Western Michigan University ScholarWorks at WMU

Dissertations Graduate College

1-2011

Evaluation of Verbal Behavior in Older Adults

Amy Gross Western Michigan University

Follow this and additional works at: https://scholarworks.wmich.edu/dissertations

Part of the Communication Sciences and Disorders Commons, Geriatrics Commons, and the Gerontology Commons

Recommended Citation Gross, Amy, "Evaluation of Verbal Behavior in Older Adults" (2011). Dissertations. 412. https://scholarworks.wmich.edu/dissertations/412

This Dissertation-Open Access is brought to you for free and open access by the Graduate College at ScholarWorks at WMU. It has been accepted for inclusion in Dissertations by an authorized administrator of ScholarWorks at WMU. For more information, please contact [email protected]. EVALUATION OF VERBAL BEHAVIOR IN OLDER ADULTS

by

Amy Gross

A Dissertation Submitted to the Faculty of The Graduate College in partial fulfillment of the requirements for the Degree of Doctor of Philosophy Psychology, Behavior Analysis Advisor: R. Wayne Fuqua, Ph.D.

Western Michigan University Kalamazoo, Michigan June 2010 EVALUATION OF VERBAL BEHAVIOR IN OLDER ADULTS

Amy Gross, Ph.D.

Western Michigan University, 2010

Older adults make up a large and increasing portion of the population (U.S.

Census Bureau, 2008). Approximately 5% of older adults have a dementia diagnosis, and language deterioration is a common symptom associated with this disorder

(Kempler, 2005). In order to assess and treat language deficits in older adults, accurate and sensitive measures of verbal skills are needed. Language has traditionally been considered an existing entity that resides inside individuals, but the entity of language is unobservable and difficult to study or manipulate. Skinner

(1957/1992) proposed that language was simply operant behavior. He classified a number of verbal operants based on their function, describing the antecedents and consequences that control various forms of verbal behavior. A conceptualization based on the function of verbal behavior allows for altering the occurrence of verbal responses by manipulating the controlling conditions. To the author's knowledge, there are no empirical studies evaluating verbal behavior in older adults using

Skinner's analysis of functionally independent verbal operants. The purpose of this study was to develop assessments using Skinner's functional verbal operants and apply those assessments to the evaluation of verbal behavior in older adults with and without dementia. The research addressed two questions. First, in what way do verbal behavior problems differ between older adults with and without dementia? Second, does language deteriorate in a pattern compatible with Skinner's analysis of functionally independent verbal operants? Thirty-one participants were categorized into Control (n=15) and Dementia (n=16) Groups based on their score on the

Dementia Rating Scale-2. Verbal behavior assessments were administered to participants on two occasions, separated by one week. Results revealed individuals with dementia performed significantly more poorly on the Tact Assessment than those without dementia; however, no significant differences were found on the other assessments. These findings suggest tacts are sensitive to language decline associated with dementia. Results also indicated participants from both groups performed better on measures of tacts than on measures of intraverbals or mands, even though topographically identical verbal responses were required across these assessments.

Differential performance across these assessments provides support for Skinner's conceptualization of functionally independent verbal operants. UMI Number: 3470404

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. UMT Dissertation Publishing

UMI 3470404 Copyright 2010 by ProQuest LLC. All rights reserved. This edition of the work is protected against unauthorized copying under Title 17, United States Code.

ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 Copyright by Amy Gross 2010 ACKNOWLEDGMENTS

I would first like to thank my advisor, Dr. Wayne Fuqua, for his guidance and support through this and many other projects during my graduate career. I also want to express gratitude for the time and assistance provided by my committee members: Dr. Cynthia Pietras, Dr. Amy Naugle, and Dr. Raymond Miltenberger. Many thanks to my research assistants: Todd Merritt, Lauren Flannery, and Erica Kasemodel. Their dedication and hours of work made this project a success. Thank you to Dr. Alan Lewandowski for his generosity with both time and clinical resources. I would also like to thank my family and friends for making me who I am today. Thank you to my parents, Roger and Cathy Mackner, for always believing in me and for providing unending guidance and support. Finally, I would like to express my gratitude to my husband, Samuel Gross, for showing tremendous patience and continuously encouraging me to pursue my goals and dreams.

Amy Gross

u TABLE OF CONTENTS

ACKNOWLEDGMENTS ii

LIST OF TABLES vi

LIST OF FIGURES vii

INTRODUCTION 1

Language in Older Adults 2

Conceptual Models of Language Decline 3

Neurological 3

Cognitive 4

Linguistic 5

Behavior Analysis 6

Assessment of Language Deficits in Older Adults 11

Purpose 13

METHOD 14

Participants and Setting 14

Measures 14

Demographic Questionnaire 14

Dementia Rating Scale-2 (DRS-2) 15

Boston Naming Test (BNT) 15

Verbal Behavior Assessment Battery (VBAB) 16

Selection Assessment 16

Echoic Assessment 16 iii Table of Contents—continued

Tact Assessment 17

Intraverbal Assessment 17

Mand Assessment 18

3D Mand Assessment 18

Procedure 21

Session One 21

Session Two 21

Target Behaviors, Data Collection, and Data Analysis 21

Interobserver Agreement 22

Procedural Integrity 23

RESULTS 26

Between Groups Differences 29

Within Subject Differences and Non-Statistical Error Analysis 30

Regression Analyses 35

Test-Retest Reliability 37

Correlations with Short-Form BNT 38

DISCUSSION 39

Group Differences 40

Functional Independence 45

Limitations 49

Future Research 51

Conclusion 53

IV Table of Contents—continued

REFERENCES 55

APPENDICES

A. Demographic Questionnaire 59

B. Selection Assessment Form 61

C. Echoic Assessment Form 63

D. Tact Assessment Form 65

E. Intraverbal Assessment Form 67

F. Mand Assessment Form 70

G. 3D Mand Assessment Form 72

H. Verbal Behavior Assessment Raw Scores During Session One and Session Two 73 I. HSIRB Approval Letter 74

v LIST OF TABLES

1. Summary of Test Instructions 20

2. Interobserver Agreement Data 24

3. Procedural Integrity Data 25

4. Participants' Level of Impairment 26

5. Assessment Score Means, Standard Deviations, and Ranges 28

6. Number of Participants with Consistent Errors Across Tact, Intraverbal, and Mand Assessments 33 7. Number of Participants with Inconsistent Errors Across Tact, Intraverbal, and Mand Assessments 34

8. Regression Analysis Summary for Total Score 36

9. Regression Analysis Summary for Intraverbal 36

10. Regression Analysis Summary for Mand 37

VI LIST OF FIGURES

1. Verbal Behavior Assessment Mean Correct 28

2. Between Groups Mean Rank Differences 30

3. Within-Subjects Mean Rank Differences 31

vn 1

INTRODUCTION

Based on results from the 2000 census, adults 65 years old and older made up 12.4% of the population (U.S. Census Bureau, 2001). Projections indicate that this age group will grow in the coming years, reaching 17.8% of the population in 2025 and 20.1% in 2050. Baby Boomers will begin entering older adulthood in 2011 and will continue to enter this age category until 2029, contributing greatly to the increase (U.S. Census Bureau, 2008). With growing numbers of older adults, the number of individuals with age-related disabilities also increases. In 2000,43% of women and 40.4% of men who were 65 years of age and older indicated some type of disability, as compared to 17.6% and 19.6% of the 16-64 year old women and men, respectively. Of the disability types, 10.8% of the older adults endorsed a mental disability, defined as difficulty with learning, remembering, or concentrating, whereas only 3.8% of working adults indicated a mental disability (U.S. Census Bureau, 2003).

Approximately 5% of adults over 65 years old suffer from some form of dementia (Kempler, 2005). Dementia is a condition characterized by memory loss and at least one other area of cognitive dysfunction including aphasia, apraxia, agnosia, or disturbed executive functioning. (American Psychiatric Association [DSM-IV-TR], 2000). There are several subtypes of dementia (e.g., Dementia of the Alzheimer's Type, Vascular Dementia), each defined by specific courses or neurological findings. Estimates indicate that the number of adults with dementia diagnoses worldwide will grow from 25 million in 2000 to 63 million in 2030 (Kempler, 2005). With the increasing numbers of older adults diagnosed with dementia, developing appropriate assessment and treatment methods is of critical importance. 2 Language in Older Adults

Aphasia, one of the cognitive dysfunctions included in the dementia diagnosis, is a "deterioration of language function (DSM-IV-TR, 2000, p. 148)." Aphasia often occurs after a specific, focal brain injury. However, the term aphasia also describes language difficulties that appear as part of another disorder or syndrome, for example, dementia (Kempler, 2005). The most common language complaint associated with dementia is word finding difficulty (Kertesz, 1994). Word finding difficulty is a broad category that can refer to a number of problems. For example, complaining of word finding difficulties might mean that the individual cannot name viewed objects or items, is unable to provide names of known individuals, has difficulties with spontaneous speech, or has other problems that interfere with producing the appropriate word in a specific context (Rohrer et al., 2008). Notably, word finding deficits may be inconsistent across different contexts. That is, a person may be able to name a photo of a spoon and read the printed word spoon, yet have difficulty finishing the phrase, "you eat with a knife, fork and ." In general, describing a language deficit as a "word finding difficulty" lacks the specificity needed to adequately characterize the problem, to determine the cause of the problem, or to inform an intervention or remedial training method (Rohrer et al., 2008). Other language complaints associated with dementia include word and sentence organization difficulties; problems comprehending the speech of others; repetitive vocalizations; reading or writing problems; or a lack of content in speech (Bourgeois & Hickey, 2009). Many of these language problems suffer from the same weakness associated with labeling a problem as a "word finding difficulty," that is, they lack specificity.

Adults without dementia often experience language changes as they age, thus, changes in language abilities are not unique to those diagnosed with dementia (Benson & 3

Ardila, 1996; Kempler, 2005). Researchers have found that confrontation naming (i.e., naming a presented item), verbal fluency, and sequential language comprehension decrease with age, even in those without a diagnosis of dementia (Benson & Ardila, 1996; Kempler, 2005). However, older adults' ability on verbal tasks is highly variable, and some do not experience any deficits in language-related skills (Kempler, 2005). Although many older adults do experience changes in language-related skills, researchers find that individuals with dementia diagnoses perform more poorly on verbal tasks than their unimpaired peers (Benson & Ardila, 1996). Because of the deterioration in speech and language of older adults with and without dementia, it is important to determine the degree to which speech and language deficits of individuals diagnosed with dementia are worse than those which are typical of normal aging. Moreover, language assessments that sample a range of behavior under several contexts need to be developed in order to: (a) evaluate functional language skills of older adults, (b) more accurately specify language problems, and (c) develop interventions that are specific to particular language deficits.

Conceptual Models of Language Decline

Neurological

Neurological explanations attempt to describe language difficulties by finding brain lesions, or damaged brain tissue (Davis, 2000). Brain lesions may be caused by specific incidents (e.g., stroke), or they can occur from general atrophy, or shrinking of brain tissue (Davis, 2000). Aphasia and dementia diagnoses can be related to focal lesions (i.e., localized), multifocal lesion patterns (i.e., multiple lesions), or diffuse brain damage (i.e., damage throughout). Examples of diagnoses related to focal lesions include Broca's and Wernicke's aphasias. Broca's aphasia is a syndrome in which there is damage to the frontal region of the left cerebral cortex, whereas Wernicke's aphasia is 4 related to lesions in the left temporal cortex (Davis, 2000). An example of a diagnosis related to multifocal lesions or diffuse damage is multiple infarct dementia, which is associated with damage in any location in the brain (Davis, 2000). The primary criticism of the neurological approach is that many syndromes or symptom patterns are not consistently correlated with independently verified brain damage (Rohrer et al., 2008). Notably, there are syndromes that have high clinico- anatomical correlation; however, in many cases, brain scans reveal ambiguous or negative findings (Rohrer et al., 2008). Moreover, many clinicians (e.g., psychologists, speech and language therapists) do not have access to brain scan technology. Thus, basing a diagnosis or treatment plan on neurological damage may prove to be difficult. Furthermore, identifying the locus of brain damage associated with clinical symptoms does not necessarily yield well-validated treatments to remedy the language-related difficulties.

Cognitive

According to Davis (2000), "Cognitive theories of language deficits are based on two primary features: knowledge and process. Knowledge is a relatively stable storage of information (p. 13)," and process is the "mind's activity (p. 13)." According to this model, a stimulus in the environment is transformed by one's brain and a mental representation of the stimulus is made. The mental representation is either stored temporarily or permanently in memory. There are several types of memory: long-term- memory, episodic memory, semantic memory, procedural memory, lexical memory, and working memory. Each type of memory is responsible for different processes or for storage of different mental representations. For example, long-term memory is the area in which knowledge is stored, whereas working memory is largely responsible for the 5 processing of information. Cognitive explanations state that language difficulties give evidence for the different types of memory, because individuals with language deficits experience language-related strengths and weaknesses, often associated with differing memory systems (Davis, 2000). For example, an individual may be able to tell about events that happened many years prior using long-term memory, but unable to process and react to new information appropriately using working memory. In terms of aphasia, cognitive theorists would assert that the language problems are a processing issue, not a loss of stored knowledge (Davis, 2000).

Cognitive theorists attribute all language functions to memory, including language acquisition and decline. Davis (2006) noted that memory is difficult to explain functionally; therefore, metaphors of mental images as photographs are used to describe the process. Not being able to provide a functional description is a weakness of the cognitive theory because these theorists must depend on hypothetical constructs as explanations for behavior. Basing a theory on hypothetical constructs is problematic because these constructs cannot be manipulated in meaningful ways to evaluate their existence or nonexistence (Skinner, 1953/1965). Without a functional explanation of language, it is difficult to adequately assess language skills and deficits or develop intervention strategies to address language-related difficulties.

Linguistic

According to a linguistic analysis, aphasia can affect any aspect of language, including features, phonemes, morphemes, syntagms, or semantics (Benson & Ardila, 1996). These aspects range in complexity from simple muscular movement, to the meanings of words. Linguists note that aphasia can affect any level of language, and some of these deficits are characteristic of particular types of aphasia. However, it is 6 often the case that the relationship between linguistic impairments and specific diagnostic categories is complex. That is, individuals with a specific diagnosis may experience multiple linguistic deficits. Furthermore, the same linguistic deficit may be present in several diagnostic categories (Benson & Ardila, 1996). Another curious aspect of the linguistic account is that linguistic deficits are not always consistent across contexts. That is, an individual may be able to give an appropriate response when requesting an item, but then fail to produce the very same response (i.e., the particular word) when asked to name a picture of the same item. Therefore, this response is at strength under certain conditions, but not others, possibly indicating the deficit is not related to specific linguistic abilities (Baker, LeBlanc, & Raetz, 2008). Because of overlapping symptoms and indistinguishable diagnostic categories, assessment and diagnosis based on linguistic deficits may have significant limitations, and therefore, may not inform specific treatment approaches.

Behavior Analysis

Behavior analysis offers a different conceptual approach to language acquisition and decline. Rather than focusing on underlying neurological causes of language problems or hypothetical entities that are claimed to account for language acquisition and decline, behavior analysis views "language" as simply operant behavior. Skinner noted the difference between language and verbal behavior by describing language, as it is traditionally characterized, as an entity that individuals acquire, but verbal behavior as vocal and non-vocal behavior that is controlled by its effect on listeners (Skinner, 1974). He asserted that analyzing verbal behavior the same way as any other behavior was more beneficial than attempting to study the entity of language (Skinner, 1974), much of which is attributed to inaccessible mechanisms that are impossible to study via direct 7 observation (e.g., various memory systems). Skinner (1957/1992) provided a classification system of verbal operants, each of which is controlled by a unique set of antecedents and consequences. A verbal operant is a type of verbal behavior, as opposed to a specific instance of verbal behavior (Skinner, 1957/1992, p. 20). There are certain controlling variables for each verbal operant (Skinner, 1957/1992). The following is a review of a subset of Skinner's verbal operants: echoics, tacts, intraverbals and mands. An echoic is a verbal operant in which the response produced is similar to a preceding vocal verbal stimulus. The consequence for an echoic is generalized reinforcement by listeners (Skinner, 1957/1992). Generalized reinforcers are those that have been paired with several primary reinforcers. Therefore, they are effective reinforcers across various conditions because at least one state of deprivation is likely to be present at any given time (Skinner, 1953/1965). An example of an echoic is as follows: the caretaker says, "say ball," and the child says, "ball," after which the caretaker provides praise.

Tacts are responses that are "evoked (or at least strengthened) by a particular object or event or property of an object or event (Skinner, 1957/1992, p. 82)." The consequence for a tact is generalized reinforcement from the listeners. For example, a caretaker says, "name this" while holding up a ball. The child says, "ball" and then the caretaker provides praise. An intraverbal is a verbal operant that is occasioned by a verbal stimulus that has no point-to-point correspondence or formal similarity with the response (Skinner, 1957/1992). That is, the preceding verbal stimulus does not match the response in any way. Intraverbals are reinforced by generalized reinforcement produced by the listener (e.g., the reinforcing effects associated with correct responding). Examples of intraverbal 8 responses include answering questions, engaging in word association tasks, or filling in the blank. Finally, mands are controlled by an antecedent condition of an establishing operation, and the reinforcer is a "characteristic consequence (Skinner, 1957/1992, p. 35)" related to the establishing operation (Michael, 1988). For example, when one is in a state of water deprivation and asks for a glass of water, the consequence is receiving a glass of water. Unlike other forms of verbal behavior, the listener does not typically provide generalized reinforcement (e.g., social praise) as the reinforcing consequence. Rather, the listener provides a characteristic consequence (Michael, 1988; Skinner, 1957/1992).

Skinner (1957/1992) proposed that each verbal operant is functionally independent, meaning that a response may occur under certain controlling conditions, but the topographically identical response may not occur under other controlling conditions. That is, an individual may be able to tact "ball" but not mand for a ball. Studies evaluating Skinner's analysis of verbal behavior have been able to demonstrate the functional independence of verbal operants in young children (see review by Sautter & LeBlanc, 2006). Sautter and LeBlanc (2006) reviewed empirical articles that evaluated verbal behavior and found that from 1989 to 2004 nine articles specifically addressed functional independence of verbal operants. These research studies evaluated if direct training of one verbal operant resulted in the emergence of another verbal operant without explicit training of that topographically similar but functionally different verbal operant. Results from these studies largely indicated functional independence of verbal operants in young children with and without disabilities (Sautter & LeBlanc, 2006). That is, training one verbal operant did not result in the emergence of topographically similar but untrained verbal operants. However, two studies from this review showed at least 9 some emergence of untrained verbal operants in children (Sautter & LeBlanc, 2006). Others have conducted research that shows speaker and listener repertoires are functionally independent (Guess & Baer, 1973). Speaker repertoires are evaluated through productive language tasks (e.g., vocal verbal behavior) and listener repertoires are evaluated through receptive language tasks (e.g., pointing to items). Guess and Baer (1973) showed that training specific listener responses did not necessarily result in emergence of similar speaker behavior, nor did training speaker behavior result in related listener behavior. However, the results of this study were variable, with some transfer of skills across speaker and listener repertoires despite direct training of only one repertoire (Guess & Baer, 1973). Further research is needed in the area of functional independence of verbal operants. Such research can be conducted during acquisition or deterioration of verbal repertoires to determine the extent to which topographically similar but functionally different verbal responses covary.

According to Skinner's analysis, verbal operants may remain functionally independent upon their deterioration. He noted that a response that was previously in an aphasic individual's repertoire might continue to occur under certain controlling conditions, but that same response might no longer occur under other conditions. He described aphasia by stating, "What has been damaged in aphasia is clearly the functional control of the behavior, and the damage respects the lines of control (Skinner, 1957/1992, p. 195)." Skinner indicated that individuals can have deficits specific to speaking or listening, and he added specificity by predicting that speaker behavior might degrade in a manner that is compatible with his classification of verbal operants. Skinner expected that verbal behavior receiving generalized reinforcers would be most likely to degrade. Thus, mands may be resilient against change because they are not reinforced with generalized reinforcers; rather, controlling conditions for mands involve establishing operations and 10 characteristic consequences (Michael, 1988; Skinner, 1957/1992). Intraverbals, tacts, and echoics all receive generalized reinforcers. In terms of these verbal operants, Skinner predicted that intraverbals and tacts would be damaged more easily than echoics because of the differences in the minimal repertoire involved with each type of verbal behavior. Because of the point-to-point correspondence and functional similarity between very small units of behavior (e.g., speech sounds) and echoed responses, the minimal repertoire of echoics is much smaller than that of other verbal operant categories. Therefore, echoics may remain intact longer than tacts or intraverbals (Skinner, 1957/1992).

Skinner's analysis of verbal behavior has resulted in relatively little empirical work (Dymond, O'Hora, Whelan, & O'Donovan, 2006). Dymond and colleagues (2006) found that 80% of articles citing Verbal Behavior from 1984-2004 were non-empirical articles. Only 4% and 1.4% of these articles were applied and basic studies, respectively. Therefore, Verbal Behavior has generated a great deal of theoretical writing and debate, but it has resulted in much less empirical work (Dymond et al., 2006). Much of the empirical work based on Skinner's analysis has focused on language acquisition, often with children with developmental disabilities (Sautter & LeBlanc, 2006). To the author's knowledge, no empirical research has been conducted to assess verbal behavior in older adults using a verbal operant conceptualization. Because of the lack of empirical evidence, it is unclear if assessments of verbal behavior in older adults will provide clear evidence regarding specific verbal behavior deficits. Furthermore, if assessments are developed and provide clear data on verbal behavior deficits, it is not known if these data will provide clinically relevant information to inform intervention methods.

In order to evaluate Skinner's predictions regarding the degradation of language, researchers must develop appropriate assessment methods. Assessment batteries must 11 include tests that assess each category of verbal behavior. Assessing the verbal operants independently allows for the comparison of performance across various types of verbal behavior. If meaningful differences in the deterioration of language are documented with the assessment measures, then the next phase of research can be undertaken, the development and evaluation of interventions linked to the function-based assessment. A similar research sequence has proven very effective in a related area, the development of functional assessment and functional analysis strategies for problem behavior and the validation of function-based interventions (e.g., Didden, 2007). The research described in this experiment is a first step in this process, the development and preliminary evaluation of a function-based assessment for language deficits in older adults.

Assessment of Language Deficits in Older Adults

Evaluation of verbal skills in older adults typically occurs when there are concerns regarding a decline in language functioning or problems with general memory. When assessing for language deterioration, evaluators examine individuals' ability to repeat spoken words or phrases, engage in spontaneous speech, comprehend language, name items, read text, and write words or sentences (Lezak, Howieson, & Loring, 2004). Several aphasia batteries have been developed that include opportunities to evaluate these skills within one assessment measure. Furthermore, independent verbal assessment measures are used to assess verbal expression (e.g., naming, vocabulary, discourse, fluency) and verbal academic skills (e.g., reading, writing, spelling, acquired knowledge) (Lezak et al., 2004). In the context of general memory concerns, evaluators administer specific verbal memory tests that include tasks such as recalling digit sequences, word lists, word pairs, and/or stories. These memory tests assess for short-term and long-term recall, as well as recognition (Lezak et al., 2004). Memory assessment batteries contain a 12

verbal memory score, which is a summation of individuals' performance on the verbal memory scales. Results from traditional language evaluations contribute to diagnosing aphasia, dementia, or other disorders. However, when a diagnosis is made, it is often assumed that deficiencies are due to brain damage or problems with underlying hypothetical constructs that contribute to the construct of language. These assessment findings and resulting diagnoses may yield little information that can guide efficacious treatment strategies. Behavioral conceptualizations may provide a framework to develop assessment measures with greater sensitivity and treatment utility. Skinner (1957/1992) suggested that researchers consider the controlling conditions of verbal behavior, rather than the topography verbal responses. Likewise, Sidman (1971) recommended analyzing the specific stimulus-response relations that contribute to verbal behavior skills, noting the influence of stimulus control on responding. Using a behavioral approach focuses the assessment on the function of verbal behavior. Traditional approaches, which focus on the topography of responses, rely on hypothetical constructs of language and assume responses are consistent across contexts (Baker et al., 2008; Davis, 2006). However, it is often observed that individuals are able to produce a word in some contexts but cannot produce the same word in other situations. This context specific performance may be due to the same topographic response being controlled by different stimuli (e.g., a word endures as an echoic but not as an intraverbal) or because of the interference of other controlling variables (e.g., stimuli strengthening competing verbal behavior). Similarly, one stimulus might influence a number of responses (Sidman, Stoddard, Mohr, & Leicester, 1971). Therefore, in order to evaluate the variables controlling verbal behavior, Sidman and colleagues (1971) recommended researchers hold either the stimulus or the response constant across testing conditions. After reviewing various behavioral 13 conceptualizations of language, Baker et al. (2008) provided a behavioral taxonomy of aphasia depicting Skinner's verbal operants and various stimulus-response relationships. The behavioral taxonomy was highly specific and revealed 40 different verbal behavior deficits that could be evaluated (Baker et al., 2008). Because of its specificity, this taxonomy might benefit those developing assessments and interventions for older adults with verbal skills deficits.

Purpose

Dementia is a condition affecting both memory and other cognitive functions, one of which is language. Although many have written about language deficits, there are weaknesses in the literature. Primarily, language deficits are assessed and categorized in ways that do not necessarily lead to effective intervention methods. Skinner's (1957/1992) analysis of verbal behavior may lend itself to assessment methods that will identify specific skill deficits, which, in turn, may inform more effective treatment recommendations, in a manner analogous to functional assessment methodology and function-based treatments. However, before determining whether intervention methods can be developed from this analysis, researchers must develop an assessment methodology based on Skinner's model of functional verbal operants and evaluate the manner in which verbal behavior deteriorates in older adults.

The purpose of this study was to evaluate verbal behavior in older adults with and without dementia. There were two primary research questions. First, in what ways do verbal behavior problems differ between older adults with and without dementia? Second, does language deteriorate in a pattern compatible with Skinner's analysis of functionally independent verbal operants? 14 METHOD Participants and Setting

Thirty-one individuals, aged 60 years and older, participated in the study. Fifteen participants met criteria for the Control Group, and 16 participants met criteria for the Dementia Group. Three additional individuals participated in the first session, but did not return to the second session. Of these three individuals, one would have been in the Control Group and two in the Dementia Group, but because they did not complete the study, they were not included in the total sample. All participants were Caucasian and 30 participants were native English speakers. One participant from the Control Group reported Latvian as his first language, but he was a fluent English speaker and had been speaking English for 55 years. The entire sample was made up of seven males and 24 females, with four males in the Control Group and three in the Dementia Group. Self- reported participant ages ranged from 60-86, with mean reported ages of 72.9 and 73.9 for the Control and Dementia Groups, respectively.

Researchers recruited participants from a neuropsychology clinic and a local senior living facility by posting flyers, hosting an informational meeting, and/or calling potential participants. Sessions were conducted in the location from which the participants were recruited.

Measures

Demographic Questionnaire

During the first session, investigators collected information on a number of demographic variables including sex, age, educational history, race, living situation, and diagnosed cognitive impairments (see Appendix A). Dementia Rating Scale-2 fDRS-21

The DRS-2 measures attention, initiation/perseveration, construction, conceptualization, and memory (Jurica, Leitten, & Mattis, 2001) and is used to screen and track cognitive status in older adults (Schmidt, Mattis, Adams & Nestor, 2005). Specific tasks on the DRS-2 include repeating digits forward and backward, following administrator commands, naming items in a category, copying line drawings, identifying similarities and differences, answering factual questions, reading, creating a sentence, remembering words or sentences after a brief delay, and remembering line drawings after a brief delay. Investigators administered the DRS-2 to participants during the first session in order to determine if they met criteria for the Control Group or the Dementia Group. For the purposes of this study, individuals who scored in the mildly, moderately, or severely impaired range were categorized into the Dementia Group. Participants scoring in the below average, average, and above average were classified as "intact" by the DRS- 2 manual (Jurica et al., 2001), and were placed in the Control Group for this study.

Boston Naming Test (BNT)

This instrument contains 60 line-drawn pictures and it measures individuals' ability to name the items depicted in the drawings, often called "confrontation naming" (Mitrushina, Boone, Razani, & D'Elia, 2005). For this study, investigators used a validated short-form of the test by administering only the even-numbered items (Fisher, Tierney, Snow & Szalai, 1999). Because this is a standard assessment often used to assess cognitive and/or language impairment, it was administered to all participants and the results were compared to the results from the other assessment measures. The short- form BNT was administered during the second session. 16 Verbal Behavior Assessment Battery (VBAB")

The investigators developed materials and procedures to assess echoics, tacts, intraverbals, and mands, as detailed in Skinner's conceptual model of verbal behavior. Materials and procedures were also developed to assess selection-based responses. Each of the primary assessments (i.e., Selection, Echoic, Tact, Intraverbal, and Mand) contained 30 items, which were developed from the odd numbered items on the full version of the BNT. The order of the 30 items was randomized across assessment measures. The 3D Mand Assessment contained five items, all of which were different from the 30 stimuli used for the other assessment conditions.

Selection Assessment

The investigator placed a series of three line drawings on the table and asked the participant to point to a certain item, an example of a selection-based response. The selection assessment is compatible with traditional receptive language tasks and evaluates the listener repertoire (Guess & Baer, 1973). Investigators assessed selection for all 30 odd-numbered BNT items. The participant had 5s to respond. Investigators recorded the item selected by the participant and the latency to the response (see Appendix B). Participants completed the Selection Assessment during both sessions (see Table 1 for specific instructions for all verbal operant assessments).

Echoic Assessment

Investigators asked participants to repeat 30 words, one at a time. The participant had 5s to respond. Investigators recorded whether the response was correct or incorrect and the latency to the response (see Appendix C). Correct responses occurred when the participant vocally repeated the word provided by the investigator within five seconds. If 17 participants responded incorrectly, the investigators recorded the incorrect response. The 30 words were the names of the odd-numbered items on the BNT. Participants completed this assessment during sessions one and two.

Tact Assessment

Investigators presented the odd-numbered items from the BNT and asked the participant to name each item. Participants had 5s to initiate responding and 10s to complete their response. Investigators recorded the response as correct or incorrect and noted the response given if it was incorrect (see Appendix D). Investigators also recorded the latency to the response. Correct responding included providing the appropriate name of the presented item within 10s. Participants completed this assessment during both sessions.

Intraverbal Assessment

Investigators described 30 items, without showing a picture of the item, and asked the participant to name the item described. The items were the odd-numbered items from the BNT (see Appendix E). The primary investigator developed the original descriptions by using definitions from the Longman Advanced American Dictionary (2000). Descriptions were revised in order to take out words that were the same or highly similar to the item name (e.g., removing "hang" from the description of a "hanger") and to add detail to the descriptions. Participants had 5s to initiate a response and 10s to complete the response. Investigators recorded responses as correct or incorrect, noting the answer given if incorrect, and recorded the latency to the response. Participant responses were considered correct if they said the name of the item described by the investigator within 10s. Participants completed this assessment during each session. 18 Mand Assessment

Investigators showed participants 30 pairs of scenes. The first scene of each pair was complete; the second scene was the same but with an item missing (see Appendix F). The investigator instructed the participant to request the missing item in order to complete the scene (i.e., the instructor said, "what do you need to complete this picture"). Upon requesting the appropriate item, the investigator gave the participant the item, and the participant placed it in the appropriate location. An error at this point in the Mand Assessment could be the result of two factors: a memory problem or a deficit in a mand repertoire. Therefore, when errors occurred, the investigator administered a remedial, selection-based assessment under the assumption that accurate performance on the selection-based assessment would rule out the memory deficit explanation, leaving the mand deficit as the most likely explanation for the poor performance. That is, if participants did not name the correct item given the original Mand Assessment procedures, the investigator showed the participant pictures of three items. Then, the investigator instructed the participant to point to the item that belonged in the missing space. If participants pointed to the correct picture, the investigator gave the participant the item to place in the scene. If they did not point to the correct picture, the next item was presented. Participants had 5s to initiate each response. Each response (i.e., vocal and selection) was scored separately as correct or incorrect, noting the response given if incorrect. If the vocal mand was correct, the selection aspect was not administered and the response was recorded as correct. Investigators also recorded the latency to the responses. Participants completed the Mand Assessment during each session.

3D Mand Assessment

Investigators gave participants a task to complete with three-dimensional objects, 19 but one object needed to complete the task was missing (see Appendix G). The investigator instructed the participant to complete the task and to ask for any additional items needed to complete the task. If the participant asked for the missing item, the investigator presented it. If the participant did not request the missing item, the investigator took the other items away and presented the next task. Investigators presented five tasks for this assessment. Participants had 5s to initiate a response and 10s to complete the response. Investigators recorded the responses as correct or incorrect and the latency to the response. If incorrect, the response given was recorded. Participants completed the 3D Mand Assessment during each session. The 3D Mand Assessment was used to probe mands for three-dimensional objects, because it was thought that these procedures might allow for more powerful establishing operations than the procedures used for the Mand Assessment. 20

Table 1

Summary of Test Instructions

Test Introduction Instructions Item-bv-Item Instructions

In this part, I am going to show you groups Selection of pictures. Then, I will ask you to point to a Point to the picture from the group. (present array) In this part, I am going to say words. I want Echoic you to repeat each one after me. Say In this part, I am going to show you pictures Name this. Tact of items. I want you to name each item. (present picture) In this part, I am going to describe a word or This is Intraverbal item. I want you to tell me what I am (e.g., This is a piece of describing. furniture that you sleep on.) Look at the first picture of In this part, I am going to show you sets of this set. Look at the second. two pictures. The first picture will be a What do you need to complete picture. The second picture will be complete this picture? Mand the exact same picture, but one piece will be (present picture 1, cover it missing. I want you to tell me what you need and present picture 2) to complete the picture. Once you do, I will (if correct, give missing item; give you the item to place on the picture to if incorrect, move to mand complete it. selection) Point to what you need to complete this picture. Mand (present array) Selection (no general instructions - this is only given (if correct, give missing item; immediately after each item missed on mand if incorrect move to next item assessment items) on mand assessment) For this job I want you to . Tell me if vou In this part, I am going to give you jobs to do need anything else to 3D Mand and I want you to complete each job. If you complete the job. are missing any items that you need to (present task) complete the job, tell me what you need and I (if correct, give item; if will give you the item. incorrect, move to next task) Note. Bold print indicates the exact instructions given by evaluator. 21

Procedure

Session One

Investigators obtained consent and then administered the Demographic Questionnaire and the DRS-2. Then, investigators administered the VBAB, which included the Selection Assessment, Echoic Assessment, Tact Assessment, Intraverbal Assessment, Mand Assessment, and 3D Mand Assessment. The order of administration of the verbal operant assessments was randomized across participants, but it was held constant across sessions for each individual participant. Investigators offered participants a break between each assessment.

Session Two

During session two, investigators administered the short-form BNT and the VBAB. The purpose of session two was to examine test-retest reliability of the verbal operant assessments. As in session one, investigators offered breaks between each assessment. At the end of session two, investigators thanked the participant for their participation and offered to send a brief report of the final research findings upon completion of the study.

Target Behaviors, Data Collection, and Data Analysis

For verbal operant assessments, investigators recorded responses as correct or incorrect. Correct responses included pointing to the correct line drawing or saying the name of the item. Other verbal behavior in which the participant engaged was not considered part of the answer for scoring purposes. If a participant offered several answers or ambiguous answers, investigators asked the participant to choose one answer 22

or to be more specific in responding. Investigators also collected data using standardized tests (i.e., DRS-2 and BNT) and did so by following the assessment protocols published in the administration manuals. Investigators did not provide verbal feedback on the accuracy of responding on the assessment measures. Between group differences were analyzed using Mann-Whitney tests, and within- subjects differences were evaluated using Friedman's ANOVAs. In addition, data were analyzed using regression analyses with age and DRS-2 scores as predictor variables and total and individual verbal operant assessment scores as outcome variables. Kendall's tau correlations were used to determine test-retest reliability of the verbal operant assessments and the degree to which the short-form BNT was correlated with the DRS-2 and the verbal operant assessments. Moreover, a Wilcoxon Signed Rank Test was used to compare performance at session one and session two. Finally, a non-statistical error analysis was completed to determine whether participants made errors on the same stimuli across all assessments (e.g., responded incorrectly on "protractor" across assessments) or if errors were inconsistent across assessments (e.g., responded incorrectly on "protractor" on the intraverbal assessment but gave the correct response on the other assessments). Data from the error analysis helped determine if verbal operants were functionally independent.

Interobserver Agreement

Interobserver agreement (10A) data were collected on the accuracy with which the participants' responses to the DRS-2, short-form BNT, and verbal operant assessments were scored. Before evaluating IOA, the independent observer was trained by viewing videotaped sessions, scoring participant responses, and discussing results with the primary investigator. Training continued until the independent observer and the 23 primary investigator reached over 95% agreement for the DRS-2 and BNT and 100% agreement for the verbal operant assessments. The independent observer viewed videotaped sessions for 10 of the 31 participants, or about 32% of the sessions. Agreement was defined as the in-session evaluator and the independent observer scoring a participant's response in the same manner on a particular trial, for example, agreeing that a response was correct or incorrect, or giving the same numerical value to a response on a specific item. The primary investigator calculated percentage of agreement by dividing the number of agreements by agreements plus disagreements and multiplying by 100. Agreement for the DRS-2 and BNT were 98.9% and 99%, respectively. The total agreement for verbal operant assessments was 99.8% (see Table 2 for individual assessment agreement and agreement ranges).

Procedural Integrity

The independent observer was trained to evaluate procedural integrity in the same manner, and met the same criterion, as previously described for IOA. The independent observer viewed videotaped sessions for the same 10 participants as were viewed for IOA (i.e., 32% of participants) and evaluated the degree to which the in-session evaluator followed the assessment protocols. Following the protocols included appropriately delivering introduction instructions and item-by-item instructions. Procedural integrity was calculated by dividing the correctly administered instructions for each assessment by the total number of instructions and multiplying by 100. The percent of correctly implemented instructions was 99.1 and 96.5 for the DRS-2 and BNT, respectively. The percent of correctly implemented steps for all of the verbal operant assessments was 99.9 (see Table 3 for individual assessment data and ranges). 24 Table 2

Interobserver Agreement Data

Assessment % Agreement" Range" DRS-2 98.9 98.4-100 BNT 99 96.7-100

Selection Session 1 100 100 Echoic Session 1 99.7 96.7-100 Tact Session 1 100 100 Intraverbal Session 1 99.7 97-100 Mand Session 1 99.3 93-100 3D Mand Session 1 100 100 Verbal Operant Assessment Session 1 Total 99.7 93-100

Selection Session 2 100 100 Echoic Session 2 99.7 97-100 Tact Session 2 99.7 97-100 Intraverbal Session 2 100 100 Mand Session 2 99.7 97-100 3D Mand Session 2 100 100 Verbal Operant Assessment Session 2 Total 99.8 97-100

Verbal Operant Assessment Total 99.8 93-100 Overall Total 99.6 93-100 "Calculated by dividing the agreements by agreements plus disagreements and converting to a percentage. 25

Table 3

Procedural Integrity Data

Assessment % Correct11 Range" DRS-2 99.1 96.1-100 BNT 96.5 91.7-100

Selection Session 1 100 100 Echoic Session 1 99.7 97-100 Tact Session 1 100 100 Intraverbal Session 1 100 100 Mand Session 1 100 100 3D Mand Session 1 100 100 Verbal Operant Assessment Session 1 Total 99.9 97-100

Selection Session 2 100 100 Echoic Session 2 100 100 Tact Session 2 99.7 97-100 Intraverbal Session 2 100 100 Mand Session 2 99.4 97-100 3D Mand Session 2 100 100 Verbal Operant Assessment Session 2 Total 99.8 97-100

Verbal Operant Assessment Total 99.9 97-100 Overall Total 99.5 91.7-100 "Calculated by dividing correctly implemented steps by total possible steps and converting to a percentage. 26

RESULTS

Thirty-one older adults completed the study, 15 in the Control Group and 16 in the Dementia Group. Participants' DRS-2 scores allowed for categorization into Control and Dementia Groups based on level of cognitive functioning; participants in this sample ranged from moderately impaired to above average (see Table 4). Participants in the Control Group (M= 10.27, SE = .396) had significantly higher DRS-2 scores than participants in the Dementia Group (M = 6.44, SE = .376), t(29) = 7.02,p< .001. The age distribution was similar across groups. Individuals ranged in age from 60-86 (M=72.9, SE = 2) in the Control Group and 61-84 (M=73.9, SE = 1.97) in the Dementia Group. The difference in mean age between groups was not significant, r(29) = -.36, p = .72.

Table 4

Participants' Level of Impairment

Categorization n Control Above Average 1 Average 2 Below Average 12

Dementia Mildly Impaired 11 Moderately Impaired 5 Severely Impaired 0

The following results are based on the first administration of the verbal behavior assessments except when specifically indicated. Data from the second administration were only used for evaluating test-retest reliability and determining differences in 27 performance between the first and second administration. Data analyses included between-groups comparisons, within-subjects evaluations, regression analyses, correlation analyses, and a non-statistical error analysis. Selection, Echoic, and 3D Mand Assessments were excluded from the between-groups, within-subjects, regression, and error analyses due to ceiling effects (see Figure 1 for mean percent correct on assessments; see Table 5 for raw score means, standard deviations, and ranges). The Mand Assessment was split into two scores, one representing the initial verbal mand (i.e., verbalizing the item that was missing from the scene) and one representing the remedial selection option (i.e., pointing to a picture of the missing item). The mand selection portion suffered from ceiling effects and was also excluded from the between-groups, within-subjects, regression, and error analyses. Therefore, Mand Assessment results are based upon the initial verbal aspect of the Mand Assessment.

VBAB results generally indicated higher average performance on assessment measures for the Control Group than the Dementia Group. Group differences were most evident on the Tact, Intraverbal, and Mand Assessments. Performance across verbal operant categories also differed. More specifically, mean performances on the Selection and Echoic Assessments were higher than on the Tact, Intraverbal, or Mand Assessments. Furthermore, average performance on tacts was greater than intraverbals, and performance on intraverbals was higher than mands (see Figure 1 and Table 5). 28 Figure 1

Verbal Behavior Assessment Mean Percent Correct

120 1100 • Control i H Dementia

Assessments

Table 5

Assessment Score Means, Standard Deviations, and Ranges

Control in-=15) Dementia (n =16) Overall (N=3l) Assessment M(SD) Range M(SD) Range M(SD) Range Selection 30(0) 30 30(0) 30 30(0) 30 Echoic 29.87 (0.35) 29-30 29.88 (0.34) 29-30 29.87 (0.34) 29-30 Tact 28.93(1.44) 26-30 26.62 (3.86) 16-30 27.74(3.13) 16-30 Intraverbal 25.93 (4.37) 17-30 24.44 (4.1) 15-30 25.16(4.21) 25-30 Mand 25.13 (5.9) 13-30 21.38(5.86) 7-30 23.19(6.09) 7-30 Mand Selection 29.93 (0.26) 29-30 29.44(1.1) 27-30 29.68 (0.83) 27-30 3D Mand 4.87 (0.35) 4-5 4.69 (0.48) 4-5 4.77 (0.43) 4-5 Note. 3D Mand Assessment has 5 items; all other assessments have 30 items.

Nonparametric tests were used because the verbal operant assessment data violated the assumptions of parametric analyses. More specifically, data in this study were not normally distributed. Although the scores on the Intraverbal Assessment did not 29 differ significantly from normal distribution, D(31)=0.142,/?=.116, the scores on the Tact Assessment, D(31)=0.307,/?<.001, and the Mand Assessment, D(31)=0.197,/?=.003, were significantly different from normal distribution. The scores on these measures were negatively skewed, clustering toward higher performance. Therefore, group differences, within-subject differences, and correlations were evaluated using nonparametric tests. Nonparametric tests evaluate ranked data rather than the raw data. Ranks are determined by giving the lowest raw data point the lowest rank (i.e., 1) and working up to higher ranks with higher raw scores. Tied ranks are acceptable (Field, 2009).

Between Groups Differences

Group differences were evaluated to help determine how verbal behavior differs in individuals with and without dementia. Mann-Whitney tests were used to evaluate group differences on the Tact, Intraverbal and Mand Assessments. A Bonferroni correction was applied so all effects are reported at a .0167 level of significance. According to mean ranks, the Control Group performed better on the Tact, Intraverbal, and Mand Assessments than the Dementia Group (see Figure 2). The Control Group had significantly higher scores on the Tact Assessment than the Dementia Group, [7=60.5, p=.007, r=-.43, but the differences were not significant for the Intraverbal Assessment, [7=89.0,/>=.l 12, r=-.22, or the Mand Assessment, [7=67.5,/?= .019, r=-.37. Notably, differences on the Mand Assessment approached significance, but using the Bonferroni correction, the groups did not differ significantly. 30 Figure 2

Between Groups Mean Rank Differences

• Control E Dementia

Tact Intraverbal Mand Assessment

Within Subject Differences and Non-Statistical Error Analysis

Within-subjects analyses and a non-statistical error analysis were used to evaluate the functional independence of verbal operant categories. Friedman's ANOVAs were used to evaluate the differences among the assessments for each group, and Wilcoxon tests were used to follow up these analyses (see Figure 3). Bonferroni corrections were applied; therefore, all effects are reported at a .0167 level of significance. For the Control Group, scores on the Tact, Intraverbal, and Mand Assessments differed significantly (%2(2)=7.9,/>=.019). Individuals performed better on the Tact Assessment than the Intraverbal Assessment, T=S.00,p=.0l5, r= -.44, and better on the Tact Assessment than the Mand Assessment, r=5.00,/?=.007, r= -.49. There was no significant difference between the Intraverbal and Mand Assessments, r=34.50,/?=.438, r= -.14. For the Dementia Group, scores on the Tact, Intraverbal, and Mand Assessments also differed 31

significantly (x2(2)=12.7,/?=.002). Individuals in the Dementia Group performed better on the Tact Assessment than the Intraverbal Assessment, T=l5.50,p=.0l 1, r- -.64. They also performed better on the Tact Assessment than the Mand Assessment, 7=12.50, p=.004, r= -.51. There was not a significant difference between the Intraverbal and Mand Assessments, r=20.00,/?=.041, r=-.36. Therefore, the pattern of findings for the Control Group and the Dementia Group was the same; performance on tacts was significantly better than on intraverbals or mands, but there was no significant difference between intraverbals and mands.

Figure 3

Within-Subjects Mean Rank Differences

• Tact Q Intraverbal • Mand

Control Dementia Group

In order to further evaluate the functional independence of verbal operant categories, investigators performed a non-statistical error analysis. If verbal behavior categories were functionally independent, one would expect to see within-subjects differences across assessments on topographically identical responses. More specifically, participants would make inconsistent errors across the assessment measures on identical 32

stimuli. Evidence against functional independence would show that individuals respond exactly the same on the same stimulus across assessments. That is, if they make an error on a word during one assessment, they would consistently make errors on that same word across all assessments, receiving a total raw score of zero on that item. All individuals performed perfectly on the Selection Assessment and performed well on the Echoic Assessment, thus no individual received a raw score of zero across all measures on any one stimulus. Yet, some individuals consistently received a score of zero on the same stimulus across the Tact, Intraverbal, and Mand Assessments, giving them a total raw score of zero across these assessments (see Table 6). Twelve participants consistently missed at least one of the stimuli across these three assessments. In contrast,

30 participants made errors on topographically identical stimuli on only one or two of the assessments but not across all the verbal operant assessments for that particular stimulus (see Table 7). These 30 participants performed inconsistently across assessments and received a total raw score of one or two across the three assessments. These findings indicate topographically identical responses occurred correctly under some assessment conditions (e.g., the Tact Assessment) but not under other assessment conditions (e.g., the same word was incorrect under the Mand Assessment). This type of inconsistency across different assessments provides some evidence for the functional independence of verbal operants. Overall, there were 284 instances when a participant made inconsistent errors on a stimulus across assessment conditions (see Table 7 total). This is compared to only

31 occasions when a participant consistently made errors on a stimulus across assessment conditions (see Table 6 total). Based on these findings, it appears the function of the response influences responding more than its topography. 33 Table 6

Number of Participants with Consistent Errors Across Tact, Intraverbal, and Mand Assessments

Stimuli Control Dementia Total Bed 0 0 0 Pencil 0 0 0 Whistle 0 0 0 Comb 0 0 0 Saw 0 0 0 Helicopter 0 0 0 Octopus 0 1 1 Hanger 0 0 0 Camel 0 0 0 Pretzel 0 2 2 Racquet 0 0 0 Volcano 0 0 0 Dart 0 0 0 Globe 0 0 0 Beaver 0 0 0 Rhinoceros 0 1 1 Igloo 0 0 0 Dominos 0 0 0 Escalator 0 2 2 Hammock 0 0 0 Pelican 0 2 2 Pyramid 0 2 2 Unicorn 1 1 2 Accordion 0 0 0 Asparagus 0 0 0 Latch 1 3 4 Scroll 0 1 1 Sphinx 1 5 6 Trellis 0 2 2 Protractor 2 4 6 Total 5 26 31 Note. Some participants responded incorrectly across assessments on multiple stimuli. 34 Table 7 Number of Participants with Inconsistent Errors Across Tact, Intraverbal, and Mand Assessments

Stimuli Control Dementia Total Bed 3 3 6 Pencil 1 4 5 Whistle 2 2 4 Comb 2 1 3 Saw 10 9 19 Helicopter 1 3 4 Octopus 3 5 8 Hanger 4 6 10 Camel 1 3 4 Pretzel 4 8 12 Racquet 5 9 14 Volcano 5 10 15 Dart 4 11 15 Globe 1 3 4 Beaver 3 9 12 Rhinoceros 5 5 10 Igloo 2 3 5 Dominos 4 6 10 Escalator 0 1 1 Hammock 5 5 10 Pelican 3 6 9 Pyramid 4 6 10 Unicorn 4 4 8 Accordion 7 10 17 Asparagus 7 7 14 Latch 9 11 20 Scroll 6 7 13 Sphinx 3 3 6 Trellis 4 5 9 Protractor 2 5 7 Total 114 170 284 Atote. Some participants made inconsistent errors across assessments on multiple stimuli. 35 Regression Analyses

Data were also analyzed using regression analyses with age and DRS-2 score as predictor variables, and total and individual assessment scores as outcome variables. Age was used as the first predictor in the regression models, as previous research indicates language changes with age, regardless of the presence or absence of other impairments (Benson & Ardila, 1996; Kempler, 2005). Total score was the sum of the Tact, Intra verbal, and Mand Assessment raw scores from the first administration. Total score was significantly predicted by age (F(l, 29)= 1 1,/J=.002), but DRS-2 scores did not significantly predict total score above and beyond age (see Table 8). Age accounted for 27.5% of the variance in total score. As age increased, total score decreased. Age alone or the combination of age and DRS-2 scores did not significantly predict performance on the Tact Assessment. However, age did significantly predict performance on the Intraverbal and Mand Assessments [(F(l. 29)=17.1,/?=.001) for Intraverbal; F( 1,29)= 10.4,/?=.003 for Mand)] (see Tables 9 and 10). Age accounted for 32.7% and 26.4% of the variance in the Intraverbal and Mand Assessment scores, respectively. As age increased, performance on Intraverbal and Mand Assessments decreased. DRS-2 scores did not significantly predict performance on Intraverbal or Mand Assessments beyond that which age predicted performance. It is not surprising that age and DRS-2 score did not predict performance on the Tact Assessment, because the data did not fit well in a regression model. Specifically, the residuals were not normally distributed for the Tact Assessment analysis; however, residuals were normally distributed for the Total Score, Intraverbal Assessment, and Mand Assessment analyses. 36

Table 8

Regression Analysis Summary for Total Score

Variable B SEB R2 Step 1 0.275** Constant 132.4 17.1 Age -0.77 0.23 -0.52 Step 2 NS Constant 118.1 17.95 Age -0.72 .22 -0.50 DRS-2 1.34 0.70 0.30 *#•p

Table 9

Regression Analysis Summary for Intraverbal

Variable B SEB R2 Stepl 0.327*** Constant 48.11 6.14 Age -0.31 0.08 -0.57 Step 2 NS Constant 45.51 6.77 Age -0.31 0.08 -0.56 DRS-2 0.24 0.26 0.14 ***p < .001 37

Table 10

Regression Analysis Summary for Mand

Variable B SEB g R2 Step 1 0.264** Constant 53 9.29 Age -0.41 .126 -0.5 b Step 2 NS Constant 45.29 9.79 Age -0.38 0.12 -048 DRS-2 0.72 0.38 0.29 #*•p < .01

Test-Retest Reliability

Kendall's tau was used for all correlation analyses because it is a nonparametric test that works well with small sample sizes and with data that results in several tied ranks (Field, 2009). Kendall's tau was used to evaluate test-retest reliability of the verbal operant assessments. The Selection Assessment was removed from this analysis because performance was constant across all participants, but all other verbal behavior assessments were used for evaluating test-retest reliability. Echoic Assessment 1 was not significantly correlated with Echoic Assessment 2,x=.199,/)= .275. Echoic Assessment performances were almost constant across participants and sessions, which probably contributed to the lack of significant findings (see Appendix H for VBAB raw scores during session one and two). However, Tact, Intraverbal, Mand, Mand Selection, and 3D Mand Assessments at session one and session two were all significantly correlated at the .01 level [(x=.403,/?=.009) for Tact; (x=.719,/?<.001) for Intraverbal; (x=.677,/7<.001) for Mand; (x=.496,/?=.005) for Mand Selection; (x=.600,p=.001) for 3D Mand]. A 38

Wilcoxon Signed Ranks Test was used to evaluate the difference between performances from session one and session two on the Total Score (i.e., sum of Tact, Intraverbal, and Mand Assessments). Participants performed significantly better during session two than session one, T=l ,p < .001, r- -.08. All but one participant had a higher Total Score during session two than session one. Improved performance from session one to session two was likely due to practice effects, as corrective feedback was not provided to participants during either session.

j Correlations with Short-Form BNT

Kendal's tau was also used to evaluate the relationship between the short-form BNT and the other assessment measures. Again, the Selection Assessment was removed from the analysis due to constant performance across participants. A significant correlation was found between the short-form BNT and the DRS-2, x=.287,/?= .039. The short-form BNT was not significantly correlated with the Echoic Assessment or the 3D Mand Assessment, possibly because of the consistently high performance across participants on the Echoic and 3D Mand Assessments. However, the short-form BNT was significantly correlated with the Intraverbal Assessment x=.330, the Mand Assessment, x=.327, and the Mand Selection Assessment, x=.365 {d\\ps, <.05). The highest correlation was between the short-form BNT and the Tact Assessment, x= .551,/?<.001, which is understandable, as these assessments required the same skill of naming line drawings. 39

DISCUSSION

The current research reports the development, administration, and results of a "verbal behavior assessment battery" (i.e., VBAB) for older adults. Investigators developed assessments based on Skinner's functional verbal operants, with procedures that simulated the controlling conditions for echoics, tacts, intra verbals, and mands, as well as nonverbal selection responses. In keeping with the recommendations of Sidman and colleagues (1971), the same 30 stimuli were used across the assessment conditions. An additional assessment measure that probed for mands of three-dimensional objects was included in the evaluation, but this assessment used five different stimuli than the 30 used in the other assessments. The VBAB was administered to older adults with and without dementia, as determined by the DRS-2.

Results from the VBAB indicated that participants in the Dementia Group performed more poorly on tacts than those in the Control Group. Therefore, tact assessments may provide a sensitive measure of language deficits associated with dementia and should be included in future evaluations of language impairment in older adults. In addition to tacts, incorporating a measure of intraverbals and mands in assessment batteries appears warranted. Although performance was not statistically different, the Dementia Group had lower average scores on the Intraverbal and Mand Assessments than the Control Group, indicating these assessments may also be sensitive to declining language skills.

In addition to group differences, results from the VBAB also revealed within- subjects differences among the verbal operant assessments, and this occurred despite using the same 30 stimuli across these assessments. This inconsistent and idiosyncratic 40 loss of functional verbal behavior provides preliminary support for Skinner's claims that verbal operant categories are acquired and lost independently of one another. The ability to identify the strengths and weakness of older adults across verbal operant categories may facilitate the development and evaluation of targeted, function-based interventions.

Group Differences

The first research question addressed the way in which verbal behavior differs between older adults with and without dementia. Results revealed individuals with dementia performed significantly more poorly than those without dementia on the Tact Assessment. On the Intraverbal and Mand Assessments, the two groups differed, with the Control Group showing higher average scores than the Dementia Group, but the magnitude of these group differences did not obtain statistical significance. These findings suggest tacts are most sensitive to performance decrements associated with dementia in older adults, yet intraverbals and mands may also be sensitive to declining functioning. Therefore, it is recommended that individuals include an evaluation of tacts, intraverbals, and mands when developing language-based assessment batteries for older adults.

The Boston Naming Test (BNT) is a test traditionally used to evaluate confrontation naming, and it has been shown to differentiate groups of individuals with and without aphasia and dementia (Lezak et al., 2004). Not surprisingly, the current results showed a high correlation between a validated short-form BNT and the Tact Assessment. The tasks on these two measures were basically the same, naming line drawings. Furthermore, the line drawings used in the Tact Assessment were taken from the full version of the BNT. Therefore, using the BNT could be appropriate for assessing language degradation associated with dementia. However, performance on the BNT has 41 traditionally been attributed to the perception of the drawings, accessing the conceptual representation, and retrieving the name (Lezak et al., 2004), all skills that are accounted for by hypothetical constructs. Errors on the BNT are described based on their topography, for example, reporting that errors are semantically, phonologically, or perceptually related to the target response (Lezak et al., 2004). Describing errors topographically and attributing performance to hypothetical constructs is problematic because these approaches often assume that the absence of a word in one context means the word will not occur under other contexts. Yet, individuals are often able to provide a specific word in some situations but not others, suggesting words are not simply lost from the verbal repertoire. The current findings support this notion. Group differences were observed on the Tact, Intraverbal, and Mand Assessments, but not on the Selection or Echoic Assessments, despite evaluating for the same 30 responses across these measures. Moreover, significant group differences were only found on the Tact Assessment. Therefore, the BNT might provide a good platform for assessing language in older adults, but it seems more appropriate to interpret the findings based on the function of the responses (e.g., controlling variables for a tact), rather than their topography.

Although results revealed significantly different performances on the Tact Assessment, the two groups performed similarly on the other verbal behavior assessments used in this study. According to statistical analyses, the groups did not differ significantly on the Mand or Intraverbal Assessments. Furthermore, the Selection, Echoic, Mand Selection, and 3D Mand Assessments suffered from ceiling effects, indicating participants from both groups performed well on these measures. Ceiling effects on various measures may have occurred for a number of reasons. Skinner predicted echoics would remain in individuals' repertoires because of the minimal repertoire required to emit this form of verbal behavior (Skinner, 1957/1992). Echoics are unique because one 42 can observe functional correspondence between stimuli and responses at the level of speech sounds, or very small units of behavior. Thus, when any echoic behavior is reinforced, the reinforcer strengthens the minimal repertoires, or speech sounds that make up the word, as well as the specific echoed response. This allows for frequent practice and reinforcement of echoic behavior (Skinner, 1957/1992). The high rate of reinforcement for echoic behavior might contribute to why it is learned early, provides the basis for language acquisition, and may be the verbal operant category most resistant to change associated with dementia. All participants performed well on the Echoic Assessment, supporting Skinner's prediction regarding the resilience of echoics.

In addition to high performance on echoics, all individuals performed perfectly on the Selection Assessment. This assessment was different from the others in that it was most similar to traditional receptive language assessments, rather than expressive language assessments. That is, participants engaged in nonverbal behavior (i.e., pointing to the picture) following the verbal instructions given by the evaluator. Sundberg and Michael (2001) noted Skinner was careful to avoid describing verbal behavior in expressive and receptive terms because these categorizations led individuals to believe the various skills came from the same underlying process. Skinner believed speaker and listener repertoires were both important, but also functionally independent and not developed from the same underlying process (Sundberg & Michael, 2001). Because both speaker and listener responses are important to functional communication, the current evaluation included the Selection Assessment. Individuals may have performed well on the Selection Assessment because of the additional sources of stimulus control that supplemented the evaluator's instructions. That is, two visual stimuli other than the target stimulus were present, and these accompanying stimuli may have provided supplementary information that assisted with correct responding. For example, one may 43 not be able to tact "sphinx" when shown a picture of a sphinx, but when shown a picture of a sphinx and two other items (e.g., tennis racquet and hanger), one might be able to choose the sphinx by eliminating the other options. In addition, pointing, which was controlled by a limited array of stimulus events (i.e., three pictures), might have been a relatively simple response compared to generating a unique verbal response, as was required during other verbal operant assessment conditions. Similar to the Selection Assessment, participants performed well on the selection portion of the Mand Assessment, and this may have occurred for similar reasons. That is, individuals may not have been able to generate a unique, vocal mand during the standard condition; nevertheless, when presented with options, they were able to use contextual cues of the scene and the three items to determine the "best fit" and select (i.e., point to) the appropriate missing item. During the remedial portion of the Mand Assessment, individuals did not engage in pure mands. Rather, they engaged in a nonverbal, selection response, but this was reinforced with the specified consequence. Therefore, the responses could have been multiply controlled by the visual stimuli and the establishing operation. If individuals have problems with pure mands or other forms of vocal verbal behavior, a selection-based response might provide an alternative way for individuals to engage in functional communication (e.g., The Picture Exchange System (Bondy & Frost, 2001)).

Participants in both groups also performed well on the 3D Mand Assessment. The current study used the 3D Mand Assessment to sample mands under conditions where individuals had to perform hands-on tasks and mand for three-dimensional objects, because it was thought that older adults may be more motivated to complete a task with three-dimensional objects than with two-dimensional line drawings. More specifically, the 3D Mand Assessment procedures might allow for more powerful establishing 44 operations than the Mand Assessment procedures. Given the high performance on the 3D Mand Assessment compared to variable performance on the Mand Assessment, it appears that the 3D Mand Assessment conditions may have created more powerful establishing operations than the Mand Assessment conditions. However, although establishing operations and specified reinforcers control mands, it is important to consider the history of reinforcement for responses when predicting the probability of responding (Skinner, 1953/1965). The Mand Assessment required some responses that probably had never been reinforced in the past (e.g., sphinx), but individuals likely had an extensive history of reinforcement for manding for the five items used for the 3D Mand Assessment (i.e., paper, screwdriver, scissors, ruler, and key). Therefore, the difference in performance between the 3D Mand Assessment and the Mand Assessment could also be due to the history of reinforcement for the responses required in each assessment. Further research examining establishing operations and/or the likely history of reinforcement for responses could help in determining the best assessment procedures for mands.

Although there are some areas in which verbal behavior performance did not differ between those with and without dementia, the current results indicated that the two groups differed significantly on the Tact Assessment. The Control Group performed significantly better on the Tact Assessment than the Dementia Group. Moreover, the Control Group had higher mean scores on the Intraverbal and Mand Assessments than the Control Group, but these differences did not reach statistical significance. The current findings suggest tacts are particularly sensitive to decline in functioning, but intraverbals and mands may also be sensitive to language difficulties associated with dementia. Baker et al. (2008) recommended using existing, validated tools for assessing verbal behavior if the assessment tools measure the appropriate controlling variables. It appears the BNT is a measure that could be used to evaluate tacts, as it requires individuals to name visual 45 stimuli and it was highly correlated with the Tact Assessment in the current study. Future evaluations should include the BNT or a similar measure, with the intention of evaluating the tact repertoire, as tacts appear to be the verbal operant most sensitive to language difficulties associated with dementia. Given the potential sensitivity of intraverbals and mands, future assessment batteries should also include and evaluation of these verbal operant categories.

Functional Independence

The current research evaluated the functional independence of verbal behavior in older adults by analyzing within-subjects differences across assessments. If verbal behavior is functionally independent, one would expect to find within-subjects differences across the measures. The same 30 stimuli were used for the Selection, Echoic, Tact, Intraverbal, and Mand Assessments, but the controlling variables differed across assessments. If an individual "loses" a word from their repertoire completely, they would perform exactly the same across all verbal operant assessment measures, that is, they would be unable to generate the word regardless of the verbal operant being assessed. However, if antecedents and consequences control responses, the same response might occur under some testing conditions but not others.

In evaluating within-subjects differences, the investigators examined the data separately for each group. However, the pattern of findings was the same for each group: participants from both groups performed better on the Tact Assessment than the Mand or Intraverbal Assessments, and performance on Mand and Intraverbal Assessments did not differ. These findings support Skinner's conceptualization of functionally independent verbal operants, because the stimuli used in each assessment were the same, but participants' performance on the assessments differed. Moreover, the error analysis 46 revealed that 30 of 31 participants were able to give topographically identical responses under some assessment conditions but not others. For example, they were able to provide the correct response on an item during the Tact Assessment, but then made errors on the same item during the Intra verbal or Mand Assessments. Twelve participants made consistent errors on at least one of the responses across the Tact, Intraverbal, and Mand Assessments. Although this finding might contradict Skinner's analysis because the performance was consistently poor across conditions, there are alternative explanations. The items most frequently missed across assessments were the more difficult items, assuming difficulty level based on the original order of stimuli on the BNT. Therefore, weak performance on these items may not reveal deterioration in functioning; rather, individuals may have never acquired these responses as tacts, intraverbals, or mands. Alternatively, participants may infrequently use these responses under any context; thus, the history of reinforcement for these responses was weak. In addition to these explanations, it is important to consider that there were only 31 instances of participants consistently making errors any one stimulus across the Tact, Intraverbal and Mand Assessments. This is compared to 284 instances in which participants responded inconsistently to the same stimulus across assessment conditions. Therefore, the data provide more support for Skinner's conceptualization of functionally independent verbal operants than for other conceptualizations.

Skinner predicted mands would be resilient to change because of the conditions controlling the mand, namely establishing operations and characteristic consequences (Michael, 1988; Skinner, 1957/1992). However, in this study, participants from both groups performed significantly better on tacts than mands. Furthermore, performance on mands was not significantly different from performance on intraverbals. Based on these findings, one might assume Skinner's prediction regarding the resiliency of mands was 47

incorrect. However, Skinner made his prediction assuming a pure mand. The Mand Assessment in this study did not assess pure mands; instead, the procedures set up a condition in which verbal behavior was multiply controlled. The intention was to create a situation in which participants were deprived of an item and, once the item was requested, the response was reinforced with the specified item, hence setting up the occasion for a mand. However, because individuals saw the full scene only seconds before viewing the scene with the missing item, they may have responded to the prior visual stimulus. In addition, the evaluator provided a verbal stimulus, "What do you need to complete this picture?" Participants may have responded to this antecedent stimulus. Some have called verbal behavior under control of this combination of conditions an intraverbal-mand-tact, because antecedents to the response included a preceding verbal stimulus, an establishing operation, and a visual stimulus (Bondy, Tincani, and Frost, 2004). It is difficult to determine the extent to which each variable controlled responding, be it the establishing operation (i.e., absence of an item to complete a scene), the presence of a prior visual stimulus, or the antecedent verbal stimulus. Because verbal behavior was multiply controlled during the Mand Assessment, it is premature to assume Skinner's prediction of the resilience of pure mands was incorrect. Although the additional antecedent stimuli might be assumed to strengthen the responses, the presence of multiple antecedent stimuli could also decrease the potency of any one of the antecedent conditions, and, in this case, possibly decreasing the influence of the establishing operation. Likewise, the participants may have been more motivated by the social reinforcers (e.g., the social interaction with the evaluator) than the specified reinforcers (i.e., the missing items). Thus, verbal behavior during the Mand Assessment was multiply controlled, and it is difficult to determine whether or not pure mands would be more resilient to language changes associated with older adulthood. 48

It is interesting to note participants in both groups had the same pattern of findings with regard to within-subjects differences. That is, results indicated tacts were stronger than intra verbals or mands, and the latter two did not differ. Age appeared to account for much of the variability in verbal behavior performance. The regression analysis indicated that age accounted for approximately 28% of the total score (i.e., total of the Tact, Intraverbal, and Mand Assessments), 33% of the Intraverbal Assessment score, and 26% of the Mand Assessment score, and adding DRS-2 scores to the models did not account for significantly more of the variation in performance. Therefore, it appears that age may have a great influence on verbal behavior performance. The age range and mean age were comparable across groups, which might explain the similar pattern of findings. However, the pattern of performance could have little to do with the decline of verbal behavior with age. The current study evaluated individuals at only one point in their lifespan, and therefore, the current findings may not reveal a change or deterioration in verbal behavior performance in older individuals. Rather, the pattern of within-subjects findings for the participants in this study might be parallel to that of the general population. That is, young adults without dementia may also perform better on the Tact Assessment than the Intraverbal or Mand Assessments. Further research might evaluate verbal behavior in younger adults to address this concern. If this research were completed and the findings in the general population match those in this study, the data would suggest verbal operants are functionally independent in sophisticated, non- impaired speakers, and later researcher would need to address the deterioration of verbal behavior with more impaired older adults.

Traditional accounts of language attribute verbal skills to hypothetical constructs and often assume specific skills (e.g., words) will be present or absent regardless of the environmental conditions. However, the current data indicated participants' performance 49

differed depending on the manipulation of controlling variables. More specifically, participants had a higher likelihood of providing the correct response during the Tact Assessment than the Intraverbal or Mand Assessments, even though identical responses were required across the assessments. These results indicate that specific words are not simply lost from one's repertoire. Instead the data provide support for the functional independence of verbal operants.

Limitations

The current findings provide evidence that individuals with dementia have more difficulty with verbal behavior than those without dementia. Furthermore, the results give support to Skinner's analysis of functionally independent verbal operants. However, these findings are not without limitations. The current sample size was small and the range of DRS-2 scores was somewhat restricted. According to the DRS-2 categorizations, most participants were in the minimally impaired or below average range, and these two categories were on either side of the cut-off for group placement. Unfortunately, two of the three individuals who did not complete the study met criteria for moderately impaired and severely impaired, which were categorizations where the sample had only five and zero participants, respectively. Including these individuals in the study would have diversified the range of abilities represented.

Another limitation is that ceiling effects on some measures made interpreting the data using statistical analyses difficult or impossible. However, these data still provide useful information for the development of future assessment methods. Because the participants in this sample were sophisticated speakers, it may have been more useful to sample complex verbal behavior. The current assessment methods evaluated elementary verbal operants but did not require individuals to engage in complex verbal behavior such 50 as using full sentences, describing events, or having a natural conversation. The assessment methods from this study may be more suitable for more severely impaired individuals, whereas complex verbal behavior assessments might be more appropriate for evaluating individuals with mild or moderate impairments. Complex verbal behavior assessments might allow for evaluation of everyday, verbal behavior difficulties experienced by sophisticated speakers. Another assessment concern is that the procedures were not similar to real life verbal behavior occurrences of older adults. Once individuals acquire verbal behavior or leave formal schooling, it is rare to name line drawings or provide responses based on a description of an item and have this type of responding reinforced (Skinner 1957/1992). Therefore, these assessments may not capture the typical difficulties that occur in older adulthood. Assessments that evaluate naturally occurring verbal behavior may be more useful for determining strengths and weaknesses. At least one study reported functional analyses of language in children with developmental disabilities (Lerman et al., 2005), and a similar approach might be appropriate for assessing verbal behavior in older adults.

A final concern with respect to the assessment methods is the difference in reinforcement under the various testing conditions. During the two sessions, investigators did not reinforce responses, except during the Mand, Mand Selection, and 3D Mand Assessments. Although this could have resulted in increased responding under these conditions, the results did not show individuals performed differently based on the reinforcement provided by the evaluator. This might not be surprising given the long and complex verbal behavior history of the participants compared to the relatively brief amount of time spent participating in the study. 51 Future Research

To the author's knowledge, this is the first study to evaluate verbal functioning in older adults using assessment strategies based on Skinner's conceptualization of functional verbal operants. As a result, there are several suggestions for future research. Future investigations might improve the assessment methods in a number of ways. Because most individuals performed very well on the Selection and Echoic Assessments, one might consider only sampling these behaviors to ensure they are in the repertoire, but then focus evaluations on more complex verbal behavior that might prove to be more prone to age-related degradation. It seems reasonable to retain an assessment of tacts, intraverbals, and mands in an assessment battery, as there was performance variability among participants on these forms of verbal behavior. In order to develop future assessment methods, one might consider using ideas from researchers and practitioners who evaluate verbal behavior of children (e.g., Greer & Keohane, 2006; Lerman et al., 2005; Sundberg, n.d.). For example, Sundberg has developed the Verbal Behavior Milestones Assessment and Placement Program (VB-MAPP), an extensive system for evaluating verbal behavior of young children with language delays, and this measure might provide insight for evaluators of verbal behavior in older adults (Sundberg, n.d.). Researchers might also develop assessments out of the stimulus and response categories proposed by Baker and colleagues (2008) in their behavioral taxonomy of aphasia. As noted previously, 40 potential deficits could be evaluated using this taxonomy, and the current research sampled only a subset of these categories. Therefore, evaluating other potential deficit areas might provide additional information about categories that are sensitive to language impairments in older adults.

Researchers might also consider altering the task presentations for verbal behavior assessments. The current study used two mand conditions, filling in the missing item 52

from a scene and completing a task with three-dimensional objects, but these are not the only two options for assessing mands. Researchers should develop alternate procedures for creating establishing operations and/or for observing mands in the natural environment. Intraverbal assessments provide a good platform for developing alternate approaches to assessment. Antecedents to intraverbals are verbal responses with no point- to-point correspondence or formal similarity to the response. This antecedent condition is quite broad, leaving many options for intraverbal testing conditions. Many educational tasks require intraverbal responding, thus one might look to educators or curriculum developers for assistance in developing intraverbal assessment methods. In creating alternative assessment methods, researchers might also consider using different stimuli across the various verbal operant categories. The current study intentionally used the same 30 stimuli for most assessments because of the goal of demonstrating functional independence across various verbal operants, even when requiring topographically identical responses. However, now that there are data to support functional independence using the same stimuli, it might be helpful to use stimuli of similar difficulty level but that are not identical. Performance improved from session one to session two for individuals in this study, indicating practice effects. Moreover, anecdotal evidence during the first session indicated that several participants learned that the same 30 stimuli appeared on each assessment, because many participants commented on the similarity of the responses required across assessments. For the current study, the change in performance from session one to session two was not problematic, as the relative strengths and weaknesses remained consistent. However, practice effects would be problematic if one wanted to track behavior change over time or evaluate behavior change due to an intervention; hence the recommendation to vary the stimuli presented across assessments measures. 53

In addition to examining the assessment procedures, future research should evaluate interventions developed from verbal behavior assessments similar to those used in the current study. By identifying strengths and weaknesses with these assessment measures, practitioners could develop intervention methods that build on the verbal skills and accommodate the language-related deficits of older adults. One might consider modeling the interventions after those used by early intervention programs with children who have language delays (e.g., Greer & Keohane, 2006). Although early intervention programs could assist in the initial development of assessment and intervention methods, one needs to be mindful of the uniqueness of the decline in functioning with older adults. Declining skills require interventions aimed to maintain current functioning and/or slow the decline, rather than treatments intended to train novel skills. Once interventions are developed based on the results of verbal behavior assessments, they need to be compared to those created without such information to determine whether the additional information actually enhances intervention effectiveness.

Conclusion

Older adults make up a substantial portion of the population and this number continues to grow (U.S. Census Bureau, 2008). Many older adults with and without dementia experience declines in language functioning (Benson & Ardila, 1996; Kempler, 2005). Traditional assessments of language functioning attribute performance to underlying, theoretical constructs and typically evaluate the topography of the response. With these strategies, individuals assume the presence or absence of skills in an individual's repertoire and largely ignore environmental conditions that may influence verbal behavior. Skinner's analysis of verbal behavior allows for assessing the function of verbal behavior, which provides a way to evaluate the conditions under which verbal 54 behavior is more or less likely to occur. The current findings revealed that elderly participants with dementia performed more poorly on the Tact, Intra verbal, and Mand Assessments than those without dementia, with differences on the Tact Assessment reaching significant levels. These findings suggest that tacts are the most sensitive measure of language degradation associated with dementia and that intraverbals and mands may also be sensitive to such functional decline. Because of their sensitivity, measures of tacts, intraverbals, and mands should be included in future assessment batteries. Moreover, results supported Skinner's analysis of functionally independent verbal operants. That is, data analyses revealed that participants performed inconsistently across assessments even though response requirements were topographically identical across assessment conditions. Therefore, it appears that the function of verbal behavior has more influence on responding than the topography of the verbal responses. Using verbal behavior assessments similar to those used in this study may help identify specific verbal behavior strengths and deficits of older adults, and the added precision of these assessments may allow for the development of individualized interventions that target specific verbal operant categories. 55

REFERENCES

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental

disorders (Revised 4th ed.) Washington, DC: Author.

Baker, J. C, LeBlanc, L. A., & Raetz, P. B. (2008). A behavioral conceptualization of

aphasia. The analysis of verbal behavior, 24, 147-158.

Benson, D. F., & Ardila, A. (1996). Aphasia: A clinical perspective. New York: Oxford

University Press.

Bondy, A., & Frost, L. (2001). The picture exchange system. Behavior Modification, 25,

125-144. doi: 10.1177/0145445501255004

Bondy, A., Tincani, M., & Frost, L. (2004). Multiply controlled verbal operants: An

analysis and extension to the picture exchange system. The Behavior Analyst, 27, 247-

261. Retrieved from http://www.ncbi.nlm.nih.gov.libraryl.unmc.edu:

2048/pmc/issues/182015/

Bourgeois, M. S. & Hickey, E. M. (2009). Dementia: From diagnosis to management - a functional approach. New York: Psychology Press.

Davis, G. A. (2006). Aphasiology: Disorders and clinical practice. Allyn and Bacon:

Needham Heights.

Didden, R. (2007). Functional analysis methodology in developmental disabilities. In P.

Sturmey (Ed.), Functional analysis in clinical treatment (pp. 65-86). Burlington, MA:

Academic Press. 56 Dymond, S., O'Hora, D., Whelan, R., & O'Donovan, A. (2006). Citation analysis of

Skinner's Verbal Behavior. 1984-2004. The Behavior Analyst, 29, 75-88. Retrieved

from http://www .ncbi .nlm .nih .gov/pmc/articles/PMC2223169/pdf/bhan-29-01 -75 .pdf

Fisher, N. J., Tierney, M C, Snow, W. G., & Szalai, J. P. (1999). Odd/Even short forms

of the Boston Naming Test: Preliminary geriatric norms. The Clinical

Neuropsychologist, 13, 359-364. doi: 10.1076/clin.l3.3.359.1742

Field, A. (2009). Discovering statistics using SPSS (3rd ed.). London: Sage Publications.

Greer, R. D., & Keohane, D. (2006). The evolution of verbal behavior in children, The

Journal of Speech-Language Pathology and Applied Behavior Analysis, 1, 111 -140.

Retrieved from http://www.thefreelibrary.com/The+evolution+of+

verbal+behavior+in+children .-a0170157268

Guess, D. & Baer, D. M. (1973). An analysis of individual differences in generalization

between receptive and productive language in retarded children. Journal of Applied

Behavior Analysis, 6, 311-329. doi: 10.1901/jaba.l973.6-311

Jurica, P. J., Leitten, C. L., & Mattis, S. (2001). DRS-2: Dementia Rating Scale-2.

Psychological Assessment Resources, Inc.: Lutz.

Kempler, D. (2005). Neurocognitive disorders in aging. Thousand Oaks: Sage

Publications.

Kertesz, A. (1994). Language deterioration in dementia. In Emory V.O.B. & Oxman, T.

E. (Eds.), Dementia: Presentations, differential diagnosis, and nosology (pp. 123-138).

Baltimore: Johns Hopkins University Press. 57 Lerman, D. C, Parten, M., Addison, L. R., Vorndran, C. ML, Volkert, V. M., & Kodak, T.

(2005). A methodology for assessing the functions of emerging speech in children with

developmental disabilities, 38, 303-316. doi: 10.1901/jaba.2005.106-04

Lezak, M. D., Howieson, D. B., & Loring, D. W. (2004). Neuropsychological Assessment

(4th ed). New York: Oxford University Press.

Longman advanced American dictionary. (2000). Harlow, Essex, England: Pearson

Education Limited.

Michael, J. (1988). Establishing operations and the mand. The Analysis of Verbal

Behavior, 6,3-9.

Mitrushina, M., Boone, K. B., Razani, J., & D'Elia, L. F. (2005). Handbook of normative

data for neuropsychological assessment (2nd ed.). New York: Oxford University Press.

Rohrer, J. D., Knight, W. D., Warren, J. E., Fox, N. C, Rossor, M. N., & Warren, J. D.

(2008). Word-finding difficulty: a clinical analysis of the progressive aphasia. Brain,

131, 8-38. doi: 10.1093/brain/awm251

Sautter, R. A. & LeBlanc, L. A. (2006). Empirical applications of Skinner's analysis of

verbal behavior in humans. The Analysis of Verbal Behavior, 22, 35-48. Retrieved

from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2774593/pdf/anvb-22-01-35.pdf

Schmidt, K. S., Mattis, P. J., Adams, J., Nestor, P. (2005). Alternate-form reliability of

the Dementia Rating Scale-2. Archives of Neuropsychology, 20, 435-441. doi:

10.1016/j.acn.2004.09.011

Sidman, M. (1971). The behavioral analysis of aphasia. Journal of Psychiatric Research,

8,413-422. doi: 10.1016/0022-3956(71)90034-3 58 Sidman, M., Stoddard, T., Mohr, J. P., & Leicester, J. (1971). Behavioral studies of

aphasia: Methods of investigation and analysis. Neuropsychologic!, 9, 119-140. doi:

10.1016/0028-3932(71 )90038-8

Skinner, B. F. (1974). About Behaviorism. Random House, Inc.: New York.

Skinner, B. V. (1965). Science and Human Behavior. Free Press: New York.

Skinner, B. F. (1992). Verbal Behavior. Copley Publishing Group: Acton, MA.

Sundberg, M. L. (n.d.). VB-MAPP. Retrieved from http://www.marksundberg.com/ vb-

mapp.htm

Sundberg, M. L. & Michael, J. (2001). The benefits of Skinner's analysis of verbal

behavior for children with autism. Behavior Modification, 25, 698-724. doi:

10.1177/O145445501255003

U.S. Census Bureau (2001). The 65 years and over population: 2000. Retrieved from

http://www.census.gov/population/www/cen2000/briefs/.

U.S. Census Bureau (2003). Disability status: 2000. Retrieved from

http://www.census.gov/population/www/cen2000/briefs/.

U.S. Census Bureau (2008). 2008 national population projections. Retrieved from

http://www .census .gov/population/www/projections/ 2008 projections .html. 59

Appendix A

Demographic Questionnaire

1 How old are you?

2 Are you male or female?

3 What is your educational history? (check one)

Completed grade High school or equivalent Some College Credits Vocational/technical school (2 year degree) Bachelor's Degree Master's Degree Doctoral Degree Professional Degree (MD, JD, etc) Other

4 For most of your adult life, did you work outside the home?

Yes No

If yes, what did you do for a living?

5 How would you classify yourself? (check one)

Arab Asian/Pacific Islander Black Caucasian/White Hispanic Indigenous or Aboriginal Latino Multiracial Would rather not say 60 Other

6 Is English your first language? Yes No If not, what is?

How long have you spoke English?

What is your living situation? (check 7 one)

Independent with no assistance Independent but with help from family/friends Independent but with hired help that comes to the home Assisted living with some meals and light cleaning provided Assisted living with all meals and household duties provided I Nursing home

Do you have any diagnosed cognitive deficits (e.g.,

8 dementia)?

Yes No

If yes, what?

Do you have vision or hearing

9 difficulties?

Yes No

YeIf yess , do you havNoe glasses or hearing aids on/in? 61 Appendix B Selection Assessment Form

Instructions: In this part, I am going to show you groups of pictures. Then, I will ask you to point to a picture from the group. For each item say: Point to the . (present array) Allow 5 seconds for a response. If the participant does not respond, move to the next item. Circle the participant's response and record the time at which the participant responds.

No Time 1 Rhinoceros Whistle Trellis response No Time 2 Saw Accordion Pyramid response No Time 3 Whistle Igloo Beaver response No Time 4 Dominos Pyramid Racquet response No Time 5 Bed Trellis Comb response No Time 6 Globe Rhinoceros Hanger response No Time 7 Dart Pretzel Unicorn response No Time 8 Pencil Comb Pelican response No Time 9 Octopus Scroll Escalator response No Time 10 Latch Sphinx Pretzel response No Time 11 Dominos Helicopter Rhinoceros response No Time 12 Hanger Hammock Scroll response No Time 13 Protractor Racquet Pelican response No Time 14 Whistle Dominos Latch response No Time 15 Volcano Unicorn Beaver response No Time 16 Octopus Protractor Helicopter response No Time 17 Volcano Igloo Asparagus response No Time 18 Sphinx Dart Pencil response No Time 19 Accordion Bed Octopus response No Time 20 Volcano Beaver Dart response No Time 21 Pelican Saw Globe response No Time 22 Asparagus Hammock Camel response No Time 23 Globe Escalator Hammock response No Time 24 Comb Asparagus Helicopter response No Time 25 Unicorn Pretzel Trellis response No Time 26 Protractor Camel Igloo response No Time 27 Saw Scroll Escalator response No Time 28 Sphinx Racquet Hanger response No Time 29 Accordion Bed Camel response No Time 30 Pyramid Pencil Latch response 63

Appendix C

Echoic Assessment Form

Instructions: In this part, I am going to say words. I want you to repeat each one after me. Allow 5 seconds for a response. Move to the next item if the participant does not respond. On each item give instruction: Say .

Response Given if Item Correct Incorrect Incorrect Time 1 Whistle

2 Pencil

3 Sphinx

4 Accordion

5 Latch

6 Hammock

7 Volcano

8 Hanger

9 Dominos

10 Pyramid

11 Beaver

12 Scroll

13 Comb

14 Bed

15 Escalator

16 Helicopter

17 Rhinoceros

18 Igloo 64

19 Camel [ | 1

20 Racquet | | 1

21 Saw | | 1

22 Pelican [ 1 [

23 Octopus | | |

24 Protractor | | |

25 Pretzel 1 1 I

26 Dart 1 1 I

27 Trellis 1 1 1

28 Asparagus | 1 1

29 Globe | | 1

30 Unicorn I I 1 65

Appendix D Tact Assessment Form

Instructions: In this part, I am going to show you pictures of items. I want you to name each item. For each item say: Name this, (present picture) Allow 5 seconds for a response. If the participant does not respond, turn to the next item.

Response Given if Item Correct Incorrect Incorrect Time 1 Comb 1 1 1

2 Protractor | [ [

3 Pencil 1 1 1

4 Dart 1 I 1

5 Sphinx | | |

6 Asparagus [ | |

7 Whistle | 1 I

8 Escalator [ | | |

9 Saw | 1 1

10 Pyramid | | |

11 Helicopter | 1 |

12 Hammock | | |

13 Beaver | | |

14 Accordion | 1 1

15 Rhinoceros [ | |

16 Racquet | 1 |

17 Volcano 18 Scroll

19 Igloo

20 Camel

21 Hanger

22 Octopus

23 Trellis

24 Pelican

25 Unicorn

26 Globe

27 Latch

28 Dominos

29 Pretzel

30 Bed 67

Appendix E

Intraverbal Assessment Form

Instructions: In this part, I am going to describe a word or item. I want you to tell me what I am describing. Read descriptoin for each item. Allow 5 sec to initiate response and 10 sec to complete it. Description Item C I Response Given if Incorrect Time 1 This is a type of aircraft with large metal blades on top that spin very fast to make it fly Helicopter 2 This is a large water bird that catches fish for food and stores them in a deep bag of skin under its beak Pelican 3 This is an imaginary animal like a white horse with a long straight horn growing on its head Unicorn 4 This is a house made from blocks of hard snow or ice Igloo 5 This is an animal that lives in the ocean with eight tentacles, but no skeleton Octopus 6 This is an instrument usually in the shape of a half-circle, used for measuring and drawing angles Protractor

7 This is a large heavy African or Asian animal with thick skin and either one or two horns on its nose Rhinoceros 8 These are small pieces of wood or plastic with a different number of spots on each half of its top side, used in playing a game. When they are standing up and one is knocked over, all the others fall down too. Dominos 9 This is a small pointed object that is thrown or shot during a game that uses a circular board with a bulls eye Dart 10 This is a small metal or plastic object used to keep doors, gates, windows, etc closed Latch Description Item 11 This is a North American animal that has thick fur, a wide flat tail, and cuts down trees with its teeth Beaver 12 This is a tool that has a flat metal blade with an edge that has teeth, used for cutting wood Saw 13 This is a round object with a map of he Earth on it Globe 14 This is a long piece of paper that can be rolled up, and is often used as an official document, especially in past times Scroll 15 This is an Egyptian image of a lion with a human head, lying down Sphinx 16 This is a flat piece of plastic or metal with a row of teeth on one side, used to make your hair look • neat C?mt> 17 This is a hard salty type of bread baked in the shape of a stick or loose knot i Pretzel i8 this is a mountain with a large hole at the top, through which lava sometimes rises into the air ^Volcano 19,This is a large stone building with four triangular walls that slope in to a point at the top, found 'especially in Egypt and Central America Pyramid 20 This is a set of stairs that move and carry people • Ifrom one level within a building to another Escalator 21 ;This is a green vegetable shaped like a small I stick with a point at one end Asparagus Description Item 22 This is a large desert animal with a long neck and one or two humps on its back Camel 23 This is a curved piece of wood, plastic, or metal with a hook on it, on which you put your clothes. It is put on a closet rod. Hanger 24 This is a frame made of long narrow pieces of wood that cross each other, used to support climbing plants Trellis 25 This is a piece of equipment used for hitting the ball in games such as tennis, consisting of a stick with netted wire in a round frame Racquet 26 This is a small object that produces a high shrill sound when you blow into it and is used by referees Whistle 27 This is a musical instrument that you pull in and out to produce sounds while pushing buttons on one side to produce different notes Accordian 28 This is a large piece of material or a net that you can sleep on, that hangs between two trees or poles Hammock 29 This is a narrow pointed wooden instrument, used for writing or drawing, containing a thin stick of black or colored substance and an eraser Pencil 30 This is a piece of furniture for sleeping Bed 70 Appendix F

Mand Assessment Form

Instructions: In this part, I am going to show you sets of two pictures. The first picture will be a complete picture. The second picture will be the exact same picture, but one piece will be missing. I want you to tell me what you need to complete the picture. Once you do, I will give you the item to place on the picture to complete it. For each item: Look at the first picture of this set. (wait 3 sec) Look at the second. What do you need to complete this picture? (present picture 1, cover it and present picture 2) Allow 5 sec to initiate response and 10 sec to complete it. (if correct, give missing item; if incorrect, move to mand selection) Mand selection: Point to what you need to complete this picture, (present array) Allow 5 sec to initiate response and 10 sec to complete it. (if correct, give missing item; if incorrect move to next item on mand assessment)

Response Given if No Item C I Incorrect Time C I Response Time 1 Bed | || | Select? 2 Saw | || | Select? 3 Dart | | | | Select? 4 Protractor | | | | Select? 5 Hammock | | | | Select? 6 Hanger | J | | Select? 7 Pencil | || | Select? 8 Pyramid | [ | | Select? 9 Volcano | | | | Select? 10 Racquet | | | | Select? 11 Beaver | | | | Select? 12 Igloo | || | Select? 13 Trellis | | | | Select? 14 Escalator | [ | | Select? 15 Dominos | | | | Select? 16 Unicorn | | | | Select? 17 Accordion | | | | Select? 71

18 Rhinoceros | | | | Select? 19 Whistle | | | | Select? 20 Latch | || | Select? 21 Camel | | | | Select? 22 Asparagus | | | | Select? 23 Octopus | || | Select? 24 Pelican | | | | Select? 25 Globe | | | | Select? 26 Sphinx | || | Select? 27 Pretzel | | | | Select? 28 Comb | || | Select? 29 Helicopter | | 1 | Select? 30 Scroll | || | Select? 72

Appendix G

3D Mand Assessment Form

Instructions: In this part I am going to give you jobs to do and I want you to complete each job. If you are missing any items that you need to complete the job, tell me what you need and I will give it to you. Present each job, Allow 5 seconds to initiate a response and 10 seconds to complete the response.

Job Item C I Response Given if Incorrect Time 1 For this job, I want you to write down your name, Tell me if you need anything else to complete the job. Paper

2 For this job, I want you to put this one in the wood so that it looks like this one. Tell me if you need anything else to complete the job, Screwdriver

3 For this job, I want you to cut out this circle. Tell me if you need anything else to complete the job. Scissors

4 For this job, I want you to tell me how long this line is, in inches. Tell me if you need anything else to complete the job. Ruler

5 For this job, I want you to open the bag and hand me the item inside, Tell me if you need anything else to complete the job, Key 73

Appendix H

Verbal Behavior Assessment Raw Scores During Session One and Session Two

Pi Group Selection Echoic Tact Intraverbal Mand Mand Selection 3D Mand 1 2 1 2 1 2 1 2 1 2 1 2 1 2 PI D 30 30 30 30 29 30 29 30 23 28 30 30 5 5 P2 D 30 30 30 30 29 29 24 28 27 26 30 30 5 4 P4 C 30 30 30 30 30 29 30 30 30 29 30 30 5 5 P5 D 30 30 30 30 28 30 26 27 20 28 30 30 5 5 P6 D 30 30 30 30 27 29 24 27 25 25 29 30 4 5 P7 D 30 30 29 30 21 29 23 27 25 27 30 30 5 5 P8 C 30 30 30 30 30 30 29 30 30 30 30 30 5 5 P9 D 30 30 30 30 22 27 19 25 19 27 30 30 4 4 Pll D 30 30 30 29 28 30 20 26 30 30 30 30 5 5 P12 D 30 30 30 30 29 28 24 24 7 15 27 26 5 5 P13 C 30 30 30 30 29 29 29 30 26 30 30 30 5 5 P14 D 30 30 30 30 24 24 23 23 18 19 28 30 4 5 P15 C 30 30 29 30 28 29 24 28 15 26 30 30 4 4 PI 6 C 30 30 30 30 28 30 29 29 27 29 30 30 5 5 P17 C 30 30 30 30 30 30 30 30 29 30 30 30 5 5 P18 D 30 30 30 29 16 18 15 19 15 17 27 29 4 3 P19 C 30 30 30 30 30 30 25 29 29 30 30 30 4 5 P20 D 30 30 30 30 29 27 27 28 22 26 30 30 5 5 P21 C 30 30 30 30 30 30 27 30 28 30 30 30 5 5 P22 C 30 30 30 30 30 30 20 29 21 30 30 30 5 5 P23 c 30 30 30 30 26 28 17 25 13 19 30 30 5 5 P24 c 30 30 30 30 29 29 30 30 29 30 30 30 5 5 P25 D 30 30 30 30 28 30 30 30 27 30 30 30 5 5 P26 C 30 30 30 30 30 30 26 29 28 30 30 30 5 5 P27 D 30 30 29 30 29 30 25 29 16 22 30 30 5 5 P28 C 30 30 30 30 30 30 24 28 26 29 30 30 5 5 P29 D 30 30 30 30 29 30 30 30 26 30 30 30 5 5 P30 C 30 30 29 29 26 30 19 27 16 26 29 30 5 5 P32 D 30 30 30 30 29 29 24 22 17 20 30 29 4 4 P33 D 30 30 30 30 29 30 28 28 25 28 30 30 5 5 P34 C 30 30 30 30 28 30 30 30 30 30 30 30 5 5 T=participants 74

Appendix I

HSIRB Approval Letter WESTERN MICHIGAN UNIVERSITY

Human Subjects Institutional Review Board

Date: June 2,2009

To: Wayne Fuqua, Principal Investigator Amy Gross, Student Investigator for dissertation

From: Amy Naugle, Ph.D., S^r^W (\|ftjM^

Re: HSIRB Project Number: 09-05-08

This letter will serve as confirmation that your research project titled "Evaluation of Verbal Behavior in Older Adults" has been approved under the full category of review by the Human Subjects Institutional Review Board. The conditions and duration of this approval are specified in the Policies of Western Michigan University. You may now begin to implement the research as described in the application.

Please note that you may only conduct this research exactly in the form it was approved. You must seek specific board approval for any changes in this project You must also seek reapproval if the project extends beyond the termination date noted below. In addition if there are any unanticipated adverse reactions or unanticipated events associated with the conduct of this research, you should immediately suspend the project and contact the Chair of the HSIRB for consultation.

The Board wishes you success in the pursuit of your research goals.

Approval Termination: May 20,2010

WalwoDd Hall, Kalamazoo, Ml 49008-5456 PHONE: (269)387-8293 FAX: (269)387-8276