sign in sign up
<<
Home , Argument from analogy, Circular reasoning

SUMMER 2014, VOL. 29, NO. 2 35

Circles and in Public Health Reasoning

Louise Cummings School of Arts and Humanities Nottingham Trent University, UK

Abstract

7KHVWXG\RIWKHIDOODFLHVKDVFKDQJHGDOPRVWEH\RQGUHFRJQLWLRQVLQFH&KDUOHV+DPEOLQFDOOHG IRUDUDGLFDOUHDSSUDLVDORIWKLVDUHDRIORJLFDOLQTXLU\LQKLVERRN)DOODFLHV7KH³ZLWOHVV H[DPSOHVRIKLVIRUEHDUV´WRZKLFK+DPEOLQUHIHUUHGKDYHODUJHO\EHHQUHSODFHGE\PRUH authentic cases of the in actual use. It is now not unusual for and argumentation theorists to draw on actual sources for examples of how the fallacies are used in our everyday UHDVRQLQJ+RZHYHUDQDVSHFWRIWKLVPRYHWRZDUGVJUHDWHUDXWKHQWLFLW\LQWKHVWXG\RIWKH fallacies, an aspect which has been almost universally neglected, is the attempt to subject the fallacies to empirical testing of the type which is more commonly associated with psychological experiments on reasoning. This paper addresses this omission in research on the fallacies by examining how subjects use two fallacies – circular and analogical argument – during a reasoning task in which subjects are required to consider a number of public health scenarios. Results are discussed in relation to a view of the fallacies as cognitive heuristics that facilitate reasoning in a context of uncertainty.

Keywords: analogical argument, circular argument, heuristic, , public health, reasoning, uncertainty

I. Introduction puns, anecdotes, and witless examples of his forbears. The study of the fallacies has changed :KHUHSKLORVRSKLFDOUHÀHFWLRQRQWKH considerably in the forty years since Charles IDOODFLHVEHIRUH+DPEOLQ¶VSRZHUIXOFULWLTXH +DPEOLQUDLOHGDJDLQVWWKHVKRUWFRPLQJVRI ZDVEDVHGXSRQDUWL¿FLDOO\FRQVWUXFWHG the so-called Standard Treatment of these examples which bore little or no resemblance DUJXPHQWVLQORJLFWH[WERRNV7KXV+DPEOLQ to the types of that are found in S ZULWHV HYHU\GD\UHDVRQLQJWKHSRVW+DPEOLQHUDLQ $QGZKDWZH¿QGLQPRVWFDVHV,WKLQN fallacy analysis has witnessed an increase in it should be admitted, is as debased, research which places emphasis on the types worn-out and dogmatic a treatment of arguments that people actually employ. as could be imagined – incredibly Examples of fallacious arguments are now tradition-bound, yet lacking in logic much more likely to come from sources and historical sense alike, and almost such as newspapers, magazines, and other without connection to anything else in media outlets than they are to originate in modern logic at all. This is the part of the misguided attempts of fallacy theorists to his book in which a writer throws away FRQVWUXFWSODXVLEOH EXWDUWL¿FLDO DUJXPHQWV logic and keeps his reader’s attention, To see that this is the case, one need only if at all, only by retailing traditional ORRNDW:DOWRQ  ZKHUHH[DPSOHVRI 36 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES fallacious arguments from actual sources are be regarded as an incorrect discussion move. in abundance. Examples include a discussion In the pragma-dialectic approach, fallacies are of argument based on an article in analyzed as such incorrect discussion moves the National Post   DSSHDOVWRDXWKRULW\ in which a discussion rule has been violated” using an article featured in Newsweek    YDQ(HPHUHQDQG*URRWHQGRUVWS  and the argument from popularity based on a )RUIXUWKHUGLVFXVVLRQRISUDJPDGLDOHFWLFV televised interview with a politician during the VHH&XPPLQJV D  Canadian Broadcasting Company programme But appearances can be deceptive. This Week in Parliament    Notwithstanding the emphasis on the role The move to more naturalistic data of the arguer in reasoning and on examining for examination by fallacy theorists has DFWXDODUJXPHQWVDVRSSRVHGWRDUWL¿FLDOO\ XQGRXEWHGO\KDGEHQH¿WVIRUWKHWKHRUHWLFDO constructed arguments, there is a very real frameworks of the fallacies that have sense in which fallacy theorists are still failing emerged. The increased authenticity of these to engage with actual reasoners. Almost frameworks is a direct consequence of efforts nothing is known about the cognitive and to engage with people’s actual reasoning. One other psychological processes that people use aspect of this improved authenticity is the to assess that group of fallacious arguments development of new criteria for the evaluation DERXWZKLFK+DPEOLQKDGPRVWWRVD\WKH of arguments. Arguments are now as likely to informal fallacies. This should be compared be characterised as fallacious if they violate to the extensive psychological literature that discussion rules in a dialogue or conversation exists on and fallacies as they are if they fall short of some deductive and the equally large literature that exists on standard of or soundness. This new induction and errors of probabilistic reasoning. emphasis on people’s actual reasoning is For example, a prominent psychologist of part of a wider pragmatic turn in the study of GHGXFWLRQ3KLOLS-RKQVRQ/DLUGKDVSURSRVHG IDOODFLHVDWXUQLQZKLFKDVSHFWVRIFRQWH[W D a mental models theory to account for people’s QRWLRQGHVSLVHGE\HDUOLHUWKHRULVWV PXVWEH SHUIRUPDQFH LQFOXGLQJIDOODFLHV LQGHGXFWLYH considered in any assessment of an argument. UHDVRQLQJWDVNV VHH-RKQVRQ/DLUGDQG%\UQH The arguer and his or her motivations for  IRUH[WHQVLYHGLVFXVVLRQ 7KHZRUNRI producing an argument is an integral part of $PRV7YHUVN\DQG'DQLHO.DKQHPDQ   context. The person who produces and receives ZDVPRVWQRWDEOHDPRQJWKH¿UVWLQYHVWLJDWRUV arguments, it seems, has never before been so to study the types of errors subjects committed important to the study of the fallacies. GXULQJSUREDELOLVWLFUHDVRQLQJ 6HH One example of this pragmatic turn is Kahneman’s 2011 engaging account of their the pragma-dialectical framework of Frans van research in Thinking, Fast and Slow.) (HPHUHQDQG5RE*URRWHQGRUVW  In contrast , no one has subjected  7KLVIUDPHZRUNGUDZVRQVSHHFKDFW the informal fallacies to experimental theory and other insights of Grice and Searle investigation, even though similar studies in developing an account of argumentation. of other areas of logic have yielded useful Speech acts that violate one of the rules results about the logical processes that people for a critical discussion and undermine ¿QGUDWLRQDOO\FRPSHOOLQJ7KHUHDUHYDULRXV attempts to resolve a difference of opinion are reasons for this omission. Chief amongst them FKDUDFWHUL]HGDVIDOODFLHV³$Q\LQIULQJHPHQW is an aversion to psychologism in logic. I of one or more of the rules, whichever party concur with the stance on psychologism taken commits it and at whatever stage in the E\*DEED\DQG:RRGV WRDSSHDU LQWKHLU discussion, is a possible threat to the resolution HPSLULFDOO\VHQVLWLYHORJLF³,QYHVWLJDWRUV of a difference of opinion and must therefore who make room for context and agency are SUMMER 2014, VOL. 29, NO. 2 37 drawn to a form of what used to be called the medical professionals who provide advice on Laws of Thought approach and, accordingly, risk to the public. There is evidence that on the are committed to an element of psychologism issue of BSE at least, medical professionals in logic. . . . Psychologism is once again an ODFNHGVXI¿FLHQWNQRZOHGJHWRJLYHDGYLFHWR open question in the research programme of the public about the disease. Simpson et al. logical theory. Its re-emergence should not  FRQGXFWHGDVXUYH\RINQRZOHGJHRI be prejudged. Better to wait and see how, BSE among 1,038 doctors in North Yorkshire, once it is up and running, a psychologically England. These investigators found that most real, agent-based logic fares as a theory of GRFWRUV  WKRXJKWWKDWWKHLUNQRZOHGJHRI reasoning” LWDOLFVLQRULJLQDO . BSE was inadequate for them to give advice to Indeed the very fact that other branches the public. of logic have survived the approaches of Alternatively, it may be a member of SV\FKRORJ\ DQGQRWMXVWVXUYLYHGWKHVH the public who is experiencing uncertainty, DSSURDFKHVEXWSRVLWLYHO\EHQH¿WHGIURP as when a person is attempting to assess WKHP VKRXOGUHDVVXUHIDOODF\WKHRULVWVWKDW WKHVLJQL¿FDQFHRIVRPHSXEOLFKHDOWKLVVXH their particular corner of logic can only be EXWODFNVWKHVFLHQWL¿FNQRZOHGJHWKDWLV enhanced by input from psychological studies required to undertake such an assessment. of the informal fallacies. At least this is my In both scenarios, a lack of knowledge can starting point for the discussion in the current be successfully bridged through circular paper. The arguments investigated in this study and analogical reasoning. To the extent that – circular argument and analogical argument – these reasoning strategies permit scientists have been extensively discussed in the fallacy and lay reasoners to move beyond gaps in and argumentation literature. But no one has their knowledge and make decisions about previously attempted to elicit responses from public health issues, these strategies can subjects concerning the conditions under be seen to serve as cognitive heuristics that which these arguments are judged to be more facilitate reasoning in contexts of uncertainty. or less warranted. This paper undertakes to do The heuristic value of these strategies for just that in the context of a task that examines scientists addressing the emergence of BSE public health reasoning. DQG+,9$,'6KDVEHHQGHVFULEHGLQYDULRXV The choice of public health reasoning SXEOLFDWLRQVWRGDWH &XPPLQJV as a context in which to examine circular 2009, 2010b, 2011, 2012a-c, 2013a-b, 2014a- and analogical arguments is motivated by HDE 7KLVLVWKH¿UVWVWXG\WRH[DPLQH the following considerations. The central how these same strategies are employed by lay claim of this paper is that many informal reasoners, who must also cope with uncertainty fallacies function as cognitive heuristics in dealing with public health problems, which can facilitate reasoning in contexts albeit uncertainty that is generated through a of uncertainty. The uncertainty in question UDWKHUOLPLWHG LQVRPHFDVHVQRQH[LVWHQW  may with scientists themselves, as when NQRZOHGJHRIWKHVFLHQWL¿FLVVXHVLQYROYHG,W they are confronting an emerging infectious LVH[SHFWHGWKDWWKHOD\UHDVRQHUZLOODOVR¿QG disease such as Severe Acute Respiratory these reasoning strategies to be a facilitative 6\QGURPH 6$56 RU%RYLQH6SRQJLIRUP cognitive resource in the face of uncertainty. (QFHSKDORSDWK\ %6( DQGWKHUHLVOLWWOHRUQR knowledge of how the pathogens responsible II. Arguments as Cognitive Heuristics for these diseases will behave. This lack of knowledge often extends beyond those While the idea that certain fallacies scientists who are directly charged with can function as cognitive heuristics during responding to a public health problem to other reasoning about public health problems is 38 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES new, the place of heuristics in reasoning heuristic information processing. The is actually well established. Reasoning systematic approach…is deliberative, KHXULVWLFVZHUH¿UVWV\VWHPDWLFDOO\VWXGLHG attends to detail, weighs alternative E\$PRV7YHUVN\DQG'DQLHO.DKQHPDQ views, and assesses argument quality ,QWKHLUODQGPDUNDUWLFOH³-XGJPHQWXQGHU in judging the validity of persuasive 8QFHUWDLQW\+HXULVWLFVDQG%LDVHV´7YHUVN\ messages. The heuristic approach DQG.DKQHPDQ S GHVFULEHGD is alert to cues (e.g., trusted groups’ number of heuristics that reasoners use in evaluation of the information) and SUREDELOLVWLFUHDVRQLQJ³0DQ\GHFLVLRQVDUH simple decision rules (if encoded in based on beliefs concerning the likelihood of memory, accessible to recall, and uncertain events…people rely on a limited deemed reliable) justifying quick number of heuristic principles which reduce intuitive judgment.” the complex tasks of assessing probabilities :LOVRQHWDO  H[DPLQHGKHXULVWLF and predicting values to simpler judgmental processing in a study of how adults assessed operations.” Although Tversky and Kahneman ULVNVDVVRFLDWHGZLWKJHQHWLFDOO\PRGL¿HGIRRG acknowledged that heuristics could be crops. useful, they were concerned to emphasise So heuristics are not a new WKH³VHYHUHDQGV\VWHPDWLFHUURUV´WRZKLFK phenomenon in the study of reasoning or even heuristics could lead. For example, one such the study of public health reasoning. But what error is known as the gambler’s fallacy, the is novel about the approach adopted in this EHOLHIWKDWUDQGRPSURFHVVHVVHOIFRUUHFW³LI paper is that no-one has previously attempted [a random] sequence has strayed from the to cast certain informal fallacies in terms of population proportion, a corrective bias in cognitive heuristics that people employ when WKHRWKHUGLUHFWLRQLVH[SHFWHG´ 7YHUVN\DQG they form judgements about public health .DKQHPDQS  )RUDQH[WHQVLYH problems. Yet, there is much to recommend and up-to-date account of their research this approach. For those philosophers and SURJUDPVHH.DKQHPDQ logicians who have subjected the informal By the time heuristics began to receive fallacies to serious scrutiny, the rewards sustained attention in the literature on risk, have been plentiful. Quite apart from being an altogether more benign view of these examples of weak, bad, or shoddy reasoning, cognitive operations was beginning to take the informal fallacies have been found to be shape. The emphasis now was less on the rationally warranted arguments within certain errors in reasoning to which heuristics might contexts of use. A sizeable literature now exists lead and more on their facilitative function on non-fallacious variants of most of the major in dealing with complex problems. Some of fallacies. The work of two fallacy theorists – these problems concerned issues of public 'RXJODV:DOWRQDQG-RKQ:RRGV±KDVEHHQ KHDOWKVLJQL¿FDQFH7UXPER  DQG SDUWLFXODUO\LQÀXHQWLDOLQWKLVUHJDUG,QDODUJH -RKQVRQ  GLVFXVVHGKHXULVWLFVLQWZR number of books and journal articles, these studies that applied a heuristic-systematic theorists have described non-fallacious forms model to the assessment of risk. These studies of petitio principii  EHJJLQJWKHTXHVWLRQ  required subjects to assess risks that were argumentum ad ignorantiam  WKHDUJXPHQW communicated in epidemiological information IURPLJQRUDQFH DQG DERXWFDQFHUUDWHV 7UXPER DQGZKLFKZHUH WKHDUJXPHQWIURPWKHVWLFNRUDSSHDOWR posed by a semi-hypothetical industrial facility IRUFH DPRQJVWRWKHUV :DOWRQ -RKQVRQ -RKQVRQ S VWDWHVWKDW :RRGV 2IFRXUVHLQHPSKDVL]LQJ “[a]…heuristic-systematic model that there exist non-fallacious variants of (HSM) separate[s] systematic from the fallacies, this is not the same as saying SUMMER 2014, VOL. 29, NO. 2 39 that there are no such things as fallacies. For is a fundamental logical law: it is the principle discussion of novel fallacious arguments in the of identity.” context of the BSE problem, see Cummings +RZHYHUDSUHPLVHWKDWLVDV   unknown as the conclusion to be proved, An altogether smaller literature which must be the case in any argument in has sought to describe how these fallacies which the premise and conclusion are one function non-fallaciously within the context and the same proposition, cannot satisfy the RILPSRUWDQWSXEOLFKHDOWKLVVXHV &XPPLQJV epistemic requirement that the premise of 2002, 2004, 2009, 2010b, 2011, 2012a-c, an argument should be more certain than, DEDHDE :KDWWKHVHODWWHU or better known than, the conclusion to be studies have revealed is that these argument SURYHG$FFRUGLQJWR%LUR S WKH forms can sustain reasoning in contexts problem with question-begging argument is that preclude other modes of reasoning, WKDWLWLV³HSLVWHPLFDOO\QRQVHULRXV´ZKHUH principally deduction and induction. These epistemic seriousness describes an argument contexts exhibit pervasive uncertainty of the in which the premises are more knowable than type commonly encountered at the outset of a the conclusion. It is the perceived failure of VFLHQWL¿FLQTXLU\RUZKHQGHOLEHUDWLRQVH[FHHG circular argument to develop the grounds for a reasoner’s available cognitive resources. In a thesis that has seen this particular argument the sections to follow, we examine the features form almost universally condemned by of two such heuristics, circular argument and WKHRULVWV)RU5HVFKHU S FLUFXODU DQDORJLFDODUJXPHQW7KH¿QGLQJVRIDVWXG\ sequences in disputation must be blocked that examines the conditions under which VLQFHWKH\³IUXVWUDWHWKHDLPRIWKH>GLDOHFWLFDO@ reasoners accept and reject these argument HQWHUSULVH´ZKLFKLV³WRGHHSHQWKHJURXQGLQJ IRUPVDUHGHVFULEHG,QUHSRUWLQJ¿QGLQJV of the contentions at issue.” If the only ground from a psychological study of the fallacies, that can be advanced in support of a thesis this study goes beyond the exclusively is the thesis itself, it is argued, then we have philosophical literature that exists on the not established or proven a response to the informal fallacies. TXHVWLRQDWLVVXHVRPXFKDVPHUHO\³EHJJHG´ for its acceptance. It seems that extensive II.A Circular Argument philosophical preoccupation with circular or question-begging argument has produced few Circular argument, also variously known as defenders of this argument form and even less petitio principii , question-begging argument to recommend it by way of rational merits. and circulus probandi , has long been a source As question-begging argument began of fascination for philosophers and logicians. to be submitted to a more systematic treatment This argument is a strange hybrid of deductive LQDSRVW+DPEOLQHUDRIIDOODF\DQDO\VLV validity and epistemic unacceptability. An it soon became apparent to analysts that argument of the form ‘p, therefore p’ is arguments with a distinctly circular form were none other than the principle of identity in not all inherently fallacious. Indeed, many IRUPDOORJLF:RRGVDQG:DOWRQ S VXFKDUJXPHQWVZHUHPHUHO\UHÀHFWLQJFHUWDLQ  GHVFULEHKRZ³>D@UJXPHQWVRIWKHIRUP QDWXUDOF\FOLFDOSURFHVVHVDWZRUNLQ¿HOGV ‘p, therefore p’ always or nearly always such as economics, mathematics, geology, beg the question, yet their formal validity is SDODHRQWRORJ\DQGHYHQHSLVWHPRORJ\ :DOWRQ LPSHFFDEO\UHÀHFWHGLQVWDQGDUG¿UVWRUGHU &XPPLQJV )RUH[DPSOHZLWK ORJLF´3HUHOPDQ S VWDWHVWKDW UHJDUGHFRQRPLFV:DOWRQ S  ³>W@KHDI¿UPDWLRQLISWKHQSVWDWLQJWKDWD remarks that proposition implies itself, is not only true but “in the majority of circular arguments 40 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES

we looked at, the circularity well-known, established theses at the disposal cannot be condemned as wrong or RILQYHVWLJDWRUV7RLQVLVWWKDWVFLHQWLVWVIXO¿OD fallacious precisely because the SULRULW\UHTXLUHPHQWDWWKLVVWDJHRIVFLHQWL¿F context of dialogue fails to indicate inquiry is to misrepresent the epistemic decisively that a priority condition standing of the propositions that are available is a procedural requirement. The to scientists in this context. A more sensible economist’s argument we began with, epistemic policy is to suspend this priority for example, should not be declared requirement until such times as a knowledge fallacious or viciously circular by a base of well-established propositions is reasonable critic unless the critic can available to investigators. In the interim cite evidence of an agreement, or at period, it is quite legitimate for scientists to least a clearly agreed upon context or use a proposition as a premise in argument that background requirement to argue only is no better established than the conclusion to in one direction or the other…If the be proved. objective (the problem) is to prove from ,Q&XPPLQJV E LWLVDUJXHG A to B, and also from B to A, there need that circular argument was used non- be no fallacy in solving the problem by fallaciously by scientists who sat on the arguing in a circle.” Tyrrell Consultative Committee on Research With regard to mathematical arguments, during the BSE epidemic in the UK. The :DOWRQ S DOORZVWKDW assumption of the question-at-issue in the “[i]n mathematics, it is common deliberations of this committee permitted practice to start at proposition A and scientists to advance their investigations until then prove B, then start again at B and such times as evidence that was independent prove that A follows. An equivalence of the conclusion to be proved could be proof in mathematics, of the if and only obtained. Circular argument served as an if type, often takes this form. Although effective and rationally warranted cognitive the form of proof is circular, in many KHXULVWLFIRUVFLHQWL¿FLQYHVWLJDWRUVLQWKLV instances such a proof is rightly context, a heuristic that enabled scientists to thought non-fallacious.” inch forward in inquiry at a time when little was known about BSE. Circular argument, it The decisive factor in an evaluation is argued, has a similar heuristic value for the of question-begging argument is now less lay reasoner who is attempting to assess the the structure of these arguments than the VLJQL¿FDQFHRISXEOLFKHDOWKSUREOHPVLQWKH SXUSRVH V IRUZKLFKWKH\ZHUHDGYDQFHG,I DEVHQFHRINQRZOHGJHRIWKHVFLHQWL¿FLVVXHV that purpose included a requirement to argue involved. It is expected that lay reasoners from better known, or more established, will judge this argument to be rationally propositions to less well known, or less warranted under certain epistemic conditions established, propositions, then the circular and unwarranted under other conditions. structure of question-begging argument is Where this argument is seen to advance some FOHDUO\SUREOHPDWLF+RZHYHULIWKHUHLVQR type of inquiry – either a reasoner’s personal such requirement for evidential or epistemic cognitive inquiry into an issue or the more priority in a particular context of argument, formal inquiry of scientists into a problem – in then an accusation of fallacy against the the absence of evidence or knowledge relating proponent of a circular argument is altogether to a topic, then this argument may be assessed OHVVWHQDEOH$WWKHRXWVHWRIVFLHQWL¿FLQTXLU\ by reasoners as being rationally warranted particularly inquiry into a newly emerging or as possessing some epistemic virtue. By infectious disease like BSE, there are few implication, if evidence which is independent SUMMER 2014, VOL. 29, NO. 2 41 of the conclusion to be proved is available to P. From these two premises the reasoner the reasoner, but the reasoner neglects to use then derives the conclusion that B also has it in favour of the question-at-issue, then it SURSHUW\3$V%URZQ S SXWVLW might reasonably be expected that this type ³$QDORJ\LVRIWHQH[SODLQHGDVDVSHFLDONLQG of reasoning would be unfavourably assessed RIFRPSDULVRQ RUVLPLODULW\ EHWZHHQWZR and judged to be lacking in rational warrant. REMHFWV HYHQWVLGHDVFODVVHVRIREMHFWVHWF  The study described below in Section III will such that the possession in common of one establish the conditions under which subjects RUPRUH FKDUDFWHULVWLF SURSHUW\DWWULEXWH are inclined to identify circles in reasoning as HWF E\ERWKREMHFWVLVEHOLHYHGWRLPSO\WKDW virtuous and vicious, respectively. the two objects probably have some other FKDUDFWHULVWLF V LQFRPPRQ´ The following example of analogical II.B Analogical Argument argument is based on the reasoning of scientists in the UK during the BSE epidemic: A second fallacy that has attracted condemnation and praise in roughly equal Scrapie in sheep and BSE in cattle measure from philosophers and logicians are similar in certain respects. is the argument from . One such Scrapie is not transmissible to FRQGHPQDWRU\YRLFHZDVWKDWRI-RKQ6WXDUW humans. Mill. Mill remarked as follows of analogy: Therefore, BSE will not be ³7KLVYHU\FRPPRQDEHUUDWLRQ>RYHUUDWLQJ transmissible to humans. the probative force of analogy] is sometimes supposed to be particularly incident to persons This particular analogical argument was GLVWLQJXLVKHGIRUWKHLULPDJLQDWLRQEXWLQ established on the basis of the most tentative reality it is the characteristic intellectual vice ODUJHO\HSLGHPLRORJLFDO HYLGHQFH± of those whose imaginations are barren…To only strain-typing studies could establish such minds objects present themselves clothed GH¿QLWLYHO\LIVFUDSLHDQG%6(ZHUHUHODWHG LQEXWIHZSURSHUWLHVDQGDVWKHUHIRUHIHZ diseases and these studies were not available analogies between one object and another WRVFLHQWLVWVLQZKHQ%6(¿UVWHPHUJHG occur to them, they almost invariably overrate in British cattle. The analogy established in the WKHGHJUHHRILPSRUWDQFHRIWKRVHIHZ´  ¿UVWSUHPLVHRIWKLVDUJXPHQWZDVWKHUHIRUHDW S  EHVWZHDNO\ZDUUDQWHG KHQFHWKHVWDWXVRIWKLV Analogical argument has been DUJXPHQWDVDZHDNDQDORJ\ 

bleeding, severe breathing  :KHUHGLGWKHRXWEUHDNRILOOQHVV GLI¿FXOWLHVDQGDIDWDOSQHXPRQLD occur? Few people who develop the  :HUHSDUDVLWRORJLVWVSDUWRIWKH disease survive it. The symptoms HPHUJHQF\WHDPIURP:+2" DUHQRQVSHFL¿FDQGDUHD feature of many diseases in the Analogical Argument Example DUHD:+2¶VLQIHFWLRXVGLVHDVH Analogical arguments were assessed scientists make a preliminary using a similar format. As with circular assessment of the situation and arguments, four questions followed each tentatively claim that a virus passage. Two of these questions examined EHORQJLQJWRWKH+DHPRVWULDWKXV information presented in the passage and family may be responsible for required either a yes/no response or a minimal WKHGLVHDVH:+2GLVSDWFKHVDQ response of just a few words. The question emergency team of virologists, targeting the analogical argument in the parasitologists and epidemiologists SDVVDJHDVNHGVXEMHFWVWRUDWHDVFLHQWL¿F to the region in order to establish if response to a public health problem as valid, RQHRIWKH+DHPRVWULDWKXVYLUXVHV moderately valid or not valid at all. A fourth is responsible for the disease. question asked subjects to explain their Apart from reports of symptoms, response to the question that probed analogical these scientists have no evidence argument. The open-ended nature of the upon which to initiate their response was intended to reveal the types of investigation. They are unable to IDFWRUVWRZKLFKVXEMHFWVDWWDFKHGVLJQL¿FDQFH obtain case histories from most in rating an analogical argument. of the victims, a large number of The following passage and questions whom are dead or too seriously examined an analogical argument with the ill to be questioned. The scientists features : are therefore unable to determine if the victims of the disease were Analogical Argument Scenario near water sources that are known In 1981, reports of a rare form WRKDUERXU+DHPRVWULDWKXVYLUXVHV RISQHXPRQLDLQ¿YHSUHYLRXVO\ In the absence of evidence, they healthy homosexual males treated DVVXPHWKDWD+DHPRVWULDWKXVYLUXV in three Los Angeles hospitals is causing the disease and proceed appeared in the Morbidity and to test the blood of hospitalised Mortality Weekly Report in the patients for the protein markers of United States. Little did the global these viruses. VFLHQWL¿FFRPPXQLW\NQRZWKHQ that it was witnessing early cases Questions on Circular Argument Scenario of what was to become known  7KHHPHUJHQF\WHDPIURP:+2 DV$,'6,QLWLDOO\WKH&HQWHUV assumed WKDWD+DHPRVWULDWKXVYLUXV IRU'LVHDVH&RQWURO &'& GLG was the cause of the illness even as not know what pathogen was they were attempting to establish if this responsible for this new disease. was the case. Please circle one answer %HIRUHWKH+,9YLUXVZDVHYHQ to the following: was this strategy LGHQWL¿HGWKH&'&LVVXHGDGYLFH L YDOLG LL PRGHUDWHO\YDOLG  on how the public could best LLL QRWYDOLGDWDOO" SURWHFWLWVHOIDJDLQVW$,'67KLV  3OHDVHH[SODLQ\RXUUHVSRQVHWR   advice included information 46 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES

about safe sex practices and the communicated to all subjects. At the outset avoidance of needle sharing of the study, all scenarios were examined by intravenous drug users. The by two public health consultants and two &'&¶VDGYLFHZDVEDVHGRQLWV academic linguists. This was done with a view belief that the causal agent of to establishing, respectively, the plausibility $,'6PXVWEHDEORRGERUQH LQSXEOLFKHDOWKWHUPV RIWKHVFHQDULRV virus given similarities between and the comprehensibility of the linguistic the population groups who were constructions used to characterize them. GHYHORSLQJ$,'6DQGWKRVHJURXSV that were susceptible to another IIIC. Subjects blood-borne virus, hepatitis B. These groups were homosexual A total of 879 subjects participated in males, intravenous drug users and the study. All subjects were between 18 and 65 recipients of blood transfusions. years of age. Subjects could be male or female, &'&VFLHQWLVWVKDGH[WHQVLYH of any ethnic or socioeconomic background experience of the hepatitis B and could be educated to either university level YLUXVE\WKHWLPHWKH¿UVWFDVHV or secondary school level. The characteristics RI$,'6HPHUJHG&RQVHTXHQWO\ of all respondents to the questionnaire are WKH\KDGOLWWOHGLI¿FXOW\DGDSWLQJ shown in Table A in the Appendix. The three public health information used versions of the questionnaire received the for hepatitis B to address this new VDPHQXPEHURIUHVSRQGHQWVYHUVLRQ$  disease that confronted them. VXEMHFWV YHUVLRQ% VXEMHFWV YHUVLRQ & VXEMHFWV 6XEMHFWVZHUHUHFUXLWHGWR Analogical Argument Questions the study through a combination of methods.   'LGWKH¿UVWFDVHVRI$,'6LQWKH86 For the most part, the participation of subjects appear in hospitals in San Francisco? was secured through a series of formal   7KH&'&EHOLHYHGWKDW$,'6ZDV recruitment activities which were undertaken caused by a blood-borne virus. Please in several venues, including public areas in circle one answer to this question: local hospitals, staff dining facilities in large :DVWKH&'&¶VEHOLHI L YDOLG LL  retail outlets and the lounge areas of private PRGHUDWHO\YDOLGRU LLL QRWYDOLGDWDOO" health clubs. A smaller number of subjects  ,QZKLFKSRSXODWLRQJURXSZHUHWKH were recruited through a technique known ¿UVW¿YHFDVHVRI$,'6GRFXPHQWHG" as snowball sampling. In this technique,   3OHDVHH[SODLQ\RXUUHVSRQVHWR   respondents to the questionnaire either offered to provide, or were asked to provide, All responses were written on the the contact details of other individuals who questionnaire which was completed might be willing to participate in the study. anonymously. Subjects were informed that Questionnaires were subsequently sent to these the task would take approximately 30 minutes individuals, some of whom recommended, to complete and that all data and responses in turn, other people who could participate ZHUHFRQ¿GHQWLDO7KH\ZHUHDGYLVHGWR in the investigation. Snowball sampling is undertake the exercise in a distraction-free a particularly effective way of recruiting environment and not to consult sources subjects for a study when target groups are such as books and the internet, as questions known to be inaccessible for a range of reasons were intended to elicit judgements from %URZQH ³$VDPSOLQJSURFHGXUH subjects rather than correct answers. A PD\EHGH¿QHGDVVQRZEDOOVDPSOLQJZKHQ deadline for return of the questionnaire was the researcher accesses informants through SUMMER 2014, VOL. 29, NO. 2 47 contact information that is provided by other informants. This process is, by necessity, repetitive: informants refer the researcher to other informants, who are contacted by the researcher and WKHQUHIHUKHURUKLPWR\HWRWKHULQIRUPDQWVDQGVRRQ+HQFHWKHHYROYLQJ³VQRZEDOO´HIIHFW captured in a metaphor that touches on the central quality of this sampling procedure: its DFFXPXODWLYH GLDFKURQLFDQGG\QDPLF GLPHQVLRQ´ 1R\S  It was predicted at the outset of the study that secondary school educated males would be particularly unwilling to participate in an investigation of this type. Level of formal education is almost certainly a key factor in the poor engagement of secondary school males in this study. +RZHYHUDQRWKHUIDFWRULVOLNHO\WREHWKHWRSLFRIWKHVWXG\7KHUHLVHYLGHQFHWKDWELRORJLFDO and health sciences are of less interest to men than the physical sciences. As a health-related GLVFLSOLQHSXEOLFKHDOWKVFLHQFHPD\VLPSO\QRWHQJDJHWKHLQWHUHVWRIPHQVXI¿FLHQWO\IRUWKHP WRZDQWWRHQJDJHLQWKLVVWXG\ 6HHFKDSWHULQWKH6FLHQFHDQG(QJLQHHULQJ,QGLFDWRUV 1DWLRQDO6FLHQFH%RDUG IRUDGHWDLOHGGLVFXVVLRQRIWKHUROHRIIRUPDOHGXFDWLRQDQGVH[ LQERWKLQWHUHVWLQVFLHQFHDQGSHUIRUPDQFHLQVXUYH\VRIVFLHQWL¿FOLWHUDF\ 'HVSLWHFRQVLGHUDEOH HIIRUWVWRUHFUXLWWKHVHVXEMHFWVWKH¿QDOUHFUXLWPHQW¿JXUHVUHYHDOHGWKDWVHFRQGDU\VFKRRO educated males did indeed constitute an underrepresented group in the study: secondary school PDOHV VXEMHFWV VHFRQGDU\VFKRROIHPDOHV VXEMHFWV XQLYHUVLW\PDOHV VXEMHFWV  XQLYHUVLW\IHPDOHV VXEMHFWV  TABLE 1: Circular argument

Results: VIRTUOUS CIRCLE POSITIVE OUTCOME Valid: 57.9% Moderately valid: 29.8% Not valid at all: 12.3%

Results: VIRTUOUS CIRCLE LACK OF EVIDENCE Valid: 28.6% Moderately valid: 43.5% Not valid at all: 27.9%

Results: VICIOUS CIRCLE NEGATIVE OUTCOME Valid: 31.1% Moderately valid: 25.6% Not valid at all: 43.3%

Results: VICIOUS CIRCLE ABUNDANT EVIDENCE Valid: 10.6% Moderately valid: 29.5% Not valid at all: 59.9%

48 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES

TABLE 2: Analogical arguments

ANALOGY

STRONG WEAK

SCENARIO

Strong analogy Weak analogy Actual scenario Actual scenario

ACTUAL Results: Results: Valid: 51.4% Valid: 10.3% Moderately valid: 41.3% Moderately valid: 37.8% Not valid at all: 7.3% Not valid at all: 51.9%

Strong analogy Weak analogy Non-actual scenario Non-actual scenario

NON-ACTUAL Results: Results: Valid: 49.1% Valid: 5.2% Moderately valid: 41.5% Moderately valid: 20.3% Not valid at all: 9.4% Not valid at all: 74.5%

of conclusion-independent evidence which IIID. Results LQYHVWLJDWRUVIDLOHGWRXVH YDOLG  Between these extremes, subjects were almost Full results for circular argument and equally inclined to judge the argument as analogical argument are displayed in Tables 1 valid when there was a lack of conclusion- and 2, respectively. Results for both types of independent evidence available to investigators argument are presented in this section. YDOLG DQGZKHQFLUFXODUUHDVRQLQJ $SRWHQWLDOO\VLJQL¿FDQWWUHQGZDVREVHUYHG led to a negative outcome in an investigation in the percentage of subject responses to the YDOLG 7KHUHZDVDQHYHQFOHDUHU four passages examining circular arguments. downward trend in the number of subjects who There was a downward trend in the number judged circular argument to be not valid at all of subjects who found circular argument to as the evidence available to scientists lessened be a valid reasoning strategy as scientists and as investigations were shown to result in a increasingly overlooked evidence which positive outcome. The percentage of subjects was available to them. Subjects were most who judged circular argument to be not inclined to judge circular argument to be YDOLGDWDOOZDVDVIROORZV DEXQGDQW a valid reasoning strategy when it led to a HYLGHQFHDYDLODEOHWRLQYHVWLJDWRUV  SRVLWLYHRXWFRPHLQDVFLHQWL¿FLQYHVWLJDWLRQ QHJDWLYHRXWFRPHWRLQTXLU\  ODFNRI YDOLG DQGZHUHOHDVWLQFOLQHGWRMXGJH HYLGHQFHDYDLODEOHWRLQYHVWLJDWRUV DQG it as valid when there was an abundance SRVLWLYHRXWFRPHWRLQTXLU\  SUMMER 2014, VOL. 29, NO. 2 49

'DWDZHUHDQDO\]HGXVLQJWKH6WDWLVWLFDO in the discussion section. 3DFNDJHIRUWKH6RFLDO6FLHQFHV 6366IRU 7KHUHZDVDOVRDSRWHQWLDOO\VLJQL¿FDQW :LQGRZV9HUVLRQ :KHUHWKHUHZDV trend in the responses of subjects to the no expectation of a difference between two passages that examined analogical argument. YDULDEOHVWKLVZDVODUJHO\FRQ¿UPHGE\DODUJH Strong analogies were consistently rated as Pearson chi-square value. For example, it was more valid by subjects than weak analogies. predicted that subjects would judge the use of This was evident across both actual and non- circular argument to be equally unacceptable actual scenarios. In this way, 51.4% of subjects ZKHQDVFLHQWL¿FLQYHVWLJDWLRQKDGDQHJDWLYH rated a strong analogy as valid and only 10.3% outcome and when abundant evidence was rated a weak analogy as valid within an actual available to investigators but was overlooked scenario. Within a non-actual scenario, a by them. It was expected that subjects would strong analogy was rated as valid by 49.1% judge the occurrence of the argument under of subjects and a weak analogy was rated as both of these epistemic conditions to be valid by 5.2% of subjects. The highest ratings equally unacceptable, hence the original RI³QRWYDOLGDWDOO´ZHUHFRQVLVWHQWO\IRXQGIRU designation of these circular arguments as weak analogies across both actual and non- ³YLFLRXVFLUFOHV´7KLVZDVFRQ¿UPHGE\D actual scenarios. Weak analogies were rated large Pearson chi-square value of 31.756. This DVQRWYDOLGDWDOOE\RIVXEMHFWV QRQ value exceeds 0.05, indicating that there was DFWXDOVFHQDULR DQGRIVXEMHFWV DFWXDO QRVWDWLVWLFDOO\VLJQL¿FDQWGLIIHUHQFHLQWKH VFHQDULR ZKLOHRQO\RIVXEMHFWV QRQ and DFWXDOVFHQDULR DQGRIVXEMHFWV DFWXDO variables. VFHQDULR UDWHGVWURQJDQDORJLHVDVQRWYDOLGDW It was also expected that subjects all. There appeared to be negligible percentage would judge circular argument to be equally differences in subject ratings of validity across YDOLGZKHQDVFLHQWL¿FLQYHVWLJDWLRQUHVXOWHGLQ actual and non-actual scenarios. For example, a positive outcome and when there was a lack responses to the passages which examine RIHYLGHQFHDYDLODEOHWRLQYHVWLJDWRUVKHQFH strong analogies were as follows: actual the designation of circular argument under VFHQDULR YDOLGPRGHUDWHO\ ERWKHSLVWHPLFFRQGLWLRQVDV³YLUWXRXVFLUFOHV´ YDOLGQRWYDOLGDWDOO DQGQRQDFWXDO $JDLQWKLVZDVFRQ¿UPHGE\DFKLVTXDUH VFHQDULR YDOLGPRGHUDWHO\ value of 3.028, which exceeds 0.05, indicating YDOLGQRWYDOLGDWDOO  WKDWWKHUHZDVQRVLJQL¿FDQWGLIIHUHQFHLQKRZ As with circular argument, responses subjects rated circular arguments under these to the analogical argument passages underwent HSLVWHPLFFRQGLWLRQV+RZHYHURWKHUQRQ VWDWLVWLFDODQDO\VLV/DUJHDQGQRQVLJQL¿FDQW VLJQL¿FDQWFKLVTXDUHYDOXHVZHUHREWDLQHG FKLVTXDUHYDOXHVFRQ¿UPHGLQLWLDOLPSUHVVLRQV IRUSDVVDJHFRPSDULVRQVZKHUHDVLJQL¿FDQW of small and negligible percentage differences difference in subject responses might have in subject responses to certain passages. For been expected. They included comparisons example, based on the above percentage between the epistemic conditions and variables in the context of passages DEXQGDQWHYLGHQFH! FKLVTXDUH   that contained strong analogies. This was between and and ZDVLQGHHGQRVLJQL¿FDQWGLIIHUHQFHLQWKHVH FKLVTXDUH  $SRVVLEOHH[SODQDWLRQRI YDULDEOHV+RZHYHUZKHUHODUJHSHUFHQWDJH WKHVHQRQVLJQL¿FDQWUHVXOWVZLOOEHFRQVLGHUHG differences in subjects’ responses occurred and 50 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES

D¿QGLQJRIVLJQL¿FDQWGLIIHUHQFHPLJKWKDYH VXEMHFWVFRQVLVWHQWO\LGHQWL¿HGWKHVRXUFHRI been expected, this did not come about. For this weakness and rejected risk assessments example, 51.4% and 10.3% rated analogical and public health measures which were based reasoning as valid in passages that contained XSRQLW HJWKHXVHRIDQDQDORJ\ZLWKVFUDSLH strong and weak analogies, respectively. LQULVNDVVHVVPHQWVDERXW%6( 0RUHRYHU Yet, a Pearson chi-square value of 1.554 was these judgements were in evidence in actual obtained for these passages. An explanation and non-actual scenarios. This suggested RIWKHVHQRQVLJQL¿FDQWUHVXOWVLVJLYHQLQWKH that subjects’ judgements about public health discussion section. problems are the manifestation of a stable Statistical analysis of a subset of the UDWLRQDOFRPSHWHQFH6SHFL¿FDOO\WKH\DUH not data using the Pearson chi-square test indicated the consequence of the unique features that WKDWWKHUHZHUHQRVLJQL¿FDQWGLIIHUHQFHV attend any particular public health problem. in reasoning based upon the gender and %\WKHVDPHWRNHQVXEMHFWVLGHQWL¿HG HGXFDWLRQOHYHORIUHVSRQGHQWV 6HH7DEOH some circles in reasoning as valid, while other %LQWKH$SSHQGL[ 7KLVFRQ¿UPVLQSDUW circles were considered to be dubious value WKH¿QGLQJVRIDQHDUOLHUVWXG\ZKLFKIDLOHG and were accordingly rated as not valid at WR¿QGDQHIIHFWRIDJHDQGJHQGHURQWKH DOO9DOLGRU³YLUWXRXV´FLUFOHVLQUHDVRQLQJ W\SHRILQIRUPDWLRQSURFHVVLQJ V\VWHPDWLF were those which enabled scientists to YHUVXVKHXULVWLF XVHGE\VXEMHFWVGXULQJWKHLU achieve a positive outcome in inquiry, where assessment of epidemiological information VXFKDQRXWFRPHPD\EHWKHFRQ¿UPDWLRQ DERXWFDQFHUUDWHV 7UXPER *LYHQ of a hypothesis about the cause of an illness WKLVODFNRIVLJQL¿FDQWGLIIHUHQFHVDQG in a population. Also, circles in reasoning the fact that subject attributes such as age, were characterized as valid by subjects gender and education are not central to the when scientists lacked evidence which was questions posed by this study, it was decided independent of the conclusion-to-be-proved. not to conduct further statistical analysis of When either situation did not obtain, that is, the effects of gender and education level on when an inquiry either had a negative outcome subjects’ reasoning. or there was available evidence that was overlooked by scientists, subjects consistently IV. Discussion judged these circles in reasoning to be not YDOLGDWDOO WKH\ZHUH³YLFLRXV´FLUFOHV  This study has revealed a previously untested Furthermore, these judgements were rational competence on the part of lay not dependent on a particular public health people as they attempt to form judgements problem or scenario – scenarios as wide- DERXWSXEOLFKHDOWKSUREOHPV,QWKLV¿UVW ranging as the emergence of a new disease experimental investigation of the informal in a tribal population and the possible side fallacies, subjects were found to be adept at effects of a vaccination for pneumonia were recognizing the epistemic conditions under equally likely to give rise to these judgements. which circles and analogies in reasoning The consistency of these judgements across were more and less rationally warranted. different scenarios once again suggests that Subjects were able to identify when analogies subjects are drawing upon a stable rational DUHVXI¿FLHQWO\UREXVWWRSURYLGHVWURQJ competence in their reasoning. rational warrant for a particular public This study provides tentative evidence KHDOWKLQWHUYHQWLRQ HJWKHGLVVHPLQDWLRQ for the claim that subjects are able to use RISXEOLFKHDOWKLQIRUPDWLRQRQ+,9EDVHG analogical and circular arguments to bridge RQDQDQDORJ\ZLWKKHSDWLWLV% :KHUHDQ gaps in their knowledge during reasoning DQDORJ\ZDVZHDNRUÀDZHGLQVRPHUHVSHFW about complex public health problems. Both SUMMER 2014, VOL. 29, NO. 2 51 analogical and circular arguments functioned QRWLFHGWKHVLPLODULWLHVEHWZHHQ$,'6 for these subjects as quick rules of thumb or and hepatitis B and the population groups heuristics which guided judgement-making VXVFHSWLEOHWRERWKGLVHDVHV´ :KLWHIHPDOH XQGHUVSHFL¿FHSLVWHPLFFRQGLWLRQV7KHVH \HDUVROGVHFRQGDU\VFKRROHGXFDWHG conditions were adverse in some respect with ³WKH&'&ZRXOGVHHPWRKDYHPDGHD scientists attempting to assess public health VFLHQWL¿FDOO\EDVHGFRQQHFWLRQEHWZHHQ risks and institute public health measures in DQH[LVWLQJGLVHDVHDQGDQHZRQH´ :KLWH WKHDEVHQFHRIHYLGHQFHDQGNQRZOHGJH RI IHPDOH\HDUVROGXQLYHUVLW\HGXFDWHG DQHZSDWKRJHQIRUH[DPSOH 8QGHUWKHVH ³,WKLQNLWZDVDJRRGVWDUWLQJSRLQWDVWKH conditions, circles and analogies were found FKHPLFDOFRPSRXQGLQTXHVWLRQZDVLGHQWL¿HG to have a facilitative function in reasoning DVEHLQJVLPLODUWRWKHFKHPLFDOLGHQWL¿HGLQ through their capacity to bridge gaps in FDXVLQJWKHDLOPHQWVVWDWHG´ %ODFN&DUULEHDQ subjects’ knowledge. Rather than suspend IHPDOH\HDUVROGXQLYHUVLW\HGXFDWHG judgement in the absence of knowledge – an ³,WVHHPVORJLFDOWRXVHH[LVWLQJNQRZOHGJH unwise cognitive policy in the public health about another similar chemical to investigate domain where there are potentially serious whether the second chemical was the source consequences of inaction for human health – RIWKHKHDOWKSUREOHPVLQWKHDUHD´ :KLWH subjects put into effect reasoning strategies IHPDOH\HDUVROGXQLYHUVLW\HGXFDWHG  which helped them to arrive at judgement Subjects were equally adept at DERXWVRPHTXLWHFRPSOH[VFLHQWL¿FLVVXHV articulating their reasons for rejecting weak Under more favourable epistemic conditions, analogies. These reasons included comments it is likely that these strategies would be found WRWKHHIIHFWWKDW³UHODWHG´GLVHDVHVFRXOG wanting in a number of essential respects. VWLOODFWGLIIHUHQWO\$VVXFKLWZDV³QRW +RZHYHUZKHQFRQIURQWHGZLWKHSLVWHPLF FRQFOXVLYH´DQGZDVQRWD³VDIHDVVXPSWLRQ´ uncertainty, subjects are clearly willing to base a conclusion about the transmissibility to embrace these traditionally fallacious RIRQHGLVHDVH %6( WRKXPDQVRQWKHODFN arguments. RIWUDQVPLVVLELOLW\RIDQRWKHUGLVHDVH VFUDSLH  A qualitative analysis of the extended to humans. Indeed, to do so was characterized responses of subjects to the passages in the DV³ÀDZHGUHDVRQLQJ´$ODFNRIVLPLODULW\ questionnaire supports the claim that subjects EHWZHHQWZRGLVHDVHV DQGWZRFKHPLFDOV ZDV are sensitive to the epistemic conditions under FOHDUO\LGHQWL¿HGE\VXEMHFWVDVFDVWLQJGRXEW which analogies and circles in reasoning are on the rational force of the weak analogies in more and less rationally warranted. Strong the passages: analogies elicited comments to the effect that VFLHQWLVWVZHUHGUDZLQJ³VFLHQWL¿FDOO\EDVHG´ Weak analogies: connections between an existing and a new ³7KH\EDVHGDQRSLQLRQDERXWDµQHZ¶GLVHDVH GLVHDVHDQGWKDWLWZDV³ORJLFDO´RU³DJRRG on the action of a ‘different’ disease. This is starting point” to use existing knowledge in QRWFRQFOXVLYH´ :KLWHPDOH\HDUVROG drawing conclusions about a new public health VHFRQGDU\VFKRROHGXFDWHG problem. The rational force of analogical ³$WWKLVWLPHLWZDVQRWFRQFOXVLYHWKDW%6( reasoning was widely acknowledged by and scrapie were related diseases – it was only subjects to consist in the extent to which a suggestion, therefore the reasoning behind the chemicals and diseases described in the the suggestion that BSE would not transmit to passages were similar in nature: KXPDQVZDVÀDZHG´ :KLWHIHPDOH\HDUV ROGXQLYHUVLW\HGXFDWHG Strong analogies: ³LWGRHVQ¶WVHHPWRPHDVDIHDVVXPSWLRQ ³7KH&'&>&HQWHUVIRU'LVHDVH&RQWURO@ that because two diseases are ‘related’ they 52 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES will necessarily act in the same way as far as RIWKHVLWXDWLRQ ZDVDUHDVRQDEOHK\SRWKHVLV WUDQVPLVVLRQWRKXPDQVLVFRQFHUQHG´ :KLWH DQGWRWHVWLW´ :KLWHPDOH\HDUVROG PDOH\HDUVROGXQLYHUVLW\HGXFDWHG XQLYHUVLW\HGXFDWHG ³7KHGUXJVKDGGLIIHUHQWFKHPLFDO ³7KH\KDGQRWKLQJHOVHWRJRZLWK7KH\KDG compositions and therefore may not have to make the assumption to test the hypothesis” been similar enough to draw that conclusion” :KLWHPDOH\HDUVROGXQLYHUVLW\ :KLWHIHPDOH\HDUVROGXQLYHUVLW\ HGXFDWHG HGXFDWHG ³7KHDVVXPSWLRQFRXOGOHDGWRRWKHUVEHLQJ In relation to circular arguments, an monitored closely to prevent further deaths” equally rich array of expressions was used :KLWHIHPDOH\HDUVROGXQLYHUVLW\ by subjects to justify their judgements about HGXFDWHG the reasoning in the passages. Virtuous or Passages which contained vicious valid circles were described in terms of good circles elicited an equally wide-ranging set of VFLHQWL¿FPHWKRGVXFKDVZKHQWKHRXWFRPH evaluative comments from subjects. Vicious RIDQLQTXLU\SURYLGHGFRQ¿UPDWLRQRID circles were described as examples of bad VFLHQWL¿FK\SRWKHVLV9LUWXRXVFLUFOHVZHUH VFLHQFHRUSRRUVFLHQWL¿FPHWKRG6XEMHFWV DOVRFKDUDFWHUL]HGDVDW\SHRI³EHWWHUWKDQ stated that these circles were problematic QRWKLQJ´UHDVRQLQJZKHQVFLHQWLVWVKDG³QR because scientists overlooked evidence and did real option” other than to use conclusion- not consider other possible causes of diseases. dependent evidence in order to advance 6RPHVXEMHFWVLGHQWL¿HGWKHUHDVRQLQJDVD an investigation. In both cases, practical W\SHRI³FLUFXODULW\´WKHODWWHUWHUPXVHGLQ considerations such as the time taken to DVWULFWO\SHMRUDWLYHVHQVH ,WLVZRUWKQRWLQJ conduct investigations and the urgency of that no subject in the study used the term DSXEOLFKHDOWKSUREOHP KLJKOLJKWHGLQWKH ³FLUFXODULW\´LQDSRVLWLYHHSLVWHPLFVHQVH RFFXUUHQFHRIGHDWKVIRUH[DPSOH ZHUH The standard meaning of this term appears to seen to justify reasoning which, under other UHÀHFWLWVODUJHO\QHJDWLYHFKDUDFWHUL]DWLRQLQ circumstances, might not be judged in such WUDGLWLRQDOORJLFDOOLWHUDWXUH IDYRXUDEOHHSLVWHPLFWHUPV 1LFKRODV5HVFKHU has discussed faute de mieux  IRUODFNRI Vicious circles: DQ\WKLQJEHWWHU FRQVLGHUDWLRQVLQFRJQLWLRQ ³EOLQNHUHGDSSURDFK«WKHVHSHRSOHZHUHWU\LQJ DQGUHDVRQLQJ6HH5HVFKHU  IRUDQ to establish a cause, not exclude a particular DFFRXQWRISUHVXPSWLRQLQWKLVFRQWH[W RSWLRQ9HU\EDGVFLHQFH´ :KLWHPDOH \HDUVROGXQLYHUVLW\HGXFDWHG Virtuous circles: ³$VWKH\KDGSULRUDVVXPSWLRQVDERXWWKH ³$OOVFLHQFHLVEDVHGRQDQDVVXPSWLRQZKLFK FDXVHRIWKHGLVHDVHWKLVPD\KDYHLQÀXHQFHG is then used to determine testable hypothesis their methodology and also biased their which can be used to prove or disprove the UHVXOWVDVWKH\PD\KDYHXQNQRZLQJO\¿WWKH DVVXPSWLRQ´ :KLWHPDOH\HDUVROG data to their assumptions to get the results XQLYHUVLW\HGXFDWHG WKH\H[SHFWHG´ :KLWHIHPDOH\HDUVROG ³,W¶VWKHPRVWREYLRXVH[SODQDWLRQDQGOHQGV XQLYHUVLW\HGXFDWHG LWVHOIWRHDV\TXLFNYHUL¿FDWLRQ7LPHZDV ³$OOSRVVLEOHFDXVHVVKRXOGKDYHEHHQ REYLRXVO\RIWKHHVVHQFHLQWKLVFDVH´ :KLWH FRQVLGHUHG VD\JHQHWLFFKHPLFDO PDOH\HDUVROGXQLYHUVLW\HGXFDWHG FRQWDPLQDWLRQHWF ´ :KLWHPDOH\HDUV ³,QWKHVLWXDWLRQGHVFULEHG LHQRFOHDU ROGXQLYHUVLW\HGXFDWHG HYLGHQFH WKHVFLHQWLVWVKDYHQRUHDORSWLRQ ³LWPD\KDYHSUHYHQWHGKDPSHUHGDQDGHTXDWH RWKHUWKDQWRVHWXSZKDWWRWKHLUMXGJPHQW RQ VHDUFKIRUDQDOWHUQDWLYHFDXVH´ :KLWHPDOH the basis of their initial preliminary assessment \HDUVROGXQLYHUVLW\HGXFDWHG SUMMER 2014, VOL. 29, NO. 2 53

³$VVXPSWLRQOHDGVWRDFLUFXODULW\ĺLILW¶V must ultimately contribute to the development ZKDW\RXH[SHFWWR¿QGDQGGLVUHJDUGDOO of theoretical frameworks that can be used other sources of evidence, it will be what is to explain the heuristic reasoning processes IRXQG´ :KLWHIHPDOH\HDUVROGXQLYHUVLW\ at work in public health and other contexts. HGXFDWHG Certainly, the heuristic processing models $OWKRXJKWKHSHUFHQWDJH¿JXUHV WKDWLQIRUPHGWKHULVNDVVHVVPHQWVWXGLHV VHH obtained in this study suggest a number of 6HFWLRQ,, RI7UXPER  DQG-RKQVRQ clear trends in how subjects rated the analogies  VHHPSRZHUOHVVWRDFFRPPRGDWHWKH and circles in the passages, these were not VSHFL¿FHPSKDVLVRQLQIRUPDOIDOODFLHVDV VXSSRUWHGE\VLJQL¿FDQWFKLVTXDUHYDOXHV cognitive heuristics addressed in this paper. 7KHODFNRIVLJQL¿FDQWFKLVTXDUHYDOXHVLQ +RZHYHUUHFHQWZRUNE\'RXJODV:DOWRQ this study may be attributed to a couple of is at least suggestive of what a theoretical factors. Firstly, public health studies that account of fallacies-as-heuristics might look examine human judgements typically employ OLNH$FFRUGLQJWR:DOWRQ  PRVWRI a sample in excess of 1,000 subjects. This the informal fallacies are associated with an study obtained responses from 879 subjects. and a corresponding 'HVSLWHH[WHQVLYHHIIRUWVWRUHFUXLWPRUH parascheme. The argumentation scheme is part subjects, it was simply not possible to do so. RIDQHZHU LQHYROXWLRQDU\WHUPV FRJQLWLYH Many subjects agreed to participate in the system which operates in a controlled, study, but did go on to complete and return conscious and slow manner. This scheme asks the questionnaire. Some individuals reported critical questions of arguments, questions that this was on account of the amount of which are likely to expose logical weaknesses, time and degree of effort which was needed if such weaknesses exist. The parascheme is a to complete the exercises. Other public shorter version of the argumentation scheme. It health studies have tended to use shorter is part of an older cognitive system which uses questionnaires in which subjects merely tick fast and frugal heuristics to achieve solutions boxes or circle responses. Also, these studies to problems. Some of these heuristics involve RIWHQRIIHUVXEMHFWVD¿QDQFLDOLQGXFHPHQW jumping to conclusions, a cognitive strategy for their participation. This was simply not that can work well enough on some occasions possible within the resource constraints of the but results in errors on other occasions. current investigation. It is likely that with a Walton demonstrates this heuristic view of larger number of subjects, several chi-square the fallacies in relation to the argument from YDOXHVWKDWDSSURDFKHGVLJQL¿FDQFHZRXOG H[SHUWRSLQLRQ DUJXPHQWXP ad verecundiam  KDYHIDOOHQEHORZWKHVLJQL¿FDQFHOHYHO The parascheme of this argument omits Secondly, the test format required subjects assumptions, exceptions and one ordinary to indicate one response from the following premise that are integral to the corresponding SRVVLEOHUHVSRQVHV D YDOLG E PRGHUDWHO\ argumentation scheme. By neglecting these YDOLG F QRWYDOLGDWDOO,WLVSRVVLEOHWKDWWKH aspects, which confer a slow, deliberative use of three response categories weakened character on reasoning, an arguer can employ trends in the subjects’ responses in terms of the DIDVWKHXULVWLFWRWKHHIIHFW³LILW¶VDQH[SHUW statistical analysis. A simpler test format that RSLQLRQGHIHUWRLW´ :DOWRQS  PDGHXVHRIMXVWWZRUHVSRQVHFDWHJRULHV YDOLG :DOWRQ¶VDSSURDFKKDVFHUWDLQEHQH¿WV DQGQRWYDOLG PD\KDYHUHVXOWHGLQVWDWLVWLFDOO\ for the view of fallacies presented in this VLJQL¿FDQWUHVXOWV&HUWDLQO\WKHVHIDFWRUV paper. It achieves an important cognitive could be addressed in any further study in the reorientation of the informal fallacies. This is a area. sine qua non of an account of the fallacies that An experimental study of this type construes them as cognitive heuristics which 54 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES have a facilitative function in public health public health domain. Given that subjects are reasoning . Notwithstanding its cognitive reasonably adept at recognizing the conditions emphasis, Walton’s approach retains a strong under which circles and analogies in reasoning normative character such that informal are more and less rationally warranted, public fallacies are evaluated against the need to health practitioners should begin to exploit consider assumptions and exceptions which this valuable cognitive resource in their are typically bypassed by their corresponding communications and interventions with the KHXULVWLFV³7RMXGJHZKHWKHUDQDOOHJHG public. Public health communications with is fallacious the members of the public are frequently ignored, heuristic has to be examined in relation to rejected or misunderstood. This results whether other assumptions and exceptions in poor compliance with a large range of need to be taken into account that may be measures which are designed to protect human DFFHSWDEOHRUQRW´ :DOWRQS  health. Cases in point are the low uptake of Yet, these merits of Walton’s approach must YDFFLQDWLRQV HJPHDVOHVPXPSVUXEHOOD be considered alongside a couple of notable 005 LQÀXHQ]D DQGWKHZLGHVSUHDGQHJOHFW GLI¿FXOWLHV2QHRIWKHIDOODFLHVZHKDYH of measures which protect against the spread examined in this paper, circular argument- RILQIHFWLRXVGLVHDVHV HJVH[XDOO\WUDQVPLWWHG -also known as petitio principii or begging LQIHFWLRQV 7RWKHH[WHQWWKDWOD\SHRSOH the question--is one of several fallacies recognize the conditions under which circles ZKLFK:DOWRQ S DGPLWV³GRQRW and analogies in reasoning are more and less DSSHDUWR¿WVSHFL¿FDUJXPHQWDWLRQVFKHPHV rationally warranted and are prepared to put RUEHQH¿WGLUHFWO\IURPVFKHPHVZKHQLW this knowledge to use in making judgements comes to analyzing them.” In contrast, false about public health problems, it would seem analogy is amenable to analysis in terms of eminently reasonable for practitioners in the argumentation schemes. Also, on the account SXEOLFKHDOWK¿HOGWRIUDPHPHVVDJHVDQG of fallacies presented in this paper, heuristics advice with these cognitive strategies in mind. are not a matter of bypassing critical questions Public health practitioners should also ZKLFKDWWHQGDQDUJXPHQWDWLRQVFKHPH jOD be aware of one further and very important :DOWRQ EXWUDWKHUWKH\DUHPHQWDOVKRUWFXWV OHVVRQIRUWKHLU¿HOGIURPWKHFXUUHQWVWXG\ through a domain of expert knowledge which That lesson is that a lack of knowledge is not is beyond the cognitive grasp of the lay the same as a lack of reasoning. Certainly, lay person. If anything, the results of this study people do not have the knowledge of expert suggest that people in fact do address, rather scientists in the public health domain and other than bypass, critical questions when they VFLHQWL¿FDUHDV7KHUHLVQRGRXEWWKHUHIRUH engage in heuristic reasoning. It remains to be that they experience a lack of knowledge. But seen if Walton’s approach can accommodate if this study demonstrates one thing it is that GLI¿FXOWLHVRIWKLVW\SHDQGRQHRUWZRRWKHUV even lay people are capable of bringing to an GLVFXVVHGLQ&XPPLQJV D ,QWKH assessment of public health problems a rich PHDQWLPHLWLVD¿UPEDVLVXSRQZKLFKWR array of reasoning strategies which serve them build a theoretical account of a heuristic-based well in arriving at judgements about those approach to the fallacies. problems. It would be a serious mistake for Quite apart from the logical and public health practitioners to confuse a lack theoretical lessons of this investigation, there of knowledge on the part of lay people with a are also lessons for public health practitioners lack of reasoning and then proceed to construct from the current study. This study has shown public health communications and advice that the informal fallacies are capable of on this basis. The result would be a series of doing some important cognitive work in the interventions which fail to engage with the SUMMER 2014, VOL. 29, NO. 2 55 rational cognitive strategies of the very people %LUR-,  .QRZDELOLW\EHOLHYDELOLW\ – members of the public – whose compliance and : a reply to is most needed in order to undertake effective Sanford. Metaphilosophy,    public health work. So, in conclusion, it 247. can be said that public health practitioners %URZQ:5  7ZRWUDGLWLRQVRI should certainly exploit the informal fallacies analogy. Informal Logic,    examined in this study. But a policy of actively %URZQH.  6QRZEDOOVDPSOLQJ circumventing these arguments is guaranteed Using social networks to research non- to lead only to poor compliance with protective heterosexual women. International health measures on the part of the public. Journal of Social Research Methodology,    V. Summary &XPPLQJV/  3HWLWLRSULQFLSLL7KH case for non-fallaciousness. Informal 7KLVVWXG\LVSDUWRIWKH¿UVW Logic,    experimental investigation of a group of &XPPLQJV/  5HDVRQLQJXQGHU arguments called the informal fallacies in uncertainty: The role of two informal public health reasoning. These arguments IDOODFLHVLQDQHPHUJLQJVFLHQWL¿FLQTXLU\ have traditionally been examined by logicians Informal Logic    and philosophers whose principal concern &XPPLQJV/  $QDORJLFDOUHDVRQLQJDV has been to describe their logical and a tool of epidemiological investigation. epistemic features. The ability of lay people Argumentation    to recognize these features and employ these &XPPLQJV/  *LYLQJVFLHQFHDEDG arguments as heuristics during reasoning name: Politically and commercially about public health issues has not previously motivated fallacies in BSE inquiry. been considered by investigators. The study Argumentation    provides tentative evidence that lay subjects &XPPLQJV/  (PHUJLQJLQIHFWLRXV are adept at recognizing the conditions under diseases: Coping with uncertainty. which circular and analogical arguments are Argumentation    more and less rationally warranted. Moreover, &XPPLQJV/ D 3UDJPDGLDOHFWLFV they are capable of exercising this rational ,Q/&XPPLQJV HG  The Routledge competence in order to arrive at judgements pragmatics encyclopedia . London: about complex public health problems. This Routledge, 337-338. ¿QGLQJKDVLPSOLFDWLRQVIRUUHFHQWWKHRUHWLFDO &XPPLQJV/ E  Rethinking the BSE models of the informal fallacies as cognitive &ULVLV$VWXG\RIVFLHQWL¿FUHDVRQLQJ KHXULVWLFV+RZHYHUEH\RQGWKHORJLFDODQG under uncertainty. 'RUGUHFKW6SULQJHU theoretical implications of the current study, &XPPLQJV/  &RQVLGHULQJULVN there are also implications for public health assessment up close: The case of bovine practitioners. To the extent that members spongiform encephalopathy. Health, Risk of the public possess this particular rational & Society,    competence and can apply it to public health &XPPLQJV/ D 6FDULQJWKHSXEOLF problems, practitioners in the public health Fear appeal arguments in public health ¿HOGLWLVDUJXHGVKRXOGVHHNWRH[SORLWWKLV reasoning. Informal Logic    competence in their communications with the &XPPLQJV/ E 7KHFRQWULEXWLRQ public. of informal logic to public health. Perspectives in Public Health    References 66-67. &XPPLQJV/ F 7KHSXEOLFKHDOWK 56 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES

scientist as informal logician. +DPSOH'  'XDOFRGLQJUHDVRQLQJ International Journal of Public Health, and fallacies. Journal of the American    Forensic Association    &XPPLQJV/ D 3XEOLFKHDOWK +DPSOH'  $WKLUGSHUVSHFWLYHRQ reasoning: Much more than deduction. argument. Philosophy and Rhetoric , 18 Archives of Public Health      &XPPLQJV/ E &LUFXODUUHDVRQLQJLQ +DPSOH'  $UJXPHQWSXEOLFDQG public health. Cogency    private, social and cognitive. Journal of &XPPLQJV/ D 3XEOLFKHDOWK the American Forensic Association , 25 reasoning: A logical view of trust.   Cogency    -RKQVRQ%%  7HVWLQJDQGH[SDQGLQJ &XPPLQJV/ E ,QIRUPDOIDOODFLHV a model of cognitive processing of risk as cognitive heuristics in public health information. Risk Analysis    reasoning. Informal Logic    650. &XPPLQJV/ F 7KHµWUXVW¶KHXULVWLF -RKQVRQ/DLUG31DQG%\UQH50- Arguments from authority in public  Deduction +RYH8./DZUHQFH health. Health Communication    Erlbaum Associates. 1043-1056. .DKQHPDQ'   Thinking, fast and slow. &XPPLQJV/ G &RSLQJZLWK New York: Farrar, Straus, and Giroux. uncertainty in public health: The use of .ODKU'   Exploring science: The heuristics. Public Health    cognition and development of discovery 394. processes . Cambridge, MA: The MIT &XPPLQJV/ H $QDORJLFDOUHDVRQLQJLQ Press. public health. Journal of Argumentation 0LOO-6   A system of logic in Context    ratiocinative and inductive HG-0 &XPPLQJV/ D  Reasoning and Robson, in The collected works of public health: New ways of coping with , vols. 7, 8. Toronto: uncertainty. 'RUGUHFKW6SULQJHU University of Toronto Press. &XPPLQJV/ E 7KHXVHRIµQR 1DWLRQDO6FLHQFH%RDUG   Science and evidence’ statements in public health. Engineering Indicators 2010 . Arlington, Informal Logic    VA: National Science Foundation. )LQRFFKLDUR0$  7ZRHPSLULFDO 1R\&  6DPSOLQJNQRZOHGJH7KH approaches to the study of reasoning. hermeneutics of snowball sampling Informal Logic    in qualitative research. International *DEED\'0DQG:RRGV- WRDSSHDU  Journal of Social Research Methodology , Seductions and shortcuts: Fallacies in    the cognitive economy. Elsevier: North 3HUHOPDQ&K   The realm of rhetoric +ROODQG WUDQVE\:LOOLDP.OXEDFN 1RWUH *OLN'&  5LVNFRPPXQLFDWLRQIRU 'DPH,18QLYHUVLW\RI1RWUH'DPH public health emergencies. Annual Press. Review of Public Health , 28, 33-54. 3HUNLQV'1  5HDVRQLQJDVLWLVDQGDV Guarini, M., Butchart, A., Smith, P.S. and it could be: An empirical perspective. 0ROGRYDQ$  5HVRXUFHVIRU ,Q'17RSSLQJ'&&URZHOODQG research on analogy: A multi-disciplinary 91.RED\DVKL HGV Thinking across guide. Informal Logic    cultures: The third international +DPEOLQ&/   Fallacies . London: conference on thinking. +LOOVGDOH1- Methuen. Lawrence Erlbaum Associates, 175-194. SUMMER 2014, VOL. 29, NO. 2 57

3HUNLQV'1$OOHQ5DQG+DIQHU-   5  A systematic theory of 'LI¿FXOWLHVLQHYHU\GD\UHDVRQLQJ,Q: argumentation: The pragma-dialectical 0D[ZHOO HG  Thinking: The expanding approach. Cambridge: Cambridge frontier. Philadelphia: The Franklin University Press. Institute Press, 177-189. :DOWRQ'1  $UHFLUFXODUDUJXPHQWV 5HVFKHU1   Dialectics: A controversy- necessarily vicious? American oriented approach to the theory Philosophical Quarterly    of knowledge. Albany, NY: State 274. University of New York Press. :DOWRQ'1  1RQIDOODFLRXVDUJXPHQWV 5HVFKHU1  3ODXVLELOLW\DQG from ignorance. American Philosophical SUHVXPSWLRQ,Q'-DFTXHWWH HG  Quarterly    Reason, method, and value: A reader :DOWRQ'1   Informal logic: A on the philosophy of Nicholas Rescher . pragmatic approach. New York: +HXVHQVWDPP2QWRV9HUODJ Cambridge University Press. 6LPSVRQ&'HDOOHU6DQG3DWWHUVRQ: :DOWRQ'1  :K\IDOODFLHVDSSHDU  6WXG\VKRZVQHHGIRUEHWWHU to be better arguments than they are. communication between government Informal Logic    and doctors. British Medical Journal , Wilson, C., Evans, G., Leppard, P. and Syrette, 313, 1146-1147. -  5HDFWLRQVWRJHQHWLFDOO\ 7UXPER&:  ,QIRUPDWLRQSURFHVVLQJ PRGL¿HGIRRGFURSVDQGKRZSHUFHSWLRQ and risk perception: An adaptation of RIULVNVDQGEHQH¿WVLQÀXHQFHV the heuristic-systematic model. Journal consumers’ information gathering. Risk of Communication    Analysis    7YHUVN\$DQG.DKQHPDQ'   :RRGV-  $SSHDOWRIRUFH,Q+9 -XGJHPHQWXQGHUXQFHUWDLQW\+HXULVWLFV +DQVHQ 5&3LQWR HGV  Fallacies: and biases. Science    Classical and contemporary readings . 1131. Pennsylvania: The Pennsylvania State 7YHUVN\$DQG.DKQHPDQ'   Belief University Press, 240-250. LQWKHODZRIVPDOOQXPEHUV,Q(6KD¿U :RRGV-   The death of argument: (ed.) Preference, belief and similarity: Fallacies in agent-based reasoning. Selected writings by Amos Tversky. Dordrecht : Kluwer Academic Cambridge, MA: The MIT Press, 193- Publishers. 202. :RRGV-DQG:DOWRQ'  3HWLWLR YDQ(HPHUHQ)+DQG*URRWHQGRUVW5 principia. Synthese     Speech acts in argumentative discussions: A theoretical model for the Acknowledgements analysis of discussions directed towards VROYLQJFRQÀLFWVRIRSLQLRQ 'RUGUHFKW The author acknowledges with gratitude +ROODQG)RULV3XEOLFDWLRQV the assistance of the following people and YDQ(HPHUHQ)+DQG*URRWHQGRUVW5 organizations in the recruitment of subjects  7KHSUDJPDGLDOHFWLFDODSSURDFK IRUWKLVVWXG\3HWHU+RPD &KLHI([HFXWLYH WRIDOODFLHV,Q+9+DQVHQ 5& 1RWWLQJKDP8QLYHUVLW\+RVSLWDOV1+67UXVW  3LQWR HGV  Fallacies: classical and 6XVDQ-DPHV &KLHI([HFXWLYH7KH5R\DO contemporary readings . University 'HUE\+RVSLWDO 0DUN/HZLV 0DQDJLQJ Park, Pennsylvania: Pennsylvania State 'LUHFWRU-RKQ/HZLV1RWWLQJKDP -XOLH University Press, 130-144. %RZOH\ *HQHUDO0DQDJHU5RNRKHDOWK YDQ(HPHUHQ)+DQG*URRWHQGRUVW FOXE1RWWLQJKDP 6LPRQ6NHOVRQ *HQHUDO 58 INQUIRY: CRITICAL THINKING ACROSS THE DISCIPLINES

0DQDJHU'DYLG/OR\GKHDOWKFOXE1RWWLQJKDP 7KHDVVLVWDQFHRI'U.HYLQ3HUUHWW &RQVXOWDQW LQ3XEOLF+HDOWK0HGLFLQH0DQFKHVWHU1+6 DQG'U9LQRG7RKDQL IRUPHUO\&RQVXOWDQWLQ &RPPXQLFDEOH'LVHDVH&RQWURODQG3XEOLF+HDOWK0HGLFLQH6RXWKHUQ+HDOWKDQG6RFLDO 6HUYLFHV%RDUG1RUWKHUQ,UHODQG LVDOVRJUDWHIXOO\DFNQRZOHGJHG6SHFLDOWKDQNVJRWR*DYLQ %URRNHV 6385VWXGHQW1RWWLQJKDP7UHQW8QLYHUVLW\ ZKRSOD\HGDNH\UROHLQWKHIRUPDO recruitment activities undertaken as part of this study. The comments of an anonymous reviewer of this paper were particularly insightful and are gratefully acknowledged. Professor Frank Fair 0DQDJLQJ(GLWRU INQUIRY XQGHUWRRNH[WHQVLYHHGLWRULDOZRUNRQWKLVSDSHU,DPPRVWJUDWHIXO to him for his contribution.

Author Information

3URIHVVRU/RXLVH&XPPLQJVLVLQWKH6FKRRORI$UWVDQG+XPDQLWLHV1RWWLQJKDP7UHQW University, Clifton Campus, Clifton Lane, Nottingham, NG11 8NS, England, UK. +HUHPDLODGGUHVVLV [email protected] .

Appendix

TABLE A: Subject characteristics

SUBJECT CHARACTERISTICS (total = 879 subjects)

Average: 43.8 years AGE Range: 18-65 years

Male: 292 subjects GENDER Female: 587 subjects

University level: 589 subjects EDUCATION Secondary school level: 290 subjects

SUMMER 2014, VOL. 29, NO. 2 59

TABLE B: Argument type: sex and education level (university/secondary)

Sex Education ARGUMENT TYPE Response M/F Uni/Sec Analogical argument: Valid: 49.4/52.2% 49.2/56.0% Strong analogy Moderately valid: 43.4/40.5% 43.7/36.3% Actual scenario Not valid at all: 7.2/7.3% 7.1/7.7%

Analogical argument: Valid: 40.4/49.7% 44.2/49.0% Strong analogy Moderately valid: 48.2/42.3% 45.8/42.9% Non-actual scenario Not valid at all: 11.4/8.0% 10.0/8.1%

Analogical argument: Valid: 7.8/12.0% 6.8/17.2% Weak analogy Moderately valid: 36.5/38.6% 40.6/32.3% Actual scenario Not valid at all: 55.7/49.4% 52.6/50.5%

Analogical argument: Valid: 8.6/3.6% 4.1/7.4% Weak analogy Moderately valid: 19.4/20.8% 21.0/18.9% Non-actual scenario Not valid at all: 72.0/75.6% 74.9/73.7%

Question-begging argument: Valid: 61.0/56.7% 59.4/54.6% Virtuous circle Moderately valid: 25.6/31.5% 27.4/35.2% Positive outcome Not valid at all: 13.4/11.8% 13.2/10.2%

Question-begging argument: Valid: 33.9/25.9% 30.0/27.4% Virtuous circle Moderately valid: 43.5/42.9% 44.7/40.0% Lack of evidence Not valid at all: 22.6/31.2% 25.3/32.6%

Question-begging argument: Valid: 39.1/27.4% 34.9/23.5% Vicious circle Moderately valid: 21.8/27.4% 26.2/24.5% Negative outcome Not valid at all: 39.1/45.2% 38.9/52.0%

Question-begging argument: Valid: 10.9/10.5% 12.3/7.2% Vicious circle Moderately valid: 29.3/29.5% 27.2/34.0% Abundant evidence Not valid at all: 59.8/60.0% 60.5/58.8%