TH TESTING DEVICE W~S :~0~ APPROVED ~BY 'THE ~s STATE, SO THAT THE READING COULD NOT BE USED IN COURT. . o ,...Will you send ue the ~s~ul~s o~ ~ht8 ~e8t? ANSWER: NO, WE WILL NEVE~ BE ABLE To ASS0CIATE ANY PARTICULJ.R READING WITH YOU. Approximately 85Z of the 8topee8 who were naked aareed to prowide .us with a sample. Most of the refusals were people who were !still very hostile •bout getting • citation, slthoush approximateI:y~ 5Z Of the refusals were people (usually female) who claimed it .was too emb~rasetng to be seen ~iving a breath ~ample at roadside. In every case,., whenever s suspect showed some hesitancy by admitting 'O to drinking, we were able to convince them of their anonymity and .obtain a breath sample. Occasionally. admitted drinkers would not blow hard enough to enable us to obtain • valid sample. :After three bad sauples ve stopped requesting additional blows. People involved in traffic accidents were never asked to provide breath samples. Thu~. we avoided the possibility of having civili suite brought against us or having our data subpoenaed. 1. r .. . ' 49 • F.ifte~ ppli~ Off~c~r~ ©o~plet~d th~ ~eld ~valu~iow, fi~!~u~ out offiCe~e, d ~5.5 ~i~ht--bdUr ~hiftm in torsi ~ during the dy, calculated ~h~ ...... ' ...... - ...... ~ ..... n.u~ber of trcff[c ~tops per Cide~Iong, defining a i"t~'Ai'fic ~o~ a~ o~e for ~hich ~ form should ha~ been completed •,...... :. o Thi~, e~ti~at~ m~y be ~l~ghtIy inflated, ~ince ~ome of •.-~he office . p~obably ~re making ~ore ~op~ th~n ~orm~l during the that . • • ~ev~r. ',~c e~-imate0 u~ing thi~ conservative figure, ~eput[e~ filIed out form~ for ~,ppro~i~ately 6.5 IZ of the ~tops lfor ~hicb they ~hould h~ve completed d~ta fo~:me. Four offic~r~ 'f£!!~d" 9~t [Orm~ at ~ rate o.f le~ th~.n ~-OZ Of.~hmt uhich ~e pToject, ed ~frem" the ridealong~. Ba~ed upon d~cua~io.ns ~ith th~ :~ q.r £o.u. ~ traffic 8er~ant~ ~ ~e [eel ~ha~ ~Ur data are ~ery :inco.~p.lete for three ~f the~e deputie~ but ~hat the fourth deput~ m~de ~ore sto.p~ thnn norma! during the ride~long~o T~he ~epu~ie~• ~de ~13 traffic~ ~top • ~uring the 59 ridealorng~o A breaR"d~.~n18 fdr "~ eaCh o.f"gr6up the d~a ~vailable from the~e s.top~ ie given in Tab:~e ...... o£ officer~ during each phase of the evaluation, ~In ~um~ary~ 6 5~ of ~he ~topees uere a~reeted durln~ each of the ~r£d, ealo~ng s~e.s.~ion~ (ae compared ~ith 7 ~Z of the ~topees for ~h.ich ~.e have d~ta fo~r.ms).o ~noth,¢r 6.8.Z of the ~to.Pees ~'ere invol'ved i~ tr~ff~ c .m,cc. id-en~, bu.t not ~rre~ted; 9~o2'~ ~.~ere not. asRed hy ~he o'£.~-icers, to provide breath ~nmple~; ll..4Z-~er.e asked to p.,rov£de bre.ath sampl~ b.ut r~fu~ed; and 66~1Z of the etopees pro.vialed •a.,no.nymous b~resth ~ampleso Thus, ~e have BAC infor~,ation ou 72 6Z of ~h'e etopeee.--those who were arrested and thoee vhe vo.luntar£ly :~rO~i.d.,:e~d~ @~.~m..pl.e%, Among the rele.s~ed otopees who ~ere a.aked to p ro~v,£Ge breath ~ample~, 85 3~ agreed. The majo.rity of the re~fusal~s ~iSid they would nat coope.ra~ because they ~ere given a citation The&ej data ~:e:re, an~.lyzed ~ith regard to three b~sic issues: (.i) W:ha:¢t ~'~ ~ Zh:e.~ nature of the. stOPee population?; (2) Is the treS, t b~atte.ry e.ffective?.; and (3,): Is large ~cale implementation of the tesot b.~,~ttery feasibl~? One o.~.,~ ~,he., o,b.j,ect~ves o~ the field. ~tudy ~as t-o det:ermine the u~ture of the. ~,tope,e po.p.ulation. The police, data forms, ~ere d,~.igned~ ~ith this, oh.jecti~e in mind in thq! information ua~ requee.,ted on the age, ~.ex, ~nd race of e~ch ~:t-o:pee~. Da~a on the ¢~h~ract~er. is,,tice of th.e ~topee POpul~tion, . derived,. £r.om the 3]28 rio,rinse completed, by., the officer~ ~e.re tabula.ted. Gi.~en that the. ~f,i:f,£cer~ d~i'd "~0"t f:ill out forms on all of their ~topees the data ~ay, be. s~o~ewbat, bia~ed. For e~ample,, certain officer.s, filled out ~,any~,,,~.~ more-=,...... forms , than. other officers,_ . ~o their bias.es~ if an.y~ could b~e reflected- in the ata pre~ented ~n thi~ report. ~owev:er. our ~.p~ulation (i , arr~ et a l'., 1980) 50 2 , i" /: TABLE 18 DATA OBTAINED FROM STOPEES DURING RIDEALONGS PHASE 1 PHASE II PHASE iII C0ntro___.__~iExperimental Control Experimental ~ontrol ~ Traffic Stops 78 I01 62 71 48 '53 Accidents 8 7. 1 2 7 3 D~I Arrest 5 6 4 8 2 ~fficer did not ask for breath sample 9 5 2 7 5 10 Refused to give sample 3 13 7 11 7 . ~ Gave breath sampls 53 70 48 41 51 \ ~he a~f di~tr!butios s of four population ~amp!~e are g!~eW iz~ T~b.~e ~p~ted of consumin.~ ~Icohol or drugs; (3) arrested slopers.; P~ople~invclved in accidents during the study ~he.. ~.opee~.,..,..,, as, a .~hole:. . tend, ~o be younger than the i'eople involved ~ ~,, accident~~,,~ or. the DWI arrestee~. Tho~e suspected of con~u~ing alc°h°l:~fa1i"~bet~een the stopees and arrestees in terms of age ~o~ev.:.er.~,,.~ , ~ ,',, for;., el! four groups the mode fell into ~he 20-26 year old age gr.ou p. ~.4op, le,over 65; represented only 1.5~ of the stopee~ and only one Person.~.,~....,~,.. in, ~b£~ age range ~.as. .suspected of: consuming alcohol prior ~01;,. , dr;~ving.,,,,., , '" P.'.eop[e 'over 60 constituted 3.4Z of the stopee ~,o~ulat~on~::,,,.,~,.~,,~ ....,, ; but accounted fox 7,.6Z of the accidents. T.&b,,le.~.9~.~,~ ,~.... , . a~o.,.,.,~ ~ndicates..., the. sex distribution of the, same -¢eu'.r ¢.a:teg0rie.s,,, :,. , . , o, of. Btopees.,.. . The 3128 stopees con, sr.iBted of 23,29 (74o.5Z)~ ma1'esl....~ " ': '" ~ and~" " 7,9.9;' (25 " 5~) females. Males in,. thi~ delta may be o~e, rre~.presented since ma, l,e officers, (only; o,ne deputy ~,a~ female)~ ~h~o~,ed, a ~,ligh~ ten.dency not to give f-emales t,icke~s, ~hich ~o,u,~d b~e; ref~lectedi in. the; number of forms, co.m,p~l,eted! fo~. ~ema, les,. O,ne,~,~,,: f e,m,ale,.:, ,, ,. out...... o,f every 19.0, female st ope.e~ ~,a.s, s~.uspec~ed. o,f.-'. co, Diaum,~,ng alco;ho, l vrior to driving,, a~, compared~ w~'th one, male, o,u:,t o,,f eve.r,y~'6~.8, ~al.e, stopees,. 'Thus. those ~,uapec,ted, o.f d.rivin, g~ after dlf.in~in~ Cons~slted~ of. 342 malesl (89.1~') and 4'2 females. ~l,0..gg)!. I~f.:~, a~.:~£e,m~a,],e,..~ , ,,,, , ;- ~a.s.,, , ~:uspected~,.,. of DWI, then,. her chances, o£ bei:n,g... ar~e-e~ted~ ~s~,e,~e ~.li-~htl.~, le.as than that o~ a mal, e suspected of DWI: O,~.-.,.:.~h,,q:.. ,..., 4,2~.. ft6m,a,Z~.e,~:'slns,pe'c:t ed. of dziving after, drin, king ,. 2] ~i.50; ) ~e;r,~ .arr,e,s, te~d~.. 0,f the. 342 me.lee eu.spected of d'r.iving af~e~ 4~~n.ki,.ng., "ii9~41:'(56.7. Z). were: a.rr.ested... Th,e DN.,I-azre,s,te.es ~e,re 9!Oo2'Z ~~ ~'I~ i~z~'~'3'~Z'.~'~.~ma~I~e- ~H.o~ever'~"~only 52 a.ccidelnts ~.ere repot, ted ~n our T~h~e% dat, e Q,n: t.h~e~ r, a cial~ make.up of the st ope.e~s~; m~y, be the mo~t bissed, o~ a~1~I~ of I. tke~. po,p,ulation data in. the, fie~I.d~ 8tud,y~. The cities, r~e~x~e~e~n~t.,ed~ in, the f0ieLd~, evaluation tended to have m,£norit.y m i~ ~ 'o;fjl"h~i-a~/h, elr~. is,tiOp.ee.. ,0ul.d.. be mino~ri,t.ie,So,. Thus, the tendency ..,.::,..Ji~i' c"e,rtla,~n/01ifi'Cer:,,~,,.,, ,,...... ,,...... to. fill, out:. man,-Y-, Bore,fo,ms. ;, .,., than' ,oth~ers could. O,u,~<,j~,, s,am,,p;,l',,e~:~.:~.,,~,, ,:~.o,f-~., st~o,pe;e~s~.., ', ,.v' '"-' co,~ns~ie;ted,.. of. 53'.3Z ~., Ca,u,ca, sian a,. pr ima;ril y, be,csu;se,,,,.~,,,,:~, ,,:~° t~o,.,,, o/f,.... the,.,.. t,h~,r-ee~,ci.ties~ from w,h:ich, m0.s,t of o,u.r~, d,a:t,a,; cam:e, ,2. ." I TABLE 19 AGE AND SEX DISTRIBUTION OF FOUR GROUPS OF STOPEES DURING FIELD EVALUATION SUSPECTED DWI INVOLVED STOPEES ALCOHOL ARRESTEES IN, OR DRUGS ACCIDENT N 3128 396 215 52 15 0.3% 0 % 0 % 0 % 16 - 19 17.2% 9.7% 9.3% 11;5% 20 - 24 24.5% 22.6% 15.8% 17.3% 25 - 29 16.6% 16,.3% 15.8% iI.4% 30 - 34 11.7% 15.4% 15.4% 17.2% , 35 - 39 7.3% 8.8% 13.6% 40 - 44 6.0% 9.1% 9.3% 7~6% , 45 - 49 4.5% 7.3% 8.7% 3~18'% ... 50 - 54 4.8% 5.6% 6.2% 17 .,3 % :. 2. 55 - 59 2.0% 1.1% 1.5% 1.9% 60 - 64 1.9% 2.1% 3.3% 5.7% 65 - 69 1.0% 0.8% 0.9% 0 % 70 - 74 0.3% 0 % i 0 % 75 + O.2% m 0 % 1.9% Missing 1.7% 1.0% 0.9% q % Male 74.5% 89.1% 90.2% 82.7% Female 25.5% 10.9% 9.8% 17.,3% ~, 53 -cowO~i,~ c e~ of largely Caucas i~m,, populations. B l:a~ck8 ,, La.ti as:,, O'r~entale, ~nd other ~inoritie.~ constitut, e 19~.OZI, ~7,,8~ 3!o9~, end i3-o3iZ of our o:opees, ~~p~ctivel, yo Xnterest~ngly, Caucss~'nw and" Latiu¢ vere much Wore likely to, be gu~pe, cted o'f " concum~nS~ &tlcohol ~,efore dri,w, ing r than B:!a'c~s ~ o,r Orieu:ta!~. ~h'e r~ ~ere one of 6,.,6-s~opees for Caucasia.n~;• one o'f 6:o8 stop e~s for L,~tlns; on~ of 17.4 mtopees~ for Blacks:;i and; one of 24.6 ~topees for Orien..~_al~,, Once a stopee ~a~. suspected. O~.f DW,',,~.~hovever, ~e fouud no grea.ter ten~en~cy for deputies to arres, t. ~c:y One zroup than .~ny o.ther. TEST BATTER'~ ~ Z-FFECT%VENESS The-,:::most crucial que~on~ t~ be ~n e~e~ ed: during the f ie 1-d~" ,~va.!ua~ion vf ~h.e ~obriety tes~ battery ~=nClude: (I) W.,i]~~ ~h~e" per:cen~age of ~topeee arrested increase, after the: test. bat~er'y' i~S. ,~.n~:roduced7 (~)" ~iI! ~oqic~ officers ~ake ~Ore accurate d.eci.s'~ioa~s wi~h re~pe¢~ ~o ~ ~AC O~ 0o~0~ ~fte~r being~ ~-~ained on the tes~ e~a,tery? (3) ~7ilI ~he lean ~AC of a~rrested drivers be reduced a'f~er ~he ~e~t b~.~ery is incro.duced? (4) Will police officers ~or~ .-.~sccurece1~ e~ti~,ate the B#.C levels of 8topees after ~eing ~ra'ined on, ~he. z~ b~t,~ry? ~5)': ~u, ad'~d'~.tion,. ~he ri, d:eal:ong date m~h'o.u'l~d~ provide ~n e~tim~te of the percen~tag, e of police stopees, as ~ opposed, co dr~vers on the h'ighw~y ~ho have been~ drinking and' vho" ~re!:egaily i~to~ic~ted. 'lh/ . ~answerix~.E ~he~e' questions, both r'i.dealong dat- and off:icer-completed: ~orm~ ~re ~vailable. The ridemlong data ere aS Co'mplete as possible mad provide ~AC distributions of released~ ~topees. ~owever~ the ri~eslung dat~ repreeent only a small sample of-.the ~rivers stopped by the participating deputies during the f-ie:Id evaluati, e~-,o In addition, these data may be somewhat b~ased~ 'beca:use.. an observer was present. The officer-completed form'w, o:n'. .the. Other hand, cover the entire field evaluation. Howeve'r, Che~e dat~, e~e le~s complete an~ do not p~ovide actual BAC information on~ , r eleased ~to~eee. As discussed before, the biggest problem with the field evalu~tid~n ~ was officer p~rSicipation. We began with 20 deputies, but had ~:o eliminate five because Of poor attitude or lack of cooPeratiOn~o Three of the r~maining deputies filled out very few data forms~ (ie~ than 40% of their pzob~b!e stops) and a fourth deputy made no DWI a,rrests 4urine: ~he' en,~ire field study. Thus~ out of" tl%'.e' or~gi, na! 20 depu~ies.o only 11 provided', u~ wit,~ ~ufficient azrew~ dat~ ~o be of value. Even among these !I officers, there was c0ns±~er~able variat~o,n: in the number Of arrests made. As a result, trends are repo:'~ed ~ but the - d'ata are not appropriate for s~gnxf-~c~nce testing; ~he assumptions, for underlying statistics w h'iCh ~ou1"d be of interest are not met by the data. However,, v~z~ue!ly ev~ry ~rend zepoz~ed i~ in the direction of improv'ed ~ iperfcr~ance re~u!-.i~g from ~he tes~ battery. The potential utility" O,f th~ ~e~ bat~.,ezy appears to be supported. ...... 5¢ on t,Jm ~ ~AtteryT By examining the procedural steps in the officers' handliug of the intoxicated stopee, we can anticipate how the test battery might increase the percentage of st,peas who are arrested. ~ Many intoxicated drivers, especially those vith a high alcohol tolerance, probably are never stopped by the police because cues for detecting them are not sensitive enough. Instead, most of .the 8topees will have made serious driving errors. Many of these driving errors may he attributable to impairment other than alcohol intoxication, such as a woman who has just had her purse stolen and is too upset to concentrate on driving; a diabetic person in need of insulin; a married Couple arguing; an elderly man driving too ~caregully9 etc. These people generally are not given sobriety :teats, because they do not smell of alcohol or because their other ~problems are obvious~ / ~If the officer detects an alcohol odor, then the driver probably .will be asked to get out of the car. Once this occurs, the officer typically w£11 continue a low-key interrogation of the st0pee .and administer behavioral tests. The officer then must make a decision to arrestor release the stopee based upon his/her estimate df hdw intoxicated the driver is. Unfortunately, the arresting officer'S decision is frequently based upon personal factors (see feasibilit'y section), rather than upon the estimated BAC of the driver:For example, during the field evaluation, approximately 5Z of the st.peas ~uapected of drinking alcohol were ~eleased despite, the fact that the stopee's officer-estimated BAC was over 0.10Z. These cases included four ~topees for whom the BAC was at least 0.20Z,:.~as estimated by the officer. t The average police officer does not, under any circumstances,~ wlsh to arrest a suspect ~ith a low BAC (i.e., below 0.10Z) and will often err by opting to release rather than risk a false arrest, The test battery probably w£11 have its greatest impact at this point by increasing the percentage of stopees who are arrested, reducing the false negatives. • J • . Table 20 gives the number o~ stopees, the number of arrestees, and the percentage of st.peas who are arrested for both group s,~bf officers, control and experimental, during each phase of the ~'field evaluation. A larger percentage of st.peas might have *been arrested daring Phase I because of the number of drinking drivers on the road during the Christmas-New Years' Holiday Season~ Indeed, the control officers arrested 6o6X of their stopeesduring Phase I, but only 2.2Z of their stopees during Phase If, The experimental group officers, in contrast, increased the percent:age of st.peas arrested from 7.7Z during Phase I to 9.1Z after,thei~ training in Phase II. The control group also increased their arrest percentage after their training from 2.2Z in Phase II t0 5.0Z in Phase llI, During Phase llI the percentage of arrestees dropped from 9.1% to 8.2% for the already-trained experimental group officers, but remained above pretraining levels. 55 J TABLE 20 STOPS AND ARRESTS MADE DURING THE FIELD EVALUATION AS A FUNCTION OF OFFICER GROUPING AND STUDY PHASE - ., .• CONTROL OFFICERs, EXPERImeNTAL OFFICERS % STOPS ARRESTS STOPS ARRESTS % PHASE ~ I 73'2 48 6.6% 775 60 7.7% Tra£ning PHASE II 319 7 2.2% 502 46 '9 .. ]'% Training PHASE I~I 359 18 5.0% • 441 36 8.2% 56 i ~ "'L When all of the data are cla~oified into trained versus untrained periods, the officers arrested 6.3Z of their stopees prio :to training and 7.6X of their stopees after training. This xepresents a 20.1Z increase in arrest rates which could have a substanti:ai effect on DNI arrests n~tion~l'ly if a large number of traiued officer~ ~ere to maintain Quch an increase. 2. ~ pollce 9~ficera ~ more accurate decisions with res vec_~ t.oo ~ ~AC of ~ after ~ trained on the test battery? The finding that police officers arrested a greater percentage of their stopees after being trained on the test battery could result from: (I) an increase in the exposure of the deputies to drinking drivers as a result of their training on the test battery (e,g., officers might seek out intoxicated drivers by staying near bars or they might alter the type of stops they make, both of which might increase the percentage of their stopees who were drinking)~; (2) a change in officers', arrest criterion after training due to increased confidence in their ability to make accurate arrest decisions; (3) pressure from superiors to perform well after they had been trained; or (4) a desire to make more arrests because they had just received training in field sobriety testing (i.e!., the Havthorne effect). . : ~ The BAC data obtained during the ridealongs may be biased. ~These data, as discussed earlier in this chapter, represent only 59 eight-hour shifts out of 685.5 shifts worked by the deputies during the three month study (i.e., or 8.fiZ of the shifts). In add£t~on, deputies may have beau influenced by the presence of an observer during the ridealongs and BAC information is available on only 72.6Z of the released and arrested stopees (although 85.3Z, of :the released stopees asked agreed to provide breath samp l,e s:),. l~evertheless, the BAC data from the ridealongs is the best:, data available to determine (a) if the deputies vere more exposed ,to drinking drivers after their training or (b) if the officer:s ~ were able to make more accurate decisons after being trained on the tea:*. battery. ' :~". Exposure t__oo Drinking Drivers Table 2l gives the number of ridealong BACs collected for each group of officers during the three phases of the field evaluation. The percentage of drinking drivers and legally intoxicated drivers is also given in the table. Clearly, our limited sample of BACs indicates that officers were not more "exposed" to drinking drivers after training than before training. Drinking drivers constituted 35.2Z of the :. before training sample of 125 BACs and 34.7% of the after training s,ample of I01 BACs. Legally intoxicated drivers constituted 18.4% of" the before training sample and 14.9% of the after training sample. Thus, the officers, if anything, are less exposed to drinking drivers after training than before -- primarily due to. the high percentage of drinking drivers (i.e., 41.9%) among police ~stopees during the Holiday season of Phase I. .. b, ~ccuracv of ~)e~s~on.s Table 22 gives decision matrices before and after training for the ridealong stopees for whom a BAC ~.s ! 57 I, " '- . TABLE 2i ~A~ b~ ,~LE~E~ S¢OPEESAS A FUNeTZON OF OFFZCE. ' .... ~a~z~¢ ~D P~SE OF ~E STUD~ .... ~NTROL OFFICERS EXPERIMENTAL OFFICERS ,~ %.DRINKING % > 10% ~BACS % DRINKING .% ~ .10% 43 41.9% 23,3% 43 41.9% 16.2% TRAINING ~H~E ~ 39 2'0~-.S~ 15.3% 4'9 - 34.?% 20.4% ~ 13.3% 22 40.9% 3.5% "-. , r TABLE 22 I. BEFORE TRAINING DECISION MATRIX Release Arrest BAC Z .zo~ 8 13 21 BAC < .i0% 104 0 zo4 -. 112 13 ' 125 II. AFTER-TRAINING DECISION MATRIX Release Arrest BAC -> i0~ 4 .9 li •I BAC < .i0% 86 2 ,! 88, • 9O !I! ii 101 ~ 59 a~du~te declsi0ns ~ith resp'~Ct to whether bt.op~@S, were ibOve o,r' b~io~ S BAC of 0.101 after thei~ ireining on the fluid sUb,i,t# ~@'~:~. Before tralnihg the deputies correctly arrested 61~9~1 6f the s~topee~ over 0.10I, but "imprOved to 69.2I after traihiug. Overall, 93,6~ of thelr decisions were correct before training d:hdI 94.11 of their decisions ~ere correct miter training. The-decision ~atrices indicate that the likelihood of a ftls~ ~ positive decision i~ extremely low (less than 21). Thus, wit~ ,field ~obriety test training the officers appear to be willing to' 16wet their criterion somewhat, but not enough so that there is any substantial change in the number of false positives. 3..._~ "~ t~e mean BA___CC of ~rrested ~ be reduce~ ~_Lf_y_ the b:a~terv i__ss ~n~'oduced? SinCe ~ borderline BACs produce most of .the decision errors, those who ~are now arrestedoften have high BACs about which there was ho uncertainty at the time of arrest. For example, the nationwide me~e. for DWI arrests ie 0.172 (NHTSA~ 1974). However, since there are many more drivers on the road with BACs in the 0.101 to 0.151 range~ than et higher levels, a test battery ~hich provides more Oertainty mnd produces more arrests in this range Should ~Ubd~tantially reduce the mean BAC of arrestees. Data relevant to this issue ~as obtained in a DOT. study of portable breath test deVi~ed (DOT-HS-891-161, ~inal Report, 1974). The investigators reported that the average BAC for DWI arrests in their county~wide areas was 0.1791 until 13 portable breat~ testing units were introduced at which time the average BAC dropped to 0.14~. A ~ensitive behavioral test battery should also lower the mean BAC of arrested drivers. ~e examined the BAC data of the DWI arreStees obtained during the ~hree:, month field evaluation. This information was available b~ 178 out of the 215 arrestees. BAC data were not available Off-32 arrestees who refused to submit to a chemical test for alcohol and o,n five Phase I blood tests that wore unaVailable to us. Table 23 gives the number of arrests, the number of available BAC~ and the mean BAC for each group of officers during each phase Of the field evaluation. These data suggest that the use of the tek:t battery had no effect on the kverage BAC. The mean BAC bf the a rrestees of the experimental group officers decreased from 0.169~ during Phase I to 0.138X after their training in Phase II~ However, the mean BAC of the arrestees of these officers jumped tO 0.1891 in phase III. The mean BAC of the arrestees of the control grodp officers did not change after the test battery was introduced at the end of Phase IIp remaining at 0.161I. Overall, the average BAC of the arrestees of untrained officers was 0.163I (i.e., for 86 ~ases) and the average BAC of the arrestees after training was 0.1601 (i.e.. for 92 cases). The unexpected occurrence of a large number of arrests of stopeee for driving under the influence of drugs makes the average BAC data 60 . . TABLE 23 • ..• • ARRESTS, AVAILABLE BACs, AND MEAN BAC •AS A FUNCTION OF OFFICER GROUPING AND STUDY PHASE CONTROL EXPERIMENTAL Arrest BAC Obtained x BA_____CC Arrest BAC Obtained ~ ~BAC -phase I 51 40 .157% 60 4O .169% Phase II 7 6 .161% 46 42 .138% Phase III 18 18 .161% 36, 32 .189% Untrained Officers .163% (86 BACs obtained) After Training o160% (92 BACs obtained) k I"~,C ~ . . 61 ! oil the arrestees ambiguous in terms of alcoSol alone. In addition, •the occurrence of 32 chemical t~ot refusals probably biases the ~ata." These t~o ~oureee of error on ~he mean BAC of arrested drivers are discussed below. h. ~ ~rrcsts, Twenty four arrestees were suspected of being under ' the influence of drugs -or under the influence of alcohol and drugs. Another six of the stopee$ w,;re suspected of having taken d~ugs, ~' but were not arrested. Four other arrestees were estimated ~: by-police officers to have BACs of 0.20Z or greater, but had actual BACs .of zero, An ~rrestee ~aust . be very .impaired for police officers, no matter how 0k~ll~d,i ~0. i e~at~nate the SaC ,at 0 20Z or greater. • , ,, ..... ~, , ...... The above cases could be excluded from the analysis, but not all of them legitimately should be excluded. Several officers routinely suspected their arrestees of being under• the influence of botb alcohol and drugs and we have no clear indication of how valid their:suspicions were.. Other officers ouspect drugs only after they see 'a low BAC reading. Th~se could ,be,.l.egit~a~ suspicions OF attempts by of~icerB to co~er~hemselveB-~r an-EYr-estee with a low BAC reading° b._~. R.efusals. Thirty two of ~he: arrestees refused C any sort of " chemical test . For example, many arresteeo with prior DWI convictions, especially those driving under suspended licenses, routinely refused all chemical tests. Sixty nine percent of the refusing drivers were over 30 years of age (as compared with only ~gx of the arrestees) suggesting that life ~xperience may play a role in refusing a chemical test. The m~an BAC, as estiL~ated by the officers, for the refusals ~aas 0.198~, as compared with a mean estimated BAC of O.ITIX for all arrestees. Since 72Z of the refusals occurred during Phase I, the ac.tual.~ BAC of all of the arrestees before training may be much higher than the mean BACs given in Table 23 for Phase I. Thus, the refusals could have substantially altered the outcome of the field evaluation. 4_~ ~ police Qffice~ ~ accurately estimate the B_~C levels of stoveep ~ter ~ ~ on the ~ battery? Police officers, trained in administering and scoring the test battery as part of the laboratory evaluation, were able to estimate the BAC of laboratory participants to within 0.03Z (i.e., the mean absolute value difference). As part of the field evaluation, we were concerned with whether or not police officers in the field would be able to do as well as in the laboratory once exposed to the test battery. In addition, ve were interested in what changes might occur in police officer estimates of BACs in the field before and after the test battery was introduced. However, we encountered severs ! problems in gathering these data. a.j.~ F~w stovees are ~eoted. Our sample of laboratory participants probably represent the stopee population quite well, but those who 62 • i • ~ere given sobriety tests in the field represent a subset of this population biased toward blgh BACs. During the entire three month field evaluation, only 322 stopees (10.3~) were given f~eld sobriety tests m~ compared with 441 field sobriety tests, given during the laboratory study. "Since ve estimate that appro~i~a:tely 30Z of the stopees had been drinking, only 37Z of the drinking dr~vers ~ho are 0topped ere given field sobriety tests. ~efore training, 10.2Z of the stopees were tested, and after training,• 10.4g of the stopees were tested. Thus~ while all participants in the laboratory evaluation ~ere given the field sobriety tests, only a small proportion of the stopees are actually given field ~obriety tests. The stopees tested are those who smell strongly of alcohol or who look intoxicated, ~o they are probably biased toward having 4 a high BAC. b. ~ of ~ office~ 0 ~ ~ •~L~_~ ~ The only stopees for whom mn actual BAC was available to compare with an officerOs estimate of the BAC were the Dk~I ~rrestees, oince BAC data on released stopeeu taken during the ridealongs were anonymous. Unfortunately, most officers filled in their data forms at the end of each shift, me they probably often knew the-actual BACs of those arrestees who were given breath teStSo Thus, ~the only valid data obtained in the field study comparing, offlcer estimated BACs ~itb actual BACs probably yore for the 73 ~rreetees who ~ere given blood or urine tests. These 73 arrestees probably represent a very different population than our l~boratory subjects who yore selected .to represent the etopee population. Approximately one third-of :the arrestees given blood or urine tests were suspected of being under the influence of drugs and all of them were considered to be highly impaired by the arresting officer. Moreover, these arrestees represent a ~uch wider range of BACs (0Z to 0.30Z) than our laboratory participants (0% to 0.18%). Thus, we would not ~expect the absolute value of the differences between the estimated'end actual BACs for these subjects to be equivalent to the laborntory eltuation. ~ ~ problems, the _accuracy ~ the ~~s'~ .... B~C ~LL~JtJi.~g~ ~nde0 ~ ]L~t ~ZKe ~ ~ ~$J~L~zg~. Table 24 give~ the absolute mean difference between the actual BAC and the estimated BAC for each officer before and after training: Also given are the number of arrestees represented by each mean. Inr many instances the officer did not have an arrestee who requested a blood or urine test during a particular phase of the study.., There were only six officers for whom ~e have data both before and after ~raining. These six officers improved their estimates by an average of 0.0175~ (s -0.028) after their training. For ~ the~ II officers for ~bom we h~ve some data, the average BAC estimate ~ was, off by 0.077X before training (s - 0.043, n ~ 7) and the, av,ereg e BAC estimate was off by 0.0537Z after training (s ~ 0.031, n - 10). The effect of training was not significant but was in the expected di.rect ion. 63 .: , • ,... TABLE 24 ,% : MEAN ABSOLUTE VALUE DIFFERENCE BETWEEN ESTIMATED BACsAND ACTUAL BACs OF ARRESTEES GIW"K~BLOOD OR URINE TESTS OFF ICER BEFORE TRAINING .AFTER TRAINING CHAN~ # 1 (C) * .15 % (I) ** .11% (i) -.0~ % # 2 (c) .045% (2) • o 5 ~ .(2). +.005% # 3 (c) .085%(4) • io % (z) +.015% # 4 (c) .07 %(3) .02 % (i) -. 05 % # 5 (E) .oz8%(6) .02 % (i) +. 002% # 6 (E) ..ll %(1) .073% ()) - . '037!% #7 (c) .06 %(6) # 8 (E) .015% (2) # 9 (E) .o53~ (v) #I0 (E) .048~ (4) • .[ ~#iI (E) ,0,42% (2) ~=. 0769% i=. 0s37% ~=-. 0175~ s=. 0434% s=.0311% s =. 01279:% i 64 '7 The anonymou~ BAC readiegs of released e~opeee eed the pol~ce obtained BACs of arrested drivers during the 59 ridealongs provides arrest probabilities which could be of some value to police agencies. The term ~tov99, in the remainder of this section, refers to those individuals ~topped by the police ~uring ridealongs for whom we were able to ~btain BAC information. Table 25 gives the probability of a police stopee being ~ith£n the listed BAC ranges. In addition~ the table ~Iso gives the probability of a 'gropes being arrested, both before and after the test battery vas introduced~ as a function of his or her BAC. a. A driver's BAC yers~s hj_~ street pJrobabilitT~ Perhaps the most interesting aspect of the data in Table 25 ig the arrest probability associated with each BAC c~tegory before and after training. Before the test battery was introduced, officers ~ere arresting half of the 8topees in the O.IOZ to 0.149Z range and the majority of the stopees above 0.15Z. No one under 0.10Z usa arrested (unless drugs were suspected). After the test battery uas introduced, all stopees over O.15Z were arrested, h~lf of the stopees betveen O~IOZ and 0.149Z sere arremted, ~nd a few stopees under 0.10Z were arrested. The probability of arrest in the 0,10Z to 0.149X range may not have changed after the teat battery ~ee introduced because many stopees in this BAC range are never ~ ~ ~ sobriety ~est. Thus, an improved test battery cannot alter these decisions. The arrest probabilities in Table 25 are quite rough, since~ they are based upon fev data points. Nevertheless, we believe that the table represents the potential change in arrests once the" test battery is introduced. b. BAC durin2 differen t phases of th___~e study. During the ,three months of ridealongsp 34Z of the stopees had been drinking and about 15Z of them were legally intoxicated. During the early morning shifts (i.e.p betveen II p.m. and 6 a.m.) 61Z of the stopees had been drinking and 26Z were legally intoxicated. Ne only encountered 56 etopees during nine early morning ridealong shifts, so these estimates ere based upon a very limited sample:. During evening shifts (ioe., typically betveen 3 pore. and II p..m.) 29Z of the stopees had been drinking and 13Z were ~,lega!ly intoxicated. Finally, part of the field evaluation occurred ~uring the Christmas Hollday season of 1979-80. We estimate that ~ during the period between December 7, 1979, and February 2, 1980, 41~ of the stopees had been drinking and 19Z were legally intoxicated. A,at0pee does not represent the average driver on the road in terms of BAC. Rational roadside survey data, for example, indicate th'at only about 6Z of the nighttime drivers are legally intoxicated (Lehman, Wolfe, ~nd Kay, 1975). Thus, our stopees were 2.5 times more likely than the average driver to be legally intoxicated. The reason for this discrepancy is that the police stopee had made one or more driving error~. 65 ,r. , k . . ~ ," ~ %, J, ~,, TABLE 2'5 -' DISTRIBUTION ,OF STOPEES ACCORDING ~TO BAC .~D ,AR~,ST PROBABILITY BEFORE AND AFTER 'TRAININGAS ,A F~UNCTION :OF BAC . [. r BAC CATEGORIES PROBABILITY ~OR PROBABILITy OF ARREST A GIVEN STOPEE Before Trainin~ After Trainin~ 7, Zero .,66~ •:000 . O00 • 0! -,049 .1,o6 •,000 . • 000 .05-.099 .080 ,.'000 • 2186 .10 ~ .•!49 .07~' • 500 • 500 .o,s3 • 62~5 1. O'O C) .20 + •.,026 • 800 1.000 66 - t.;' • ', . i¸'' ,' L" %. TABLE 26 MOST COMMON REASONS FOR STOPPING A DRIVER DHRINGTBE FIELD EVALUATION REASON %OF STOPS Speeding .514 Ran stop sign .179 Ran stop light .087 On inappropriate area .060 Equipment violation .051 D Weaving .043 Drifting .034 Not in marked lane .017 Accident .017 No lights .015 Near accident .013 Stops in lane without cause .011 Looks intoxicated .010 Bright lights .009 Driving too slow .008 67 c~.' S~;___C~vcFeus ~ o~ ~ e!cr..or..~ T.a,,bl~e; ~6,~ ~i~v,e~s~ ~h~e.. 1!5~, m'o~s~t~ e:ommo n ~ dr iv in S e freer s made ~yl ~I! I~ ~t ope e s~ d~ur ~n~g~ ~,,e, ~ ~e~l!d~ ev;a:l'u~a,.t:ion and the p~oba~bi~l it.y o>f', occu,~en,ce, d~ur~n,g ~h~e~ ~x~e~l~d~" evslSuation~., No,re than: half of t~he po,l~ice~ stops,: ~ere for speed~i~n~g~,~. ~£~nce most of the part icipat/ng d eputie~s~ hia~d~ ~adar eq~u~iFped~ c a~r~s~., H's~rr£:e et al,. (I~980): r e,s,timate th~at thee, p:r.o,b,a,b:il~,i.ty~ o,~ s,om~e.'on~e~ d~i.lwi~:n,g 10. mpb o~er the s;pe~ed'~ l:imit, having; a.: BAC ower 0;,.I~OZ~ i~a~ about 0oS7% Ba.s~ed, upon our po.l~ce of,£i~c:e~r esCi/mates o~f the BAC~ o~. the stopees, only- 5,.IZ of the, speed,era ~er.e o,we.r 0.1OX,, which i~s,~ pro,b~a~bly less th:an th,e. percen.tsge~ o'f, l~ega4~l'y, iu, to.x.icaCed' d~,rivers~ o~n~ the~"roe,d. On. thee. other h~and,, Earris~ es,~tima~t.e,s; ~h~e pro.ba~b~,i-l~i:ty, of,~ someone~ stopped for, w,e~ving, having a, B~A,~,' o~ 0).,I~0-Z~ o:5 grea~ter to' b~e~ 0~60. During the field e valuation~ 5B.5~i of those s:topped for ~eav. ing: were e,stlmated: to b~e l:ega:l:l.y~ ~nto.~ Co" FEASIBILITY V'irtu~a~1.1y every, po,l~ice o~ffi'cer kn, own to, us~ w~ho: ~s interested~ in~ enforcing. DWI lays, recogn~ize.s the ne~ed, for a research based, et.a,ndar d iz ed Field sobriety t:e s~t b~attery. ~hu s~,, overal ]~ acceptability of an impro.ved test ba~.tte~ry seems highly fav.orab.~e~... A number of critical is:sues, concer.ning ~ t:h~e, fe~a~s:~ib,ility of the test ~ ba,,ttery e-t£11, exist an,d s.hould, be a~d~d~r.es~s~ed' befo~re w~id:espread~ introduction of the, test battery occurs~.., These issues include:. (I) the police attitude toward. DWI a, rres,ts~; {2) police acceptance of s.tand~rdized r~ admi:nl, s:tra~t, ion. and: s:co~n,g~ tech,ni~q~ues; a,n.d~ ('3;);~ pres-et BAC crit'eri,:a: for the test ha:tierS. L :Police Agtitud¢ ~owa~rd ~ Arrests A police officer's attitude toward DWi 8rrests is of extreme i'mportance in determining whet h~er o~ hot a~ s~tan~dard:ized~ fie1~d~ sobriety test battery ~ill he u se;d~. Law~ enforcement o ffi~cer~ generally reflect society's attitud!es t Ow~'~d drunk drivers. Little~ (1968) found that while most people ~ntervi:ewed disapproved o~f" DNI~,~ they~ were not particularly concern~e.d,, a~bo~u,t, any co.n~s,eque,n~ce~s t,o~ themselves. The drunk driver is not particu,larl..y, viBible and: thee consequences of drunk drivi;ng- do n.o~ ~!mpa~ct d:irectly on. mos~ people. Consequently, the public cons~id!er,s police sctlviti~es other than traffic patrol,, ,~uch as pro tect,~'n~g fives and property from crlminals, as being, of "prime £mportanceo Freq~uently, even the •d-runki driver who kil"Ls is not considered, to, be a crim£na.l by the public,, or even by s.ome police off~cers~ but merely,' someone who waS)~ unfortunate. ' PUbl iC attitude is high l,y influential, i-~n~ d!e,~term in in 8 po,lice attitude,8 toward' DI#:I, T,he p0tenti, al ~nfluen'ce on law en,forcement is probably greateot at the munic£pal 1,ewel where police respond / directly to community demmnds. In areas ~ith heavy crime rates and small budgets, the DUI problem is likely to be virtumlXy ignored° Xn districts with lower crime rates, •eucb as those ps~ticlpa,t£u~ in the field Qvalu~tlon~ more emphasis usually i~ placed on ~tr~ffic enfor cement ~ including DWI enforcement° Even then~ however D persons Setting tickets for-hazardous uov~ng violations f~equently complain that the police should be c~tching criminals instead of harassing innocent citisens. Individual police officer~ may also have their own personal reasons for not arresting for DWIo One participating deputye for ezsmpl'e•. • insisted that his primary life interest .was in making his marriage work so that he ' avoided anything that ~ight force him tO ~ork 0re,time, including DNI arrests° Other reasons police avoid such arrests include: they drink and drive themselves; they don't fully understand the consequences of alcohol, impairment ; the arrest process requires too much overtime for which they do not get extra pay; they receive poor support in the courts; DWI enforcement is not encouraged by their immediate supervisor; they prefer other kinds of enforcement activities; and/or ~any :~ •.other reasOnSo Factors influencing DWI " ~rrests have been ,.studied previously in other ~HTSA contracts (IqETSAB 1972; Youn~ and~ .Co., 1974; Oates, 1974; Hawkins et al~ 1976). ,~ A standardized field sobriety test battery is not a cure for ~oor police attitudes. Officers who avoid Dl~I arrests will probably continue to avoid them for the same reasons° Officers ~ho use the test battery and find that it makes their job easier and helps them set convictions may make more arzest~ once they are ~iven ~he ~ .test battery ~s a tools '~ -. t~ A number of factors also could cause the introduction of the .tes.t battery to have a negative effect on police attitudes, including: (i) Officers may find they are ~rrestlng more drivers under,~ 0.IO~ ~equirlng them to fill out nn arrest report even though the driver is released at the station. (2) Officers may find that mor,e arrests in the 0.I0~ to 0.15~ range are being plea-bargained since they are more plentiful. Plea-bargaining discourages ~. paoli.ca officers from making similar arrests. (3) Note DWI ~rrests :may c~u~e a back up Of cases in the courts and result in considerable plea-bargaining regardless of the BAC. , . 2.~ Polic~ Accevtance of ~and~rd~e¢~ bdmin~s~zatio~ an___dd~ Sc~.~ ~ocedures .... ~ ~ost officers concerned with DI#Z enforcement see the ueed for ?~.a •,tand~dized te~t battery, in the ~en~e that every officer~would administer the same tests in the same ~ay. Hovever~ officers, are reluctant to use an elaborate scoring system or even any-~scoring system° This resistance appears to be the result of a re luct:ance to a~e ~nythin~ very complicated and the probable lack ~ of underetandin~ of ~he benefits and purpose of standardized scoring° The training of officers during the field evaluation was very extensive° SCRI staff members were convinced that every• office~r 69 " ~ " completing the training could correctly sdmin~-st:-er and ~ m:c'ore~ t h:el test battery. Unfortunately. s~ome offi~cera, forgo~: @r £gin~or, ed',, mo~s~t~, of the administration procedures~, e~cept those associated' w,ft~b nystmgmus, by the time their second post-traluing r i de a~l~on,g~ occurred• These officers appeared to believe tha,t they were stl;l~l~ ~,~dministering "the onl-le8 stand rest w~ or the "walk and tur, n~ t:es~t, "~ ,~nd i~ thaW. differences in the |Srdn.inJ;strat~on procedure were ~. ~nimpor t an~'t o .SCRI observers, when present duri.ug r/.dealongs, requested that e..~l~ sobriety teats be scored immediately, N'evertheless, we sus:~pe~c:t~ that many officers filled Out their scoring sheets at the end o£ ~ -their ~hift or at the time they completed~ the, arrest report for that..individusl• Most police officers have, remarkable memor£es~ for" deta:iil~ but we still suspect that man'y advantages of standia~r:dii, zedi scoring are lost when the scoring is l~ft to memory• - ~. j Failur, e to have sobriety testa whlch s're cons'~Btently administered~ and scored probably results in, the a~cqu:irtal of numerous DWI :defendants• Pressure from the courts and from police superlors for :consistency i8 one possible way for sta.nda, rdized procedures to be ado, pred.. In order for this to hap:pe,n, we believe t,har the standardized admini~trmtion and sc6~ ink procedur es should be incorporated inca the police arrest forms• / % The ~obriety test battery was introduce~d in,to the field evaluatio'n~ Us~,.ng .arrest criteria that ~ere se.C tO a BAC of O•lOZ during the~- .labOratory~studies. Several problems a~rose vith~ ~hese criteria Fir.st.p ,leboratory procedures ere as exact as'passibis, while, a~rrest, proce~dures tend to err in, f~vor of the a, rregtee• For exsmFIe,. ~n~ -"she ..laboratory a BAC reading of 0.099Z: is round~ed to O•10Z ex-ce:p,~ .in figuring decision matrices ~here 0•099X is treated as being less than 0.10Zo For ~ DMI errestee this reading 'woul,d be considered, 0•09~ at all times• Second, the field sobriety test is des~.gned! .to help the po~lice ~ officer estimate whether the stopee i~ legally intoxicated at th,e, time of the testing. Unfortunately. anactua.t BAC rending may? no'~ be obtained for over an hour ~fter the decks,ion t.o arrest is,~ m,sd!e~., Thus, 8 stopee with a BAC correct, ly est~mated~ at 0•12Z may h"av.~e a~ reading of o.098z..(i~,e.o!~ w~i~,~h ~:r00:i~d~e!,d .to 0•09Z) when an actua:,l~, chemical test finally is,:ob~ai,~ned~ In ~bSt ca,8.es, this individu~ai ~ould be released immediately and no cha~rges wouId be filed Occasionally, an officer in California ~nay still follow through Vith ~' an mrrest if the ,chemi'cal test is ~n.~.~.he. 0.08~ to O,09Z range. One officer informed um~i;of such a~cas~ ~i~g th~ field .tudy. The~ prosecutor handling t'~" came, 'vith~i~~-!ic0n~nlting the arresting: officer, merely asked ~he defendant if: he .~bul~d accept two moving violatlons. The defeudan.~ akg,hed for j~ib~ ~ ~ mpeedi, ng ticket a, nd it yam granted. ,~ ~ . ~ -. "~ .• .2 rl | SC•RI has adjustable arrest criteria associated with " the, Lest battery. Local law onforcement officials might select their"own ~rrest criteria, based Upon what their courts will accept. Otherwise, many low IAC dr~yers may be arrested resulting •in, more plea-bargaining and negative police attitudes toward using' the standardized test battery. ~, .~ I ,, . . ° 71 '. v. uLp- r ..:. ONCLUS ...... :The major objectives of this project have been to' (I) complete the laboratory development and validatiou of the sobriety test battery, whic~h was initially identified under Contract No. DOT-HS-5-01242, and to ('2) assess in the field its feasibility and effectiveness when used by the police for estimating BAC and facilitating the .identification of those drivers with BACs greater than or equal ~,. 0.10Z. ": ~ ~. ' ~.:~ Idministratlon, scoring, and interpretation procedures and criteria for: the three-test battery have been refined and evaluated. Under Laboratory co~.itions, and in the hands of adequately trained ,personnel, the* ~,~st battery is a sensitive index of BAC and of impairment. Based on exhaustive analysis of the laboratory evaluation data, we conclude that the tests are optimally developed and etandardigeds and no further laboratory work io recommended The laboratory data indicate that police officers established an average test performance criterion auc~ that they made "arrest" decisions ~t a mean BAC of O.08Z and higher'. Their ~ of I BACs differed from actual BACs, as measured by Intoximeter, by 0,,03Z (s = 0.005Z). They alno were able to correctly classify 8lZ ~ the laboratory subjects in term~ of being above or below 0.10Z BAC. Reliability measures produced correlations in the range of 0.60 to 0.80 for test-retest reliability and also for interrster z~liability. This project has confirmed the findings of DOT-~IS-5-01242 that gaze nTstagmus is an outstandingly useful tool for the officer at roadside. An additional i~,portant finding is that 'angle of onset' 0 £ " ,the characteristic jerking motion of the eyes, as a sole me.ssur e, enabled officers to -correctly classify 78Z of the laboratory subjects. For this measure to be maximally useful~ officers shouldbe trained to es*timate the •angle of onset with considerable precioion. With precise measurement of the angle of onset 88Z of the laboratory participants could have been correct]y classified. The second project objective0 evalugtion of the test battery in the fields also has been met with a limited sample. Additional field evaluation is recommended. The limited field evaluation was carried out as a three-phase study. Officers were assigned to an:experimental or control group, end over three time periods filled out data forms on all stopees. The variable of interest for the different time periods was nuntrained" on the three-test bat~tery versus "trained" to administer and score the tests. SCRI staff members also collected data by riding with participating officers to observe test administration and scoring and to obtain anonymous breath aanples .for BAC analysis from stopees who were released 72 Yhe questions addressed by the analysis of the field data" ~were: {1) ~id the nunbar of arrests increase miter pollea office~8 were ~rainod to ~se the ~e~t battery? (2) W~re the ,file@re b~tter/~ble to dlacrlminate 0.1@Z B£Co ~s a result of umim~ ~he te~tbstteryT {$) Did the mean ~£C of arrested drivers • decline after introduction of the test b~ttery? (~) ~ere the officers better able to'detect lmpair~ent as s r~sult of using the test b~ttery? Definitive easters to the questions cannot be offered, based on the'limited nature of this field study, but the data do clearly ~suggest positive results due to use of the battery. A 20~ increa~e ~n arrest rate~ occurred. Officers were able to make' more accurate geci~ions relative to BACs of 0o10~ and it appears that•they were better able to estimate BACs. B__~. RICOMI~UDATIO~$ /. ~ajor effort is needed for a subsequent field evaluation, repeating essentially the same study design with s sample which is both l~rger and broader. Are~s which caused difficulty in obtaining ~sta and which are therefore crlticsl issues in de:sign :of additional study, include the follo¢Ing: l._~. Foliee Attitud~ ~nd~p_L~j~ / -" ~. ~. Zxtremely sexious problems result when there is a lack of int:erest and cooperation by individual officers, by supervisory personnel, or by agencies. Good data, and ultimately effective utiliza.tion of the test battery on a large scales requires motivation at these various level, to cooperate witb the research and to give I~ high priority to the arrest of alcohol-impaired drivers. • . .... t The greatest impact of the tests will be realized .if~, law enforcement agencies and officers, recognizing the sensitivity of nystagmus as an index of BAC~ routinely check the eyes of at.peas. As the data from the project have demonstrated, many alcohol-impaired drivers are being releaBed without any testing~ at roadside. A routine examination of all ,topees for nystagmus would more effectively detect the drinking driver than the current observational uethods which rely on odor~ slurred speech, or other ? .. obvious signs of intoxication. , ~, Experience in the Los Angeles urban area. where traffic density is O rel~tively heavy~ indicates that eight traffic stops per shift is the maximum avermge nuuber which can be expected. A project schedule should be baaed on this estimate. , . The disposition of arrested DWI cases by the courts is xmpor£ant data which hm8 not been ~ealt with in this or earlier studies. Officers~ at the present, often express frustration over what. •they perceive as lack of support by. the courts and the futility of arrestimg .DWIes who will plea-bargain a lesser charge~ and experience only minimum penalty. The situation may be elthe~ 73 worsened or improved by stny more arrests sad ~=~®et~ ~t lo~er SACep depending on action taken by the col~t~; ~'Clearly; intQrmctions with the courts is an important component of effective DWX "deterrence, and thus" should be included In the field evaiuation. The project schedule should be long enough to pernit deveiopnent of contacts with the judiciary and the final period.dieposit ion~ of DtYI charges which arise dur ins the evnlustion ~_L ~ Considerations Many I~w enforcement agencies continue to operate units with t~o officers, Particularly on nighttime shifts. For example, both the California lligbwey Patrol end the Los Angeles Police Department have ~two officers in traffic patrol unite. If such agencies ~re involved in the field evaluation (and to routinely exclude ~ll of .those ,with t~o-officer unite uould introduce unacceptable biases ..into the data), then the number of officers would double, and Clearly there will be a substantial increase in the coBts of training and supervision. Obta,lni,ng law enforcement cooperation is a major effort, in and of itself; requiring considerable ti~e. The various .agencies vhich b~ve ~orked cooperatively with SCRI during the execution of two DtTI p'~ojeCtS have had serious concerns about legal issues involved in t "~° field evaluation, including the following: (1) If permission is .Siren to obtain breath samples, the agencies require guarantees that the samples be anonymous. Their legitimate concern i8 that if a ~river whose BAC exceeds O.IOZ i~s released and subsequently is tr~olved in an accident, the BAC reading may be ~ubpoenaed as zv~denCe ~nd the police agency could be held liable for having rele,a,s:ed an impaired 'driver.. (2) Stopeea may feel embarrassed and ;h~ara~ase'd by being asked for a breath sample, tKerncies typically ere:acutely aware of public relations"problens and thu.s object to in~roduci.n;g r'esearc,h procedures ~hich the public will not like. (3), Ill, the field ,study .reveals that officers actually are releasing *a l'a'rK'e proportion Of high BAC drivers, then t~his information may b~come Wi~dely ,known and may be used ,as c:ritic:ism against ~he a gen~cy. ~h,e.se ~esues are 'neit:her tr,ivial .nor easily re.solved. I.f rhe agen~cy's policy m~kers rule ~t~hat parti ciption in the research is 'nbt a!p,pr~ove.d:, then ,li~t ale re,course "'ram,sins.. T.he au~hor i~y of aSe;n~cy ,direc~tor:s is ab~olu~te,, ~n.d local u.nits o.f ,state pol.ice, for e x~am p'l ~e., ~ii 1 ~1 n~o,t 'c OOp e,r~a.t .e vl ,t!h,o.u.t fu :1 1 ,ap ~pr,o v 'a 1 o f ~.h e 'a p pr ~op :r .i~a t ,e :s.u!pe r v l ~s:o.r's and ~ dm i n i.s :tr a t,or e. The ~ xdea,l,ong ~sy~et~em i~ 8n ~por~t,ant c ouponent of the field study pFen. SCRI reco'm,~ende •that ,ee:f fi,cient • ipersonne 1 b,e a~ ~igned t,o t:he pro,j ec t tO pe:r.m.~ ~ one ob8 • rv,e r 'f~or ,e.m~c~h ~s i~x t.r a f f i.c ,p:a t r o.1 ,u n i t .s. -.. . I n ~s un',ua ry ,, S'CRi reC,omle:n,d~® t~h,~t ~tih:e ~'f i'e"~rd eV,~ l~u,a t ~i,o n ,o:f the th~ee~'t'e~s:~ :h~a,tte:ry •ib,e ~c'ompl,e'.t~ed w:i,ti~ a m~olr ~'e"ffo,r~t o .A period ,of 1B 'mO,n:t~ s ~i S .re~c,o~ue~nd,e;d ,i+n ,or~d,e~,r t,o ~ca'r,ry ,out ;t~h e ot u.dy on ~s n~e'ti~on-e:i~d~e bas~8 vi~:tih ~d:ive~r~.e =l'aw en~o~rc,em~e;n'.t ~a~g.e~n..c.i~ea. ;t , , R~FK~ENCE8 A~chen, Go D~ff@r~n~ typ~s of alcohol nystagmus.' • ' ~...r~..~,..~ Otolervn~ol, ~LI~,IULI~, ~s~, ,.L,A~_, s9-78. ,~ ., ,•• £schen, C. and Bergs~edt, N. Positional alcohol nyot~gmus (P~) in man following repeated alcohol dose~. Acts Otolary~ol, Aschan, C., Bergstedt, M., ~nd Coldberg~ L. The effect of ~ome antihist~mini c drugs on position~l alcohol nystagmus. Ar~ O~olarvnz@l ~_~, 1958, 140, 79, q A~chan, G., Berg~tedt~ M., Coldberg, L., ~nd Laureii: " L . Positional ny~agmus in m~n daring ~nd after alcohol intoxication. q u~rterly Journal o__~ Studies on Alcohol~ 1956~ 17~ 381-~05o Baloh, R., ~onrad, B., and Honruhia, V ' Vestibulo-ocular function in p~tient~ ~ith cerebellar s~rophy. ~~, 1975, 2~_, 160-168. ~yB~rdy,.A.~ zn Saults~'~l@ma E.~ Huhmar, E " end Lehtovo~ra, ~ . Pos~ural • ~orm~l v~lue ~ffect o~ ~lcohol xngestion. ~Lr.O_I.@ML~ Scand£n~viG~ ~~ 1978, ~7, 278. Barnes, E., Cooke, N., K£ng, A., and P~s~more~ Ro Observations On the metabolism of alcohol in mano ~ ~rnal of Nutrit~@n, 1965~ 19, ~85-489. Begb~e, C, The effects of alcohol end of varying amount o~ visual in£ormatlon on a balancing test. ~, 1966, ~, 325-~33. Beitel, G., Sharp, M., and Glarez, W. Probability of arrest • whi.l.e driving under ~he influence of elcoholo Journa~ .~_~ Studied,on Bender, N., and ,'Brian, F. The influence of barbiturates on various forms of nystagmus. American ~ ~ O_pthalmologv, 19~6 , 29° 1541-1552o Burns, N., ~nd Mo~kowitz~ H. Psychovhv~cal J~.~_ ~or ~_! a~restt (DOT-H$-S-01242). Washington, : D. Co: U.S. Department: of Transportation~ NHTSA0 19770 Q Cahalanp Do, Cisinp Io, and Crossleyp H. American drinking New Brunswick: Rutgers Center of Alcohol Studiesp 1969. • Coll£mo, W. Man£pulst£on of arousal and its effects on human vestibular ny~tagmu8 induced by caloric irregation and~angular accelerations. ~d~, 1963, 124-129. Collins, W.B 8¢hroeder, Do, and HLII~ R. Some effects of .... alcohol. on vestibular responses. Advances £n 9to-Rh£no-L~ryn~olo~v~ 1973, 75 ~19, . 29 5-303. Cronbacb, L. ~ of ~cholo~ica] ~rper ~nd ~ow, 1970. " ~" Franke~ Roo Hen~ley, U., Renaley, W., 8farmer, C., and Teoo R. The --r.elstlou~hip between alcohol dosage and performance decrem,ent"£u Fregly, A., Bergstedt, N., and Graybiel, A. Relationships between blood alcohol, positional alcohol nystagmus sad post~ural -equi.!ibriuu. _OJerter~y ~ of ~ nn ~, 1967, Fregly~ A., Cr~ybiel, A. and Smiths N. Walk on floor eyes c10eed (~OFEC): A new addition to an etszia test battery. ~, 1972, 395-399 ...... Goldberg, L. Effects and after-effecta of alcohol, tranquilizers, '~nd~ fatigue on ocular phenomena. In J.D.J. Harvard (Ed.), ~_~ an___dd road traffic. London: Br~tish Nedical Association, 1963. Goldberg, Lo Behavioral and physiological effects of alcohol on ~an. ;Ps~vch.oS~nat~e ~, 1966, ~.~., 70-595. ~uilford, J. ~nd Fruchter, B° F_nl~d~me n t a I ~L~ 'P-~R.CJt~]~L~ ~ ~ucetion~ ~ew York: McGrsw-Hill~ 1978. '~ • " '.i~ :: . Iiarri~, D., Dick, R.. CaGey, S., ~nd Jerosz, C. ~ _Pd• ~etecCi~p uethods, (Contract ~o. DOT HS-7-1538). ~ash~ngton, D:. C°: Departuent of Transportation, ~HTSA, 1980. Hawkins, T., Scrimglour, C., ~renck, R., and Dreyer, C. ~ ~tnd~n~s, (Contract l~o. DOT ~8-5-01155). ~a0hington, D. C.~: Department of Transportation, NHTSA, 1976. Rill, R., Collins, ~., and 8chroeder, D. Influence of alcohol on positional ny~taamu s over 32 hour periods. ~L~JLJ~ o£ O~olo~y. and L_~cvn~olo~v. 1973, 82, 103...... Houard~ I. and Te~pleton, W. ~ ~ orientation. ~ev York: ~iley, 1966. ~ ...... Lehman, R., Wolfe, Ao, and Kay, R. ~ ~ ~ of ~O~i~_~ ~ ~ ~urye?s~ ~970-1974. Ann Arbor: University of Nichigan, Highway Safety Research Institute, 1975. Lebti, H. The effect of blood alcohol concentration of the onset of. gaze ny6tagmu~°. Blutslkohol, 1976, 13, 411-414. ~ Liden., C., Lovejoy, F., and Costello, C. Mine cases of poisoning -- phenyclidine. ~ of ~he AmerJ~Ml ~ /klLF_~_~~.D_~ 1975, 2~t, 51:3-516. 76 "+. ~j • ~ +. +• -+ Little, J. A theory and empirical study of what deters drink+ing drivers, if, ~hen. ~nd why. #~dm~i~r~tion ~ gravies. 1968, ,Z~, 23-57 ; 167-193 . Maebe, Y. Diagnosis of ~lcohol i~toxication by She optokinetlc ~e~t. ~~.gg~of+stue~ee o~ Alc~ohol, 1969, 30, 1-16. ~o~ey~ ~o~ ~nd ~yleeo U. Heavy ~ater nystagmus and effects ~ of ~lcohol. ~gggg~_, 1976, ~__~ 604-~05o Money~ E. and ~yles, W. ~otion mick.ness ~nd other we~cibulo-ga~tric illness. In R. ~auton (Ed.) : The veoti]Tul~r q • ew York: Academic Press, 1975. ~osko~ita~ ~o Laboratory ~tudies of the effects of alcohol on some v~zlables rel~ted to driving. Jouznal p_~ Research, 1973~ ~. 185-199. f . Murphree, H.~ Price, L.m ~nd Greenburg, L. Effect of congeners in alcoholic beverages on the incidence of nystagmus. Ous~rterly Jourpal ~LL $~ud~e6 on Alcohol~ 1966, 26, 201-213o National R ighwmy Traffic ~nd Safety AdministratiOno AlcQhgl is~fetv pro4ect~: ~valu~tion of ~ezstionp -- 1972. V~.~_.~I~: I~_LL ~e~ailed analv~e. (D0~-~S-800-874) ~ashin~ton, Do C., 1972o ~a~ioual Eighw~y T~affic ~nd Safety Administration. Alcohol ~DOT'HS-801-707). Wa~hington~ D. C., 1976. Ni~iokik~jien, C. The influence of an auditory task on ~omber~s te~t in healthy people and neurological patients. P~ocedureG g~Lfl ~ ~ ~n~nation~ ~e Fo~.roer~nh~f Smoleni~e. Septembe~ 0 1973. Pp. 11-19. ' i ' ' Oaten. ~. ~ inflcenci~ & arrests ~o~ slcobo~ ~elated traffic Y_~ (Contract MOo DOT-HS-801-230), ~sshing~on, ~ D. 'C.: Department of Transportation, NHTSA, 1974 .... , ,~:~ Oostervel~ W. Effect of gravity on positional alcohol nystagmus (PAN)° C~in~cal Aviation en_~d Aerosn~ce Medecine, 1970, ~_~_I~ Q '557-560. ~ ~ : Penttil~, A.~ Tenhu~ M.~ and gataja, M. Clinical cgamination ~ gggg-g JL~ suspected drnnken driving. Statistical ~ud Re~earcb ~ureau of TALJA. I~o Roobertinkatu 20, Helsinki 13, ~£nland, 1971. ~,.+. Peuttila. A.o Tenhu, M., and gataja, M. Examination o_~ alcohol ~Btoz~.G~ion ~ £gggg of ~usDect~d ~LLp_ILg_gg driyers _11. Liikenne~urvaD Zso loobertinkatu 20, 00120 Relaink£ 12, ~nland, 1974. 77 ++ . ". , + ..~ . , .+ . / !.. ;'R~eh~.b~'~s~ ~ Co B~rbiturmCe myet.~gSue and the meeh~en.Ka.a~8 of " ~tsu~ei t io'n. , ...... i~a~shb.as,~, C. The r'ela~£ons~hip b,et~e-en ,Sacc a.~d:ic mn~d smooth trec,k~'i~ng e~y"e,. • uovementao,~ ~ o~ ~ 1961, ,~. 32"6-358. :~R~binson, ,,D° Eye ~ovemen~ control in pr'ima'tee- ~,. ]'96:8, ~.~ 1219 ~. 12240 '~ yb:a~c k.,, , ~. mnd Do,d, Po A~ t er e,f ~'e c t~ of various alcoholic ~b:ev er,~ge ~ on positional nys t'agmu e 'an~ cor~olis acceleration,. ~A,'A-A~-7 I-6 ). 0~l~homm C:~'~y : ,Fe~ e r,al AV ~a t,io,n ~dm in i's't r'mtio n~, ~!!'9 ;7,l,a.. p o n Se s;~ I.l :a C ce le ~ ~ t io n .(FA~-A~-71-16); Oklaho~ C~ty: ~edera~ ~•via~ion Admi i 971b o iniStr~tion, $i~peon-Cra~ford~ T., ~nd ~late,r~ ~S. Eye Siena in dr~inking driver~: Clinicel examination :in. relaZton uuepected to blood a~lC°hol" ~ ~ ~ 3ou~n~S~,, 197i, ~7~, 92-96. SUmmere~ L. en~ Harris, D. ~ ~ ~ -- ~olume~ ~DO~HS-803-582). ~m~hington. ,D. ~ ~ (Contract No. Tra, n~port~tlon~ .~BTSA, 1978. C.: Depmrtm~nt of T°'g i~•, J- Electronyata~mo~r~ewbv~: ~ aavect~ ~ ~~ ~prlnigfield,. Illo: C.C. Thomas. 1'976. Une,da~ Yo and S~akates K. ~lcohol and t~he oculomotor system. .'~, O~ ~ R_~p_R~p_g.~~ an___dd ~arynxolo~v,~ 1978, 8..~7,. 392-398. W~lkinson, I., Eime, Ro and Purneil, M. ~lc'oh~o,1 and human eye ~ovement. ~, 197~, ~7, 785-792. W•i,18on, ~o, Barboriakp J., ~and K,a,s~s, U. E££ectm of alc0holi,c beverages end cogeners on peyc~h'omotor skills in old and youn~g eu'bj~cts. ~ .~ ~ ~udie~]~ ;on ~ 1970 115-129. Young, ~Ao C. and Co. ~ b~i~hwev ~ ~ Q~fense 'ediudicetion. (Contract No. DO"T-HS-801-21'6). Uaahington, D.C.: Dep~tment of Transportation, MHTSA, |'97~. : ! 78 APPENDIX A A. Alcohol And ~ys~a~us ~yetagmu8 refers to a jerking of the eyes which may be pendular (equal on both sides) or asymmetric with a slow end fast phase (To~lla~ 1976)o Alcohol appears to influence a number of different kiads of nystaSmus~ including: positional nystagmus (Aschan~ 1958; Goldberg9 1963), post-rotational nystagmus (Schroder, ~1971b), c~loric nystagmus (Schroeder, 1971a), optokinetic nystagmus (5chroedsr, 1971a)~ gaze uystagmus (Aachen, 1958; Lehti, 1976),. If all of these forms of nyetag~us ere congidered~ then• the literature on ~lcohol ~nd nystagmus i~ quite large and somewhat ©outredictory. ~o~ever~ by studying the mechanisms producing nystagmue~ the literature can easily be sorted. ~eentially, alcohol can influence nyst~gmus in t~o ways: (1) mechanically by acting on the vestibular 8ystem~ and ,~ (2) ~eurologically. J-L Ve0t~bular ~ (See Ho~ard and Templetonp 1966) In ~an, three semicircular canals, joined at right angl~ are located in each inmer ear. The canals are filled ~it~ fluid, called endolymph. A ~elling or ampulla is located in- e~ch ~anal and contains the ~eneory transducer of the canal. Zssentiallly,. t.he cilia of a number of ~ensory cells project into e common gel~tfnousl ~ss~ ~he cupula. This ©~pula is hinged at one end, so that~ it ~can e~ng from side to side ~ith the ampulla. Xn the upright position, ~he c upula for~e an effective seal, preventing the leakage of e~dolymph pes~ tha~ point. The semicircular canals respond to anguIar ~cceleration, such ~as i:n head ~ovemen~ which causes the endolymph to lag behind the head ~ ~ovement (ion., the fluid move~) ~nd deflects the cupula. Deflection of the cupula d~chargee the ~enso~y cells and provides• the ~ensetion of movement. ~ith constant angular accelerati0~n, the ~ystem provid~ accurate ~nforuatlon for the first ten seconds or ~o and then undere~ti~ete~ the a~ount of acceleratiOns Xf the person ie then held at a constan~ velocity, then the cupule catches up to the skull novement (i.e., it returns to normal position) and ~he sensation is one of elo~Ing do~n and eventually (in ~bout ~0 ~econds) of stopping. If the person is stopped, then he Or she niXl sense a sudden acceleration ~n the opposite direction because ~he he~d is no~ slo~er th~n the endoly~ph, which causes the cupuIa~: ~o deflect in the opposite direction. Xf the person :remains stopped~ then the eupula returns to its level position giving~, a~ senaation of slowin~ down and stopping. .. .." ' 8inca ~h~ three ~emicircuXar cenal~ in e~ch ear are at right angle~ ~e can sense angular acceleration in ~ny direction. ~hen visual information conflicts with the ~en~ation of moti0n,...... ene feels ~ssy sod may feel sick. However, ~he mere sensation~of 79 ~. i movement may produce illness in some iudividualSo T'h~e: Vestibular system interacts w.,ith the viaual system hy p, roducin:g altern~ating fast and o~low "eye movements (i.e. , nyetagmue) iln addition to the sensation of movement.~ Nystagmus is produced be'cayuse the eyes lag. behind the angular acceleration, so s "brain center" makes periodic adjustments in order to maintain adeq~Jate fore:el fixation. For example, ,,he can move one's head back end forth and still maintain fixation. Unfortunately, angular acceleration is not the only at£mulus which will cause cupular deflectiono The cupula and endolymph both have the same specific gravity° A very slight change in the specific gravity of either the fluid or the cupula say result in a cupular deflection, because the system becomes sensitive to gravity with certain head positions. Money and Miles (1975) claim that a change in the specific gravity of 3 parts in 100,000 will make the system aenaitive to gravity. Alcohol:. and some other drugs can alter the balance in specific gravity (Money and Miles, 197/+; 1975)o The base of the cup ula has a rich blood supply. Foreign substances in the blood will diffuse rapidly into the cupula because of its proximity to the blood and alter the specific gravity of the cupula with respect to the endolympb. The direction of the nystagmua (ioeo, the fast phase) will depend upon whether the drug makes the specific gravity of the cupula greater or leas than that .of the endolymph. For .~example, within one hour after consuming alcohol a positional alcohol nystagmus (PAN) will occur. That is, if from supine /position one rolls one°s head to the side (ioe.p so that the cupula im 'subject to gravity)~ a nystagmusp called PAN I~ occurs in which the fast eye movements are down (e°g., Aschan and Bergsted0 1975), Approximately four hours after drinking, the nystagmua stops. This is~. probably because sufficient alcohol has defused into the endolymph so that its specific gravity equals that of the cupula. Finally, as alcohol is elilinated from the blood stream, the endolympb ends up with a greater concentration of alcohol than the cupula. At this point, a positional nystagmua occurs in which the ' fast eye movements are up (PAN IX). PAN lI may persist up to 20 hours after consuming alcohol -- long after alcohol has been eliminated from the bloodstream (Hill, Collins, and Schroeder, 1973). In fact, under conditions of increased gravity, PAN Ii has been found up to 40 hours after drinking alcohol (Oosterveld, 1970). The change in specific gravity also explains why the presence of congeners in alcohol can increase the amount of positional nystagmus (Murphree, Prlce~ Greenberg, 1966; Ryback and ' Dowd, 1970). Excellent reviews of the PAN phenomenon are contained ~ in A achan, Bergatedt ~ Coldbergp and Lsurell (1956) ; Fregly, "":' Bergo t edt , and Graybiel (1967); Hill, Collins, and Schroeder (1973) ; Aschan and Bergstedt (1975); Aschan (1958) ; and Coldberg (1963) , PAN I intensity provides a rather good indication of the peak BAC (Coldberg, 1963)p but not of the duration of the intoxication. PAN 80 .:-.. II intensity has been correlated ~ith hangover effects (Co!dber'g . 1963)~ Alcohol affects nystsgmus in an indirect way -- by inhibiting ~ the neural mechanisms involved in maintaining visual fix~ttion. In some instancesp visual fixation acts to inhibit nyatagmus. ThUs, if a vestibular signal tells one that rotation is occurring ~hile'visUal information eonfiicts~ then the visual information usually wins but often at the expense of producing nausea. , Irrigating the ears ~ith warm or cold water starts the endolymph fluid moving and produces a nystagmus called caloric nV~t¢~mus (e.go, Schroeder~ 1971a). Visual fixation will inhibit this nystagmusp but not after taking alcohol (Schroeder ]971a). Similarlyo rotational nystagmus or post-rotational nystagmus can also be suppressed by visual fixation. But fixation again is ineffective after taking alcohol (Schroeder~ 1971b). Both rotational and caloric ny~tagmusD however, are also reduced by Io~ levels of arousal, suggesting the alcohol, suppression may also be due to the sedative effect of the drug (Collins, 1963; 1973). In all of the ~bove examples, ny~tagmus is produced by vestibular activation and alcohol acts to suppress that nystagmus. However,. alcohol reducea nystagmus that is not produced by vest~ibular activation. Optokinetic nystagmusp for example, is produced by, watching a rolating drum covered with alternating black and white vertical strips (Hisoi, Hishida~ and Naeba, 1969). It consists.of slow component in the direction of the moving object (or strips) and a quick phase in the opposite direction. Mizoi~ Rishida, and Maeba (1969) describe four phases of optokineti~ nystagmus: Firs.tl. the slow eye movemento keep up with the movement of the~.0bjec.t,. Second, the slow phase eye movements accelerate, but cannot keep UP with the stimulus. Third, the slow phase attains its maximum speed. An average person can typically follow a moving object up to 30 degrees per second. Finally, the eye movement•.fa£!So ' Alcohol impairs optokinetic nystagmua by reducing the maximum speed that can be obtained (Mizoi et al., 1969) o ., The slow eye movements mentioned in connection with opto.kinetic nyatagmus are called "smooth pursuit" movements (Rashbass, 1961; Robimsono 1968)o, This system for moving the eyes (I) requires .a moving stimulus; (2) £s virtually autonomic; and (3) is concerued primarily with matching the speed of the eye with the speed.of the target (Robinson, 1968). These movements appear to funct:ion in providing a stable image on the retina (Rashbass, 1961). Smooth movements do nothing to correct for the position of the.ta.rget , ~hicb lathe function of the much faster "saccadic" eye movement" system (Rashbass~ 1961; Robinson, 1968). .~ ..:.: The smooth pursuit system appears to be particularly vulnerable to the effects of alcohol (Wilkinson, Kime, and Purnell, 1974),: This system normally can tr~ck movement ~t up to 30 degrees per second,. Alcohol~ however, reduces the maximal tracking speed and, i,n ~UffiCient concern,rat!on, may eliminate Smooth p4rau~t ~ovement e enti~e!y. When the gAC iQ high enough, only the secc~dic syste~ '(which adjusts the eye for tarKet position when the position ,difference is above some threshold) reualns. Thus, ~t a Qufficiently high BAC, one can on!y follow a Loving object ~series of saccmdic jerks, with a Rashbass (1959) claims that th e inability to !aintal ~ v~suel f.ixation is responsible for gaze nystagmus, a jerking movement of the eyes when they are deviated laterally. He argues that only th e smooth pursuit system is involved in bringing the eye to a single spot. When the eyes are deviated~ tO the side, slow drifting novements will occur toward the center depending upon the amount of is,era! deviation and the ability of the smooth Pursuit system to counteract these drifts. When the smo?th purmg~t ~ystem is inhibited by drug~ ~uch as alcohol ~r barbiturates, the slgw dr~fts become large enough that saccadic jerks are required to maintain ,the.lateral gaze. ~..... , ..... ,..- ...... Ca:ze nystsgmus can be seen in 50-60% of o!l individuals if their ,eyes ~ are deviated to the extremes, but it is considered to be path°~°gic~! when it occurs at ~ess extreme (i.e., 40 degrees) d~'~ietions (Toglia. 1976). Gsze nyatagmus occurs with some types o~ ~rain damage (Baloh, Konrad, 9nd Honruba, 1975), bu t ~t pr0vides, )~tie !0calizing value in detecting the brain da~ag e ~xcept to d~rect, one's attention away from t h~ periphera~ ~labyrinths of the vestibular ~ystem. The data of Baloh et al (1975) does support ~hbass' theory ~n that pathologice ~ gaze nystagmus corre!~tes ~h ,'f~xation instability. F~ve of their six Patients with f~xmtion inkt~bility also showed Pathological gaze nystagmus. , ~Gaze ~ystagmus occurs under several different drugs, including alc,oholl (i.e,, Aschan, 1958), barbiturates (e,g., Bender, O|Brien, ' ': 1946)" antihistamines (Aschan, Bergstedt, and Coldberg, 1958) and phencyclidine (Linden, Lovejoy, and Costello, 1975) Anumber of ...... o~her drugs may also produce gaze nyetagmus, but most of the ~!i @~£dence, ...... is contained...... in clinical case reports- ...... Although some articles mention the occurrence o£ alcohol gaze nystagmus, few detail which Parameters are imPortant, Leht~. (1976) in~icated that the eagle of onset from the midpoin t of the visual • field decreases as a function of increasing BAC. His data suggest .~'~~~.~'..i~b~'at at e. BAC of 0 IOZ, • ~...... ,.~Y'~,...,.... . ~- . . gaze nystagmus wil! occur mt abou.t 51 :~-. '~ negroes and, mta BAC of 0.20g, gaze uye.¢egmua will occur at about • 29 degrees. The correlation between the angle of onset and ,he BAC Was - .788 for 56 indivldumle. ' .... i- Most Other studi~ in which gaze .nystagmus has been measure d involve a cutoff point of 30-40 ~degrees. Use of a cutoff mm.y ~P!ain some of their conclusions. Fo~ exemplar Aachan (1958) used cu~0ff of ~0 degrees mad reported that $a~e uy,~tmgmu~ had a distinct threshold BAC of approximately O.06Z. Umedm and Sakata (1978) used a cutoff of 30 ,~grees end concluded that it was one of 82 the least sensitive eye measures of alcohol intoxication. These conclusion8 are not at mll surprising in view of the data that gaze nystagmus will occur at approzimately 41 degrees at a B~C of 0.1OZ. Aschan (1958) ha6 di6tingulohed between n "fine n gaze nysta"gmus I and a "course m 8are nyot~gmus. The latter tends to be m elo~, large amplitude movement of about lO degr~.eVo Fine nyetagmus tends to be a much smaller amplitude of about 4 degrees. We would expect that the difference in amplitude would only occur at ~ sufficiently high • AC for ~accadic eye movement (i.e.~ ie addition to ~ Bmooth movement~) to be impaired (Wilkinson et al~ 1974). When ~ the ~accmdic system is impaired~ a Imrger drift off targe~ may"be required for saccadic correction. Aschan (1958) also reports that g~ze nystagmus is more evident with monocular fixation than with binocular fixation. He reported that ~ubjects shotting monocular gaze nystagmu~ at 20 degrees would not show binocular gaze nystagmus until 40 degrees. Toglia (1976) reports that gage nystagmus tends to be greater in the left eye upon gazing to the left and in the right eye upon gazing t0 the right. Thes~ t~o ph@nomena may be the same. :, A1coh?.l ~ Balance While many studies use balances and coordination tests xn conjunction with alcohol impairmentp only a few studies have tried to manipulate important parameters in these tests. Balance tests of various sorts ~how large individual °.differences in the performance of sober indlviduals (i.e., Coldberg~ 1963), with older subjects (60-85 years) having much more difficulty than young (21-35 years) subjects (Wilson, Barboriak, and Koss, 1970).: Wilson et el (1970) observed that alcohol (mean BAC = 0.06Z) improved performance in the older subjects, but impaired performance in younger subjects. Both groups of subjects were tested for baseline performance and then given alcohol. The improvement seen in :the intoxicated older subjects may be due to the fact that balance tests show distinct learning curves (Goldberg, 1963), and the slider subjects have much more room for improvement (i.e., the ,baseline performance of older subjects was ten times worse than that of ~the younger individuals). It should be noted that Bardy, Elomaa, Huhmar, and Lehtovaara (1978) reported that age (between 18 and~ 67 years) had no significant effect on body sway. A number of variables, in addition to alcohols increase body, sway. These variables include exercise (Barnes, Cooke ~ King, and Passmore, 1965), sleep loss (Coldberg, 1963), increasing the room temperature from 65-68 F to 79-86 F (Goldberg, 1963), eating (Coldberg, 1963)~ and tranquilizers and antihistamines (Goldberg, 1966)o Is contrmsto ~ijiokikjien (1973), found that "controlled mttention ~ (i.e.~ counting background clicks) decreased body sway, One of the most important parameters in tests of balance, and muscul~r coordination is vision. Clo~ing the eyes makes all,of ~the b~lance tests much more difficult for sober and intoxicated iudividuale (Coldberg, 1963; Franks et al, 1976; Begbie,. 19.66; 83 Fregly~ Bergsted ~nd Cr~ybiel, 1967). ~egb!e (I~66) £nv~stiggted =balancing on a ~oving mtaud = under four conditlono: (1) eyes closed, lights off° (2) monitorin S an os©illoscope ~ith the lights off (i.e., no peripheral vision)~ (3) monitorlng an oscillo~cop e with lights on (i.e., llmited peripheral vision), ~nd (4) eyes open, lights on~ no ta~R (i.e., full perlpheral vision). The ¢ond£fions~ in ter~s of difficulty, were ~anked in the order presented ~i.eo9 eyes closed, lights off was the most difficult). .~: These ~data suggest that peripheral vision plays a particularly important role in maintaining balance. 1_.a_ ~alk-The-L~n~ Very few studies have looked specifically at the walk-the-l~ne tests. Fregley, Grsybiel, and Smith (1972) found that most individuals of both se~es could make 30 heel-to-toe steps ~ith their eyes closed sad arms folded across their chest without side stepping. In a second study, Fregley, Bergsted, and Craybiel (1967) found that ~alk-the-line performance (i.e.~ on 8-foot long, 3/4 inch rail ~ith eyes open) showed the maximum amount of deterioration just before ~ubjects reached their peak BAC of 0.I0~ and returned to normal in about two hours. 2_~. ~ne-Le~Stand O~ij a few studies have looked at variables affecting the one-leg- stand test. Fregley et al (1972) found that the leg used made no difference in the amount of time one could stanA on one leg (eyes cl~ed)o Most of Coldberg's findings on standing steadiness =/olved this tegto Thus, variables such ss sleep loss, alcohol, tranquilizers, food intake nnd warm temperatures appear to influence one's ability to stand on one lego ~oreover, the test is very difficult even for sober individuals with the eyes closed. 84 APPENDIX B INSTRUCTIONS TO SUBJECTS, QUESTIONS ASKED SUBJECTS, 4%ND SCORING AND DECISION SHEETS USED IN THE LABORATORY EVALUATION WALK AND TURN Instructions to the stopee Please assume a heel-to-toe position on the line with t your arms at your sides (demonstrate). When I tell you to, make nine heel-to-toe steps on the line in front of you, turn around, and return in nine heel-to-toe steps. Watch your feet at all times, making sure that you ~alk in a straight line and that every step is heel-to-toe, like this (demonstrate). Do you understand? (One rep ~ etition of one or two parts of the instructions is fine, but the entire instructions should not be repeated unless there is an obvious language problem.) Now begin and count your steps outloudo ONE-LEG STAND Instructions to the stopee: Please stand with your heels together and your arms at your sides (demonstrate and do not resume until the sus- pect is in the correct position). When I tell you to, I want you to raise one leg about 6 inches off the ground and hold that position while you count rapidly from i001 to 1030 (demonstrate). Do you understand? Now begin by raising either you right or left foot. NYSTAGMUS Instructions to the stop~e: I ~n going to check your eyes. }'leasu keep your llead still and follow this object (indicate what the stimulus is) to the side with your eyes. Keep your head straight and do not move your eyes back to center until I tell you to do so. ~ ,C 85 Par~icipanlt # Sex Officer ~ate of bfrth / Date .. . .. • Approx. weight QUEST IONS " Without looking, what time is it now? Actual time Have you been drinking? ~ How much? Are you too drunk to drive_? Which did you last eat? What did you eat at that tlme ) When did you last sleep? How many hours? Do you have"any physical defects? Yes No If yes, describe- Are you ill? Yes NO: Are you hurt? Yes No If yes, what:is: wrong? Have you recently been to a doctor? Yes :tlo ; a dentist? Yes No If yes, when? Reason for seeing doctor or dentist Are you taking medicine? Yes No If yes, what? Last dosetaken when? a,m. p,m. OBSERVATIONS CLOTHES: Orderly___ Mussed____ Soiled__ Disorderly___ Disarranged Describe BREATH (odor of alcoholic beverage): Strong__ !Moderate Faint :None ATTITUDE: Excited Hilarious Talkative ~ Carefree Sl~epy..... Combative Indifferent I.nsult~i.ng__ Cocky..... Cooperative Polite Other UNUSUAL ACTIONS: Hiccupping____ Belching___ Vomlting____ Fighting__ Profanity Other SPEECH: ' Incoherent.___ Mumbled S1urred___:~Confused Thick tongued Stuttered Accented Good 'Fai~r Other COLOR OF PACE: Normal Flushed Pale___ Uther EYES-'Normal Watery Bl,oodshot___ B6 Subject Time Date BAC Zye Problems C~ntact lenses Balance Problems Sc0ring Sheet for Sobriety Test Battery % Ao Walk and Turn i. Cannot keep balance while listening to instructions 2. Starts before instructions are finished. ~3. Keeps balance but does not remember instructions 4. Stops while walking to steady self 5o Does not touch heel-to-toe while walking l 6. Loses balance while walking (i.e., steps off line) 7. Uses arms for balance 8. Loses balance while turning 9. Incorrect number of steps • i0. Cannot do the test (equal to I0 checkmarks) A. TOTAL Bo O_ne L~ Stand i. Swaying while balancing 2o Uses arms to balance 3. Slightly unsteady 41. Quite unsteady 5.• Starts before instructions are •finished • 6. Puts foot down • ~ 7. Cannot do/or test discontinued (equal to 7 checkmarks) B. TOTAL A.+B. TOTAL Co Alcoh01 Gaze Nystagmus .(AGN) RIGHT EYE LEFT EYE Io Onset of AGN at less than 45 ° and with at least 10%~of •the white showing. 2. Estimated angle of onset. 3o Eyes cannot follow smoothly 4. AGN at maximum lateral deviation: •~ a. absent R L b. minimal ' R L c. moderate R L d. heavy R L 5oAGN at maximum lateral deviation is moderate or stronger 87 C. TOTAL -?-- • SU~tMARY .OF SCORING : NUMBER OF CHECK~U%RKS WALK AND TURN ONE-LEG STAND BALANCE TOTAL NYSTAGMUS '%E=, DECISION CR.ITERI A based upon our pilot work ~ A; 3 or,more checks on balance plus at least a score of 2 on the ny- stagmus will correc-~ classify about 75% of those above : ~ i .10% and will incorrectly classify about 15% of those below .10% " B. 2 o[ more checks on balance plus at least 2 on nystagmus will cor- rectly classify about 75% of those above .075% and will incorrectly classify about 10% of those below 075% c. 1 or more checks on balance plus nystagmus onset of 50 ° or less will • • •" , .... • correctly classify 80% of those above~ .05% and incorrect L .... ly classify about 15% of those below ~05%. A. ESTIMATE THIS PERSON'S BAC TO WITHIN .01% ON A SCALE OF 1 TO• i0 (l=uncertain; 10=very sure) ESTIMATE YOUR CONFIDENCE IN YOUR ESTIMATE OF THE BAC. " } B. IS.THIS PERSON IMPAIRED BY ALCOHOL? .. YES ON THE SAME. SCALE WHAT IS YOUR CONFIDENCE NO IN THE ABOVE? i / C. WOULD YOU ARREST THIS PERSON UNDER YOUR NO~ CRITERIA? YES NO "i c: • i 88 "" '-c '%c . , • j t 4"'.