This article was downloaded by: [University of Otago] On: 15 July 2015, At: 09:23 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place, London, SW1P 1WG

The Clinical Neuropsychologist Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ntcn20

Factor Structure and Construct Validity of the Behavioral Dyscontrol Scale-II ab

Robert D. Shura , Jared A. Rowland

cd

ab

& Ruth E. Yoash-Gantz

a

Mid-Atlantic Mental Illness Research, Education, and Clinical Center (MA-MIRECC), W.G. ‘Bill’ Hefner Veterans Affairs Medical Center, Salisbury, NC 28144, USA b

Department of Psychiatry and Behavioral Sciences, Wake Forest School of Medicine, Winston-Salem, NC 27157, USA c

MA-MIRECC, Research & Education Service Line, W.G. ‘Bill’ Hefner VA Medical Center, Salisbury, NC 28144, USA d

Click for updates

Department of Psychiatry and Behavioral Sciences, Department of Neurobiology & Anatomy, Wake Forest School of Medicine, Winston-Salem, NC 27157, USA Published online: 04 Feb 2015.

To cite this article: Robert D. Shura, Jared A. Rowland & Ruth E. Yoash-Gantz (2015) Factor Structure and Construct Validity of the Behavioral Dyscontrol Scale-II, The Clinical Neuropsychologist, 29:1, 82-100, DOI: 10.1080/13854046.2015.1007169 To link to this article: http://dx.doi.org/10.1080/13854046.2015.1007169

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Downloaded by [University of Otago] at 09:23 15 July 2015

Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

The Clinical Neuropsychologist, 2015 Vol. 29, No. 1, 82–100, http://dx.doi.org/10.1080/13854046.2015.1007169

Factor Structure and Construct Validity of the Behavioral Dyscontrol Scale-II Robert D. Shura1,2, Jared A. Rowland3,4, and Ruth E. Yoash-Gantz1,2

Downloaded by [University of Otago] at 09:23 15 July 2015

1

Mid-Atlantic Mental Illness Research, Education, and Clinical Center (MA-MIRECC), W.G. ‘Bill’ Hefner Veterans Affairs Medical Center, Salisbury, NC 28144, USA 2 Department of Psychiatry and Behavioral Sciences, Wake Forest School of Medicine, Winston-Salem, NC 27157, USA 3 MA-MIRECC, Research & Education Service Line, W.G. ‘Bill’ Hefner VA Medical Center, Salisbury, NC 28144, USA 4 Department of Psychiatry and Behavioral Sciences, Department of Neurobiology & Anatomy, Wake Forest School of Medicine, Winston-Salem, NC 27157, USA The Behavioral Dyscontrol Scale-II (BDS-II) was developed as an improved scoring method to the original BDS, which was designed to evaluate the capacity for independent regulation of behavior and attention. The purpose of this study was to evaluate the factor structure and construct validity of the BDS-II, which had not been adequately re-examined since the development of the new scoring system. In a sample of 164 Veterans with a mean age of 35 years, exploratory factor analysis was used to evaluate BDS-II latent factor structure. Correlations and regressions were used to explore validity against 22 psychometrically sound neurocognitive measures across seven neurocognitive domains of sensation, motor output, processing speed, attention, visual-spatial reasoning, memory, and executive functions. Factor analysis found a two-factor solution for this sample which explained 41% of the variance in the model. Validity analyses found significant correlations among the BDS-II scores and all other cognitive domains except sensation and language (which was not evaluated). Hierarchical regressions revealed that PASAT performance was strongly associated with all three BDS-II scores; dominant hand Finger Tapping Test was also associated with the Total score and Factor 1, and CPT-II Commissions was also associated with Factor 2. These results suggest the BDS-II is both a general test of cerebral functioning, and a more specific test of working memory, motor output, and impulsivity. The BDS-II may therefore show utility with younger populations for measuring frontal lobe abilities and might be very sensitive to neurological injury. Keywords: Behavioral Dyscontrol Scale-II; BDS-II; Executive function; Veteran; Validity; Factor.

INTRODUCTION The Behavioral Dyscontrol Scale (BDS; Grigsby & Kaye, 1996) is one of the few measures to quantify dynamic motor behavior, a subdomain of executive function that is thought to involve inhibiting automatic behaviors, detecting inconsistent behaviors, and correcting behavioral action in accordance with error detection (Garavan, Ross, Murphy, Roche, & Stein, 2002). The BDS has shown utility in a Address correspondence to: Robert D. Shura, Hefner VAMC, 11M-2 MH/BS, 1601 Brenner Ave., Salisbury, NC 28144, USA. E-mail: [email protected] (Received 11 September 2014; accepted 9 January 2015)

© 2015 Taylor & Francis

Downloaded by [University of Otago] at 09:23 15 July 2015

BDS-II STRUCTURE AND VALIDITY

83

variety of contexts, including predicting functional independence (Grigsby, Kaye, Kowalsky, & Kramer, 2002a, 2002b; Kaye, Grigsby, Robbins, & Korzun, 1990; Suchy, Blint, & Osmon, 1997), differentiating frontal from non-frontal lesions in traumatic brain injury (TBI) patients (Leahy, Suchy, Sweet, & Lam, 2003), and distinguishing normal aging from dementia and mild cognitive impairment (Belanger et al., 2005; Hall & Harvey, 2008). The BDS was developed as a measure of executive function evaluating the capacity for independent regulation of behavior and attention, with the aim of providing an ecologically valid assessment of functional independence. The initial normative and validity studies focused on geriatric samples, creating a ceiling effect when used with younger individuals. The BDS-II scoring system was subsequently designed which increases the ceiling in order to improve the normality of scores, improve variability in higher-functioning individuals, and more effectively discriminate high-functioning individuals. Leahy and colleagues (2003) found the BDS-II scoring system significantly improved the negative skewness present when using the original BDS scoring system in a sample ranging in age from 18 to 71. A more recent study contrasted the BDS and BDS-II scoring systems in non-elderly Veterans (mean age = 35) and found the BDS-II had significantly improved psychometric properties and internal reliability (Shura, Rowland, & Yoash-Gantz, 2014). Criterion validity of the new scoring system has been demonstrated through prediction of functional outcome (Suchy et al., 1997), and lesion location and discrimination of mild TBI (Leahy et al., 2003; Suchy, Leahy, Sweet, & Lam, 2003). Although these studies suggest the new scoring system improves the validity and reliability of the BDS, effects on the factor structure have not been examined. Original analyses of the BDS found three factors: ‘ability to use an intention to guide behavior’ (Items 1, 2, 5, and 6), ‘aspects of cognition that have been associated with frontal lobe functioning’ (Items 3 and 4), and ‘capacity for inhibition’ (Items 7, 8, and 9; Grigsby, Kaye, & Robbins, 1992, pp. 889–890). The initial study included a principal component analysis (PCA) of the BDS in a sample of 229 Veterans aged 60 to 105. The authors used the scree method to determine number of factors and an orthogonal Varimax rotation for the final solution. A later study by Ecklund-Johnson, Miller, and Sweet (2004) used confirmatory factor analysis (CFA) on the BDS scores of a mixed inpatient/outpatient sample aged 16 to 92 years. The authors tested four models and found that none adequately fit the data, although when allowing for inter-factor correlations, their results supported the original three factors. Interestingly, the authors separated out a younger sample (less than 65 years old), but were unable to complete the CFA due to non-normality of data. This was likely due to the ceiling effect of the original BDS scoring system in younger populations (Shura et al., 2014). An additional finding from Ecklund-Johnson et al. (2004) was the high inter-correlations of the factors, suggesting that oblique rotations should be used instead of orthogonal rotations in exploratory factor analysis (EFA), which contrasted with the initial study by Grigsby et al. (1992). This factor structure has not been re-examined with a younger population since the creation of the BDS-II scoring system. Thus, the primary aim of this study is to examine the latent factor structure of the BDS-II in a sample of non-elderly participants. A number of studies examined aspects of validity of the original BDS factors using varying methods and populations. A criterion validity study with the three

Downloaded by [University of Otago] at 09:23 15 July 2015

84

ROBERT D. SHURA ET AL.

original BDS factors used a sample of 68 participants aged 31 to 91 (Suchy et al., 1997). The Stroop Color and Word Test Part C and Mini-Mental State Exam were significant predictors across all three factors, especially Factor 1 (items 1, 2, 5, and 6, which they labeled Motor Programming). The Wechsler Adult Intelligence ScaleRevised (WAIS-R) Similarities subtest was an additional predictor of Factor 2 (items 7, 8, 9, which they labeled Fluid Intelligence), and the Milwaukee Card Sorting Test was an additional predictor of Factor 3 (items 3 and 4, which they labeled Environmental Independence). Additionally, the authors stated that the lack of relationship between the BDS factors and WAIS-R Information subtest provided evidence for divergent validity. Another study (Diesfeldt, 2004) used CFA and item response theory to confirm the BDS construct validity in a sample of 693 community adults aged 50 to 94. The BDS loaded with three other measures of executive functioning (Expanded Mental Control Test, Category Fluency, and Graphical Sequences) compared to a test of Orientation and Delayed Picture Recognition, which loaded on an episodic memory factor. The two factors were highly correlated, suggesting that they were not independent and failed to show divergent validity for the BDS (Diesfeldt, 2004). A novel finding from that study was that the BDS Item 9 violated the criteria of double monotonicity, and eliminating this item improved the measure. However, there was no difference in validity analyses when the item was dropped. Limitations of these two studies include the older age of participants, the larger age range of participants, and the small number of executive function tests used. Only one study has examined the validity of the three original BDS factors using the updated BDS-II scoring system. In a sample of 49 post-acute TBI participants aged 18 to 71, Suchy and colleagues (2003) used regression analyses to examine the relationships between the BDS factors calculated using the BDS-II scoring system and three other measures of executive functioning (Trail Making Test Part B, Controlled Oral Word Association Test, and the Stroop Color-Word Interference Test), three measures of processing speed (Trail Making Test Part A, Stroop Word, and Stroop Color), and the Grooved Pegboard as a measure of manual dexterity. The authors found that the BDS Factor 3 (Fluid Intelligence) added significant variance above the other measures in classifying frontal versus non-frontal lesion patients, and Factor 1 (Motor Programming) added significant variance in classifying mild to moderate versus severe TBI patients. Again, the small number of executive measures used and small sample size were limitations to this study. Of note, two studies were completed using an electronic version based on the original BDS (Suchy, Derbidge, & Cope, 2005; Suchy, Eastvold, Whittaker, & Strassberg, 2007), which provided initial validation of the instrument; however, the electronic version was compared to the original scoring system, does not directly reflect the original BDS items, and is less accessible than the original version. In summary, previous studies used samples with large ranges of participant age. Grigsby and Kaye (1996) noted that there appears to be a decline in BDS performance beginning around age 60; however, in participants below age 60 there was no relationship between BDS performance and age. Therefore, collapsing across large age groups might confound validity studies. Thus, the second aim of this study is to examine the construct validity of BDS-II factors using measures of numerous cognitive domains that are common to clinical neuropsychological practice, and a sample younger than the age at which decline was speculated to occur in the BDS.

BDS-II STRUCTURE AND VALIDITY

85

METHOD The prospective studies from which these data were drawn were reviewed and approved by the affiliated Institutional Review Board. The welfare and privacy of human participants was protected and maintained. Voluntary verbal and written informed consent was obtained prior to initiation of any study activities. Statistical analyses were completed using SPSS 21 and a freely available Monte Carlo program for parallel analysis (Watkins, 2000).

Downloaded by [University of Otago] at 09:23 15 July 2015

Participants The current study included 164 participants selected from a larger multi-site study of cognitive functioning in post-deployment veterans being conducted by the MidAtlantic Mental Illness Research, Education, and Clinical Center (MA-MIRECC). Inclusion criteria include military service since 10/1/2001. Exclusion criteria include combat exposure prior to 1985, neuropsychological evaluation in the past 6 months, presence of psychotic symptoms, presence of current substance abuse or dependence, history of pre-deployment non-military related PTSD, and history of moderate to severe TBI prior to or since deployment. Participants in the current study participated at a single site. Tests were administered by a North Carolina licensed master’s-level psychological associate, doctorate-level psychologist with training in neuropsychology, or neuropsychology postdoctoral fellows. All staff were supervised by a board certified neuropsychologist. Tests were administered in a fixed order and in a standardized manner in accordance with the tests’ manuals. Data were collected between June 2006 and April 2014. The sample included the complete sample reported in Shura et al (2014), plus the addition of 19 participants who validly completed the ongoing protocol since that publication. Participants selected for the current analysis were required to have complete BDS data as well as score above 82.5% on the immediate recall, delayed recall, and consistency subtests of the Word Memory Test (WMT; Green, 2005), a test of performance validity. Table 1 presents demographic, military history, and basic diagnostic data for

Table 1. Sample descriptive statistics Variable

Mean (SD, range)

Age Years education Male Caucasian Service connected Deployment mTBI Any current DSM-IV diagnosis Current PTSD Current depression

34.94 (9.00, 21–60) 13.90 (1.77, 11–19)

Total (%)

141 (86.0%) 120 (73.2%) 94 (57.3%) 32 (19.5%) 86 (52.0%) 55 (33.5%) 27 (16.5%)

N = 164. SD = standard deviation; mTBI = mild traumatic brain injury; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders – IV (American Psychiatric Association, 2000); PTSD = post-traumatic stress disorder.

86

ROBERT D. SHURA ET AL.

the included sample of 164 participants. BDS-II total scores were not significantly correlated with age (r = –.10, ns) or attained education level (r = .08, ns). There were no significant differences in BDS-II total scores when comparing across gender (t = –.08, ns) or minority status (t = –.70, ns).

Downloaded by [University of Otago] at 09:23 15 July 2015

Measures Behavioral Dyscontrol Scale-II. The BDS is a 9-item executive functioning measure that evaluates dynamic motor control, alphanumeric sequencing, and insight (Grigsby & Kaye, 1996). The BDS-II includes the same nine items. Items 1 and 2 require alternating tapping once with one hand and twice with the other; item 3 is a go– no-go task; in item 4, the respondent taps his hand once when the assessor taps twice, and vice versa; items 5 and 6 measure the capacity for motor procedural learning; item 7 involves mimicking examiner movements using the same hand as the examiner; item 8 is an auditory alphanumeric sequencing task, similar to Trail Making Part B; and item 9 evaluates the patient’s insight regarding performance on the test (for more detailed information, see the test manual; Grigsby & Kaye, 1996). The original BDS items were scored from 0 to 2, except for the insight item (scored on a 4-point scale), for a maximum score of 19. With the BDS-II, items are scored from 0 to 3, for a summative total raw score range of 0 to 27, with lower scores reflecting poorer performance. Initial reliability of the original BDS was reported to be adequate in an elderly sample (α = .87; Grigsby & Kaye, 1996); however, reliability in a post-deployment sample was significantly lower with the BDS-II (α = .62, Shura et al., 2014), likely due to differences between the BDS and BDS-II scoring systems, homogeneity of the post-deployment population, and relatively healthy neurological status of the younger sample. Basic normative data for the post-deployment Veteran population were also provided by Shura et al. (2014). The test is available from the author directly. Additional test selection. The selection of tests from the available standardized battery for use in the validity analysis involved a number of considerations. The aim of this process was to select tests with the strongest psychometric properties and that allowed the evaluation of as many cognitive domains and sub-domains as possible. We used a reductionist approach and began by first removing from consideration the self-report measures, the BDS-II, the Word Memory Test (WMT; used to exclude invalid cognitive performance), and the Wechsler Test of Adult Reading (WTAR). We then removed tests for which psychometric data was unavailable or unacceptable. Next we sorted the possible outcome variables from the remaining tests into cognitive domains based on descriptions in Lezak, Howieson, Bigler, and Tranel (2012): sensation, motor output, processing speed, language, visual spatial reasoning/construction, attention, memory, and executive functions (note that this list deviates slightly by separating processing speed into a distinct domain from attention). Although many of the tests include verbal aspects, none specifically measured language as identified by Lezak et al. (2012; i.e., aphasic processes), thus we were unable to sample the language domain. Within each domain, tests were then sorted into subdomains based on Lezak et al. (2012) and descriptions from each test manual. Several domains/sub-domains included more than one test. In this case we chose the test with the best reliability to represent that domain/sub-domain. Domains, sub-domains, the tests selected to represent each

BDS-II STRUCTURE AND VALIDITY

87

domain/sub-domain, as well as the test reliability estimates are presented in Table 2. Raw scores were used for all outcome variables. We hypothesized that the BDS-II outcome measures would correlate with measures of executive functioning (based on prior studies of the original BDS), complex attention and processing speed (as these abilities are foundational for executive functioning and many overlap across domains), and motor output (due to the motor component of the BDS-II), suggesting convergent validity. In contrast, we hypothesized that measures of sensation, visual-spatial ability, and memory would not correlate with the BDS-II measures, suggesting divergent validity. Table 2. List of selected tests, domain and sub-domain of each test, and reliability estimates

Downloaded by [University of Otago] at 09:23 15 July 2015

Domaina Sensation Motor Output Processing Speed Language Visual-Spatial Attentional

Memory

Executive

Sub-Domaina Olfaction Speed Dexterity Processing Speed Construction Basic Attention Vigilance Divided Attention WM: Complex Tracking WM: Sequencing Verbal Learning Verbal Recall Verbal Recognition Visual Learning Visual Recall Visual Recognition Abstract Reasoning Concept Formation Planning Interference Impulsivity Self-Regulation

Measure UPSIT Tap Dom Peg Dom WAIS-III DSC none WAIS-III BD CPT-II OM CPT-II ISI TMT Ratio PASAT WAIS-III LNS CVLT-II 1-5 CVLT-II LDFR CVLT-II d’ BVMT-R Trial 3 BVMT-R DR RCFT Recog WAIS-III SIM WCST-64 Errors RCFT Copy Stroop Ratio CPT-II COM COWAT

r b

.93 .77c .86c .84d .86d .94e .51e .79/.89f .90g .82d .82h .88h .86h .95i .87i .87j .86d .77k .68l .82/.73m .83e .83n

r = reliability estimate per citation; WM = working memory; UPSIT = University of Pittsburg Smell Identification Test; Tap = Finger Tapping Test; Dom = dominant hand Peg = Grooved Pegboard; WAIS-III = Wechsler Adult Intelligence Scale-III; DSC = Digit-Symbol Coding; BD = Block Design; CPT-II = Conners’ Continuous Performance Test-II; OM = Omissions; ISI = Hit Reaction Time Inter-Stimulus Interval Change; TMT Ratio = Trail Making Test Ratio score; PASAT = Paced Auditory Serial Addition Test Total score; LNS = Letter Number Sequencing; CVLT-II = California Verbal Learning Test-II; 1–5 = Trials 1–5 Correct; LDFR = Long Delay Free Recall; d’ = Total Recognition Discriminability; BVMT-R = Brief Visuospatial Memory TestRevised; DR = Delayed Recall; RCFT = Rey Complex Figure Test; Recog = Recognition Total Correct; SIM = Similarities; WCST-64 = Wisconsin Card Sort Test-64; COM = Commissions; COWAT = Controlled Oral Word Association Test (CFL). aAdapted from Lezak et al. (2012). bDoty, 1995. cDikmen, Heaton, Grant, & Temkin, 1999. dWechsler, 1997. eConners & MHS Staff, 2004. fReliability for Part A/Part B; Dikmen, Heaton, Grant, & Temkin, 1999. gCrawford et al., 1998. hDelis et al., 2000. iBenedict, 1997. jMeyers & Meyers, 1995. kGreve et al., 2002. lStrauss et al., 2006. mColor/Color-Word; Golden & Freshwater, 2002. nRuff et al., 1996.

Downloaded by [University of Otago] at 09:23 15 July 2015

88

ROBERT D. SHURA ET AL.

Test descriptions. The University of Pittsburg Smell Identification Test (UPSIT; Doty, 1995) measures olfaction, and was selected as the only available test of sensation. The total score was used as the outcome variable, with higher scores reflecting better performance. Finger tapping is a measure of speeded motor output (Reitan, & Wolfson, 1985). Dominant hand performance (Tap Dom) was used as the outcome variable with higher scores representing better performance. The Grooved Pegboard test (Reitan & Davison, 1974) is a measure of fine motor dexterity. Dominant hand performance (Peg Dom) in seconds was used as the outcome variable with higher scores representing poorer performance of the task. The Wechsler Adult Intelligence Scale-III (WAIS-III; Wechsler, 1997) is a clinician-administered battery for the assessment of intellectual ability. Four subtests were used as outcome variables. Digit Symbol-Coding (DSC) is a timed performance test of processing speed, visual scanning, and motor control. Block Design (BD) is a measure of visual-spatial reasoning and construction. Letter Number Sequencing (LNS) is a verbal, alphanumeric measure of working memory. Similarities (SIM) is a test of simple, abstract verbal reasoning. Higher scores on WAIS-III subtests reflect better performance. The Conners’ Continuous Performance Test-II Version 5 (CPT-II; Conners & MHS Staff, 2004) is a computerized test evaluating sustained attention, reaction time, and impulsivity. Outcome variables for the CPT-II included raw scores for Omissions (OM, lack of response to targets), which is considered a measure of inattention, Commission Errors (COM, responses to non-targets), which is considered a measure of impulsivity, and Hit Reaction Time Inter-Stimulus Interval Change (ISI, reaction time adjustment to differing stimulus presentation timings), which is considered a measure of vigilance. Hit Reaction Time (HRT) was also used in post hoc analyses. Higher scores on these variables reflect poorer performance. The Trail Making Test (TMT) consists of two parts: Part A evaluates processing speed and Part B evaluates set shifting or divided attention (Reitan & Wolfson, 1985). The ratio of Part B to Part A raw scores in seconds (TMT Ratio) was used as the outcome variable, providing a measure of divided attention controlled for to processing speed (see Oosterman et al., 2010). Higher scores reflect greater discrepancy between the two subtests and poorer performance on Trails B when accounting for processing speed. The Paced Auditory Serial Addition Test (PASAT; Gronwall, 1977) is a measure of auditory processing speed and working memory. A total score was calculated by summing total correct across the first two trials and used as the outcome variable. The first two trials were used as per test instructions, the test is to be discontinued if these trials are not tolerated; thus a portion of the sample was missing data for the final two trials. Higher scores represent better performance. The California Verbal Learning Test-II (CVLT-II; Delis, Kramer, Kaplan, & Ober, 2000) is a well-validated test of verbal learning, recall, and recognition. Three outcome variables were selected: the Trials 1–5 Correct (1–5) as a measure of verbal learning, the Long Delay Free Recall (LDFR) as a measure of verbal recall, and Total Recognition Discriminability (d’) as a measure of recognition. Higher scores on the CVLT-II variables represent better performance. The Brief Visuospatial Memory Test-Revised (BVMT-R; Benedict, 1997) is a test of visual learning and memory. Two scores were selected: the Trial 3 score as a measure of visual learning and the Delayed Recall (DR) score as a measure of visual recall. Higher scores reflect better performance on both

Downloaded by [University of Otago] at 09:23 15 July 2015

BDS-II STRUCTURE AND VALIDITY

89

measures. The Rey Complex Figure Test (RCFT; Meyers & Meyers, 1995) is a test of visual-spatial construction ability, visual memory, and planning. The Copy score (Copy) was selected as a measure of planning, and the Recognition Total Correct score (Recog) was used as a measure of visual recognition. Higher scores on these two variables reflect better performance. The Wisconsin Card Sort Test (WCST; Heaton, 1981) is a measure of abstract reasoning, concept formation, and set shifting. This study used a computerized short form with a maximum of 64 card presentations (Greve, 2001). Total Errors (WCST-64 Errors) was selected as a measure of concept formation, as that measure has one of the highest reliability of the short form measures. Higher scores on this measure reflect poorer performance. The Stroop Color and Word Test, Adult Version (Golden & Freshwater, 2002) consists of three parts: The Word Reading and Color Naming trials evaluate processing speed, and the Color/Word Inhibition trial evaluates inhibitory abilities and resistance to interference. The ratio of Color/Word Inhibition to Color Naming was used as the outcome variable, providing a measure of inhibition controlled for variance due to processing speed (Stroop Ratio; see Lansbergen, Kenemans, & van Engeland, 2007). Higher scores on this measure reflect less interference and thus better performance. Last, the Controlled Oral Word Association Test (COWAT; Ruff, Light, Parker, & Levin, 1996) is a measure of phonemic verbal fluency. The total of the three CFL trials was used, with higher scores reflecting better performance. RESULTS Factor analysis of the BDS-II Initial factor studies used PCA with the original BDS scoring system and in samples very different from the post-deployment population. Given the differences in demographic characteristics and scoring system, as well as a lack of clear theoretical support for a particular factor structure, EFA was utilized to examine the factor structure in the current sample. No factor analyses have been previously conducted using the BDS-II scoring system, thus there was no pre-existing model or theory from which to work, and EFA was more appropriate than CFA. EFA was run on the nine items of the BDS-II using Kaiser’s criterion, the maximum likelihood extraction method as suggested by Costello and Osborne (2005), and an oblique direct oblimin rotation due to the intercorrelations found by EcklundJohnson et al. (2004). Power analysis for a conservative, small effect size confirmed that the minimum number of participants required (152) was surpassed. The KaiserMeyer-Olkin measure of sampling adequacy was .69, which exceeded the recommended value of .6 (Kaiser, 1974), and Bartlett’s Test of Sphericity was significant (151.40, df = 36, p < .001), supporting the use of factor analysis with these data. Results of this EFA indicated a three-factor solution that accounted for 52.60% of the variance and was consistent with initial studies of the BDS that found a three-factor solution. However, examination of the scree plot for these data suggested the use of two factors. This was further supported by results of a Monte Carlo parallel analysis using the initial eigenvalues from the EFA, which found two factors with initial eigenvalues exceeding the corresponding criterion values in the random matrix (9 variables, 164 participants, 100 iterations). Therefore, two factors were forced for a second exploratory model.

90

ROBERT D. SHURA ET AL.

Table 3. Pattern, structure, and communality coefficients for Behavioral Dyscontrol Scale-II factor analysis forcing two factors Item

Downloaded by [University of Otago] at 09:23 15 July 2015

1 2 3 4 5 6 7 8 9

Skilla Dynamic organization Dynamic organization Disinhibition Echopraxia; perseveration Motor procedural learning Motor procedural learning Echopraxia; monitoring Attention flexibility Insight

Pattern coefficients

Structure coefficients

Factor 1

Factor 2

Factor 1

Factor 2

–.636 –.637

.209

–.635 –.593

.417 .516 –.344 .506

.403 .530 .308 .558

–.261 .412 .438

.383 .482

Communalities

–.213 –.408 –.323 –.284 –.278

.404 .367 .164 .282 .200 .332 .085 .154 .248

a

N = 164. Factor loadings < .2 are omitted.

(Grigsby & Kaye, 1996).

Results of the two-factor analysis explained a total of 40.90% of the variance (Factor 1 = 25.59%, Factor 2 = 15.31%). The two factors were significantly but weakly correlated at r = .24. Pattern, structure, and communality coefficients are shown in Table 3. The model supports the inclusion of Items 3, 4, 6, 8, and 9 in Factor 1 and the inclusion of Items 1, 2, 5, and 7 in Factor 2. Cronbach’s α was computed for the total score and two resulting factors: total score, α = .59; Factor 1, α = .54; Factor 2, α = .53. Table 4 presents descriptive data at the item and scale level. Notably, this sample performed quite well, with only item 8 (alphanumeric sequencing) showing a mean below 2. This suggests that despite increasing the maximum score for the BDS-II, the possibility of a ceiling effect remains for neurologically healthy samples.

Table 4. Item and scale statistics for the Behavioral Dyscontrol Scale-II Score Frequencies Item/Scale

0

1

2

3

Mean (SD)

1 2 3 4 5 6 7 8 9 Total F1 F2

1 0 1 0 0 2 0 3 0

0 7 12 2 21 23 11 64 2

41 47 50 25 75 71 52 41 27

122 110 101 137 68 68 101 56 135

2.73 (0.48) 2.63 (0.57) 2.53 (0.66) 2.82 (0.41) 2.29 (0.68) 2.25 (0.74) 2.55 (0.62) 1.91 (0.90) 2.81 (0.42) 22.52 (2.73) 12.33 (1.94) 10.20 (1.51)

SD = standard deviation; Total = Behavioral Dyscontrol Scale-II total score; F1 = Factor 1 score; F2 = Factor 2 score.

BDS-II STRUCTURE AND VALIDITY

91

Downloaded by [University of Otago] at 09:23 15 July 2015

Validity Analyses A series of correlations were run among the BDS-II Total score, the two BDS-II factors, age, years of education, and 22 measures of cognitive functioning in order to examine construct validity in this non-elderly sample of post-deployment veterans, which had not been previously examined in the literature (Table 5). Demographic variables were included in the analysis because the outcome variables were raw scores rather than demographic-based normative scores, and several variables were significantly correlated with age and/or education. A hierarchical linear regression was conducted for each of the BDS-II outcome variables (Total score, Factor 1, and Factor 2) to examine their relationship to performance on specific cognitive tests spanning multiple cognitive domains. For each outcome variable, the model included demographic variables (age, gender, race, and years of education) in the first step, as well as all variables significantly correlated with that outcome variable in step 2. All assumptions for regression were met, and power analysis confirmed that the minimum number of participants required (131) was surpassed for a conservative, small effect size in the regression with the most variables (F1, with four demographic variables in Step 1 and 12 predictor variables in Step 2). Table 6 shows results of the final step for each model, excluding demographics. When examining the BDS-II Total score, Step 1 was not significant, explaining 4.4% of the variance. After Step 2 the model was significant F(15, 131) = 5.07, p < .001, and represented a significant improvement over Step 1, F change = 6.08, p < .001, and explained 36.7% of the total variance. In the final model, PASAT (Standardized Βeta = .34, p < .001) and Tapping Dominant (Standardized Βeta = .21, p = .006) were statistically significant predictors of the BDS-II total score. Results for the BDS-II Factor 1 score were similar: Step 1 was not significant and explained 4.3% of the variance. After Step 2 the model was significant, F(16, 128) = 3.64, p < .001, represented a significant improvement over Step 1, F change = 4.18, p < .001, and explained 31.2% of the variance. In the final model, PASAT (Standardized Βeta = .28, p = .005) and Tapping Dominant (Standardized Βeta = .25, p = .003) were statistically significant predictors. Results for BDS-II Factor 2 showed a different pattern. Step 1 was not significant and explained a total of 2.8% of the variance. After Step 2 the model was significant, explained 21.3% of the variance, F(9, 138) = 4.15, p < .001 and represented a significant improvement over Step 1, F change = 6.50, p < .001. In this model, PASAT (Standardized Βeta = .31, p = .002) and CPT-II Commissions (Standardized Βeta = –.20, p = .013) were significant predictors.

Post hoc analyses Four post-hoc analyses were conducted with the BDS-II variables. First, BDS-II item 8 is a measure of alphanumeric sequencing, similar to the TMT Part B except that item 8 is verbal and does not involve a motor speed component. However, the TMT Ratio score was not significantly correlated to any of the three BDS-II scores. Correlations were run between all nine BDS-II items and TMT A, TMT B, and TMT Ratio. TMT Ratio was not significantly correlated to any BDS-II item. TMT B was significantly correlated with item 8 (r = –.29, p < .001) and item 9 (r = –.18, p = .021), but

92

ROBERT D. SHURA ET AL.

Table 5. Correlations among the Behavioral Dyscontrol Scale-II, demographic variables, and other neurocognitive measures Domain

Measure

Demographics

Age Education BDS-II Total BDS-II F1 BDS-II F2 UPSIT Tap Dom Peg Dom WAIS-III DSC WAIS-III BD CPT-II OM CPT-II ISI TMT Ratio PASAT WAIS-III LNS CVLT-II 1-5 CVLT-II LDFR CVLT-II d’ BVMT-R Trial 3 BVMT-R DR RCFT Recog WAIS-III SIM WCST-64 Errors RCFT Copy Stroop Ratio CPT-II COM COWAT

BDS-II

Sensation Motor

Downloaded by [University of Otago] at 09:23 15 July 2015

PS V-S Attentional

Memory

EF

Age

Yearseducation

BDS-II total

BDS-II F1

BDS-II F2

— .17* –.10 –.13 –.02 –.09 –.03 .15* –.13 –.15 –.05 –.11 .01 .04 .01 –.07 –.04 –.21** –.26** –.21** –.35** .01 .15 –.12 –.21** –.21** –.10

.17* — .19* .16* .14 .07 .00 –.01 .25** .10 –.03 .04 –.04 .22** .19* .09 .06 .06 .05 .07 .03 .17* –.09 .11 –.03 –.15 .11

–.10 .09 — .84** .73** .01 .29** –.09 .34** .22** –.23** .05 .02 .47** .45** .17* .18* .15 .07 .12 .03 .20* .00 .03 –.04 –.25** .21**

–.13 .08 .84** — .24** .10 .27** –.11 .35** .23** –.26** .06 –.06 .40** .42** .26** .21** .20* .12 .17* .05 .17* –.05 .03 –.05 –.15 .25**

–.02 .06 .73** .24** — –.11 .18* –.02 .17* .11 –.09 .02 .12 .35** .28** –.02 .05 .02 –.03 –.01 –.01 .14 .07 .02 –.00 –.26** .06

BDS-II = Behavioral Dyscontrol Scale-II; F1 = BDS-II Factor 1; F2 = BDS-II Factor 2; PS = processing speed; V-S = visual-spatial/construction ability; EF = executive functions; UPSIT = University of Pittsburg Smell Identification Test; Tap = Finger Tapping Test; Dom = dominant hand; Peg = Grooved Pegboard; WAIS-III = Wechsler Adult Intelligence Scale-III; DSC = Digit-Symbol Coding; BD = Block Design; CPT-II = Conners’ Continuous Performance Test-II; OM = Omissions; ISI = Hit Reaction Time Inter-Stimulus Interval Change; TMT Ratio = Trail Making Test Ratio score; PASAT = Paced Auditory Serial Addition Test Total score; LNS = Letter Number Sequencing; CVLT-II = California Verbal Learning Test-II; 1–5 = Trials 1–5 Correct; LDFR = Long Delay Free Recall; d’ = Total Recognition Discriminability; BVMT-R = Brief Visuospatial Memory Test-Revised; DR = Delayed Recall; RCFT = Rey Complex Figure Test; Recog = Recognition Total Correct; SIM = Similarities; WCST-64 = Wisconsin Card Sort Test-64; COM = Commissions; COWAT = Controlled Oral Word Association Test (CFL). n = 154 to 164. *= p < .05 (two-tailed); **= p < .01 (two-tailed).

none of the other seven items, supporting the similarity of these two measures of timed alphanumeric sequencing. However, TMT A was significantly correlated to a number of items: item 2 (r = –.24, p = .002), item 4 (r = –.16, p = .046), item 8 (r = –.28, p < .001), and item 9 (r = –.16, p = .048). The similarity in correlation strength between item 8 with TMT A and TMT B, together with the lack of correlation with the TMT ratio, suggest that variability in processing speed rather than divided attention accounts for the variability in performance on item 8.

BDS-II STRUCTURE AND VALIDITY

93

Table 6. Final step of three hierarchical regression models Model

Measures

B

p

Tap Dom WAIS-III DSC WAIS-III BD CPT-II OM PASAT WAIS-III LNS CVLT-II 1-5 CVLT-II LDFR WAIS-III SIM CPT-II COM COWAT

.08 .00 .02 –.11 .03 .10 .01 .07 .02 –.03 .02

.006 .998 .402 .164 .000 .243 .845 .519 .715 .418 .433

Tap Dom WAIS-III DSC WAIS-III BD CPT-II OM PASAT WAIS-III LNS CVLT-II 1-5 CVLT-II LDFR CVLT-II d’ BVMT-R DR WAIS-III SIM COWAT

.06 .00 .01 –.06 .01 .04 .04 –.05 .22 .02 –.01 .02

.003 .764 .595 .296 .005 .487 .091 .600 .420 .876 .807 .315

Tap Dom WAIS-III DSC PASAT WAIS-III LNS CPT-II COM

.02 –.00 .01 .04 –.05

.255 .782 .002 .455 .013

Downloaded by [University of Otago] at 09:23 15 July 2015

BDS-II total

BDS-II Factor 1

BDS-II Factor 2

R2

R2 change

F change

p of F change

.37

.32

6.08

.000

.31

.27

4.18

.000

.21

.19

6.49

.000

Demographic statistics for Step 2 are not included as all were non-significant. B = unstandardized beta; BDS-II = Behavioral Dyscontrol Scale – II; Tap Dom = Finger Tapping Test dominant hand; WAIS-III = Wechsler Adult Intelligence Scale-III; DSC = Digit-Symbol Coding; BD = Block Design; CPT-II = Conners’ Continuous Performance Test-II; OM = Omissions; PASAT = Paced Auditory Serial Addition Test Total score; LNS = Letter Number Sequencing; CVLT-II = California Verbal Learning Test-II; 1–5 = Trials 1–5 Correct; LDFR = Long Delay Free Recall; SIM = Similarities; COM = Commissions; COWAT = Controlled Oral Word Association Test (CFL); d’ = Total Recognition Discriminability; BVMT-R DR = Brief Visuospatial Memory Test-Revised Delayed Recall.

Second, item 3 (a go–no-go task) contains a component of impulsivity, but was not in the factor related to the measure of impulsivity. To further explore this, Item 3 was correlated with CPT-II Commissions, Stroop Ratio, Stroop Color, and Stroop Color-Word. Item 3 was not significantly correlated with Commissions or Stroop Ratio, but was significantly correlated with Stroop Color (r = .29, p < .001) and Stroop ColorWord (r = .23, p = .004). This provides further support for the role of psychomotor speed in BDS-II performance and demonstrates that variability in processing speed rather than impulsivity accounts for variability in performance on this item.

Downloaded by [University of Otago] at 09:23 15 July 2015

94

ROBERT D. SHURA ET AL.

In order to explore the relationship between CPT-II variables and the BDS-II factors, correlations were run between the two factors and CPT-II Hit Reaction Time (HRT). HRT was not significantly correlated with either factor Factor 1 (r = –.12, ns); Factor 2 (r = .09, ns). However, CPT-II Commissions was significantly and negatively correlated to CPT-II Reaction Time (r = –.55, p < .001), suggesting that in this sample more commission errors were related to faster reaction times, thus suggesting that CPTII Commissions reflect impulsivity more than inattention. Finally, since the PASAT is a complex task that likely requires several different cognitive processes to complete successfully, it is difficult to make inferences about the cognitive processes involved in the BDS-II from this relationship. In an attempt to clarify this relationship, the regression analysis for the BDS-II Total score was re-run without the PASAT to examine how other variables might relate to performance in its absence. Step 1 was not significant, explaining 5.7% of the variance. After Step 2 the model was significant F(14, 142) = 4.83, p < .001, and represented a significant improvement over Step 1, F change = 5.57, p < .001, and explained 32.3% of the total variance. In the final model, Tapping Dominant (Standardized Βeta = .17, p = .026) and WAIS-III LNS (Standardized Βeta = .28, p = .001) were statistically significant predictors of the BDS-II total score. Although this model explained less variance than the model using the PASAT, the results are much clearer. The replacement of the PASAT by WAIS-III LNS suggests that the working memory component of the PASAT was the primary aspect related to BDSII performance. Similar results were seen when the PASAT was dropped from the models predicting BDS-II factors, demonstrating the consistency of the working memory relationship and confirming its relevance to performance on the BDS-II.

DISCUSSION This study examined the factor structure and construct validity of the BDS-II in a sample of non-elderly post-deployment veterans. Results suggest a two-factor solution for the BDS-II, and demonstrate that performance on the BDS-II is related to measures of working memory, motor output, and impulsivity. This study is the first to examine the factor structure and construct validity of the BDS-II scoring system, and the first to do so for either the BDS or BDS-II in a sample of exclusively non-elderly participants. The original BDS was validated in geriatric samples; however, few studies have examined the validity of the updated scoring system, referred to as the BDS-II. Of the three studies that have examined the BDS-II scoring system, only one used an exclusively non-geriatric (age < 65) sample (Shura et al., 2014), and the other two studies both included the same sample of 49 participants with a large age range (age 18 to 71; Leahy et al., 2003; Suchy et al., 2003). Thus, the results of the current study address this gap in the literature on the BDS-II. Unlike research on the original BDS supporting a three factor solution (Motor Programming, Environmental Independence, and Fluid Intelligence; Ecklund-Johnson et al., 2004; Grigsby et al., 1992), our analyses suggest a two-factor solution provides the best fit for the data. There are several possible reasons for this discrepancy with previous results including differences in statistical analyses (discussed below), age ranges of samples, sample sizes, as well as differences in the psychometric properties of the BDS and BDS-II scoring systems (Leahy et al., 2003; Shura et al., 2014). The initial

Downloaded by [University of Otago] at 09:23 15 July 2015

BDS-II STRUCTURE AND VALIDITY

95

factor structure of the BDS was developed using PCA and orthogonal rotations (Grigsby et al., 1992; Hall & Harvey, 2008). A subsequent study generally supported the initial three-factor solution, but only if inter-correlations were allowed in the model (which would not be consistent with use of orthogonal rotations; Ecklund-Johnson et al., 2004). The current study addressed these methodological issues in several ways, including parallel analysis and scree examination to determine factor number, the use of a maximum likelihood factor analysis which avoids the inflated variance estimates resulting from PCA, and an oblique rotation which is more appropriate considering the likely inter-correlation of the items (Ecklund-Johnson et al., 2004). Consistent with the current findings, Hall and Harvey (2008) found a two-factor solution using the original scoring system and an orthogonal design with varimax rotation; however, the authors only included the first seven items (excluding the attention flexibility and insight items) and the mean age of the sample was 79.3. Thus the results are not easily comparable to our study using all nine items of the BDS, the BDS-II scoring system, and a sample with a mean age of 34. In qualitative comparison, the structure of Hall and Harvey’s two-factor solution was not consistent with the structure of our final two-factor solution (see Table 7). Because prior factor analyses were completed using the original BDS, comparisons to the current results using the BDS-II should be undertaken cautiously. In particular, when compared to item loadings of the BDS three-factor solution, the item loadings of the current factor solution are not intuitive. This may partially be due to the model forcing two rather than three factors, thus reducing specificity; however, further inspection is warranted. Factor 2 broadly incorporates items evaluating basic motor skills with inhibition control components: simple alternating movements, a finger movement sequence, and mimicking the movements of the examiner while inhibiting the urge to mirror the movements. In comparison, items included in Factor 1 are more complex and may have a stronger relation to executive functions such as set-shifting (alphanumeric sequencing). Of note, the two motor procedural learning items were split between the factors. Luria conceptualized these dynamic organization tests as measuring

Table 7. Behavioral Dyscontrol Scale-II items, skills measured, and factor relationships

Item 1 2 3 4 5 6 7 8 9

Skill

Previous research: 3-factora

This study: 3-factor

Previous research: 2-factorb

This study: 2-factor

Dynamic organization Dynamic organization Disinhibition Echopraxia; perseveration Motor procedural learning Motor procedural learning Echopraxia; monitoring Attention flexibility Insight

MP MP EI EI MP MP FI FI FI

2 2 1 1 3 1 3 1 1

SMRB SMRB MPS MPS MPS MPS MPS

F2 F2 F1 F1 F2 F1 F2 F1 F1

Skill descriptors from Grigsby & Kaye (1996); MP = Motor Programming; EI = Environmental Independence; FI = Fluid Intelligence; SMRB = Simple Motor Repetitive Behaviors; MPS = Motor ProblemSolving. a(Ecklund-Johnson et al., 2004); b(Hall & Harvey, 2008).

Downloaded by [University of Otago] at 09:23 15 July 2015

96

ROBERT D. SHURA ET AL.

‘kinetic melody,’ involving an inhibition component, and being sensitive to frontal lesions (Luria, 1980, pp. 423–424). The current results suggest that finger alternations (item 5) requires fewer executive resources, possibly resulting from the sample being studied. Examining how the BDS-II factors relate to performance on other cognitive tests does not completely support conclusions via qualitative examination of each factor’s included items. Results of correlations and regressions for Factor 1 were similar to those of the BDS-II total score: Factor 1 significantly correlated to measures from all domains except olfaction, and Finger Tapping Test and PASAT were the best predictors of Factor 1 in regression analysis. Factor 2 had significant correlations with motor output, processing speed, and executive function domains; however, there were no significant correlations with olfaction, memory, or visual-spatial variables. In addition, regression analyses revealed that PASAT and CPT-II Commissions were the best predictors of Factor 2. The different relationships of the factors to other cognitive tests suggest functional specificity, but the relationship between Factor 2 (simple motor items) to Commissions (a test of impulsivity) is counterintuitive as the BDS-II disinhibition item (item 3) loads with Factor 1. The CPT-II manual (Conners & MHS Staff, 2004) states that Commissions can signify either impulsivity (in the context of fast reaction time), or inattention (if reaction time is slow). However, post hoc analyses revealed that in this sample, CPT-II Commissions was negatively correlated to HRT, supporting that those with a higher number of errors of commission were committing the errors in the context of faster reaction time. This further supports that Factor 2 is related to impulsivity, not inattention. Additionally, item 3 was found to correlate to processing speed variables, and not to measures of inhibition. Thus, although Factor 1 would seem to more measure inhibition based on item descriptions, analyses support inhibition as a component of Factor 2. Additional analyses also suggest that Factor 1 items are heavily related to processing speed (TMT A and B correlated to item 8; Stroop Color and Color-Word correlated to item 3), consistent with a general mental efficiency and processing speed component of the factor. Future studies might more thoroughly explore this relationship. From a functional neuroanatomy perspective, the current results provide evidence that the BDS-II actually evaluates a range of abilities representative of the functioning of the frontal lobes rather than executive functioning specifically. For example, the BDS-II Total was found to be significantly correlated with measures of motor output, processing speed, working memory abilities of complex tracking and sequencing, abstract reasoning, impulsivity, and self-regulation; all processes involving the frontal lobes. Further investigation using regression analyses revealed that motor output (Tap Dom) and working memory/complex tracking (PASAT) were the best predictors and explained the most unique variance in the BDS-II total score. These results fit well with our previous conceptualization of a basic motor control factor (posterior frontal regions) and a complex executive factor (prefrontal regions). Of note, the BDS-II was not related to executive function measures of concept formation, planning, or interference control, reflecting the non-unitary nature of executive functioning and functional specialization within the frontal lobes. Results demonstrated that performance on the PASAT is the strongest predictor of performance on each of the BDS-II outcomes after accounting for other tests and demographic variables, suggesting strong overlap in the underlying cognitive processes.

Downloaded by [University of Otago] at 09:23 15 July 2015

BDS-II STRUCTURE AND VALIDITY

97

Post hoc analyses demonstrated that in the absence of the PASAT, WAIS-III LNS was the strongest predictor of BDS-II outcomes, suggesting that the working memory component of the PASAT is the most relevant aspect for BDS-II performance. A survey of neuropsychologists found that the PASAT was the third most popular test of attention (Rabin, Barr, & Burton, 2005), and is frequently correlated to other complex attention measures, such as Trails B, Digit Span Backward, and Digit-Symbol Coding (Tombaugh, 2006). The PASAT heavily relies on processing speed, attention, concentration, executive control, and working memory, and neuroimaging studies have correlated performance on it to numerous areas of the brain (Tombaugh, 2006). Thus, although the relationship between the BDS-II and PASAT suggests the BDS-II is a broad measure of general cerebral functioning and mental efficiency, when the PASAT was eliminated in post hoc regression analyses, LNS became a significant predictor. This suggests that although the BDS-II is related to many different cognitive abilities, working memory seems to be the strongest ability tapped by the BDS-II. Support for divergent validity was provided by results showing that the BDS-II Total score was not significantly correlated with olfaction or visual memory variables. However, there were significant correlations with verbal learning and recall and visualspatial reasoning/construction. Also, there were significant correlations among the BDS-II Total score and at least one measure from each of the major cognitive domains evaluated except sensation, suggesting the possibility that the BDS-II is a broad measure of general cognitive functioning, or perhaps demonstrating the involvement of functions of the frontal lobes across a wide range of cognitive processes. Our hypothesis that the BDS-II would only correlate with executive functioning, complex attention and processing speed, and motor output was therefore not supported. Instead, the BDSII appears to be both a general test of cerebral functioning, and a more specific test of working memory and motor output. Given the differences between the current findings using the BDS-II and previous findings using the BDS, replications will be needed before firm conclusions about the BDS-II factor structure can be reached. Ideally, future studies would examine the proposed factor structure in samples with higher proportions of minorities, women, and non-veterans. Similarly, the current study did not include individuals in the elderly age range, the population for which the BDS was initially developed; thus future studies should examine our proposed two-factor solution in elderly or mixed age samples. Additionally, future studies could explore the effects of various psychiatric and neurological conditions on BDS-II scores. The current results suggest the BDS-II may not be useful for identification of focal deficits, but might be ideal as a screening instrument examining the possibility of deficits in cognitive processes mediated by the function of the frontal lobes. There are several limitations of this study that should be considered when interpreting the results. First, the sample consisted of mostly male veterans who deployed to Iraq and/or Afghanistan, and was heterogeneous with respect to psychiatric symptomatology and history of mild TBI. It is possible that results might differ between participants with or without psychiatric diagnosis or history of TBI. Finally, although we excluded participants who failed a performance validity test, we did not control for symptom validity, which could be examined in an additional study to determine the relationship between the BDS-II and exaggerated emotional distress.

98

ROBERT D. SHURA ET AL.

ACKNOWLEDGMENTS

Downloaded by [University of Otago] at 09:23 15 July 2015

This research was supported by resources of the W.G. ‘Bill’ Hefner Veterans Affairs Medical Center, the Mid-Atlantic Mental Illness Research Education and Clinical Center, and the Department of Veterans Affairs Office of Academic Affiliations Advanced Fellowship Program in Mental Illness Research and Treatment. There are no conflicts of interest to disclose. The prospective studies from which these data were drawn were reviewed and approved by the W.G. (Bill) Hefner VA Medical Center (Hefner VAMC) Institutional Review Board. The welfare and privacy of human participants were protected and maintained. Disclaimer: The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs, the Department of Defense, or the U.S. government.

REFERENCES American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington DC: American Psychiatric Association. Belanger, H. G., Wilder-Willis, K., Malloy, P., Salloway, S., Hamman, R. F., & Grigsby, J. (2005). Assessing motor and cognitive regulation in AD, MCI, and controls using the Behavioral Dyscontrol Scale. Archives of Clinical Neuropsychology, 20, 183–189. Benedict, R. H. B. (1997). Brief Visuospatial Memory Test-Revised: Professional manual. Lutz, FL: PAR Inc. Conners, C. K., & MHS Staff. (2004). Conners’ Continuous Performance Test II (CPT II) for Windows: Technical and software manual. North Tonawanda, NY: MHS. Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10(7), 1–9. Crawford, J. R., Obonsawin, M. C., & Allan, K. M. (1998). PASAT and components of WAIS-R performance: Convergent and discriminant validity. Neuropsychological Rehabilitation, 8, 255–272. Delis, D. C., Kramer, J. H., Kaplan, E., & Ober, B. A. (2000). California Verbal Learning Test (2nd ed., adult version): Manual. Bloomington, MN: NCS Pearson, Inc. Diesfeldt, H. F. A. (2004). Executive functioning in psychogeriatric patients: Scalability and construct validity of the Behavioral Dyscontrol Scale (BDS). International Journal of Geriatric Psychiatry, 19, 1065–1073. Dikmen, S. S., Heaton, R. K., Grant, I., & Temkin, N. R. (1999). Test-retest reliability and practice effects of expanded halstead-reitan neuropsychological test battery. Journal of the International Neuropsychological Society, 5, 346–356. Doty, R. L. (1995). The Smell Identification Test administration manual (3rd ed.). Haddon Heights, NJ: Sensonics Inc. Ecklund-Johnson, E., Miller, S. A., & Sweet, J. J. (2004). Confirmatory factor analysis of the Behavioral Dyscontrol Scale in a mixed clinical sample. The Clinical Neuropsychologist, 18, 395–410. Garavan, H., Ross, T. J., Murphy, K., Roche, R. A. P., & Stein, E. A. (2002). Dissociable executive functions in the dynamic control of behavior: Inhibition, error detection, and correction. NeuroImage, 17, 1820–1829. Golden, C. J., & Freshwater, S. M. (2002). Stroop Color and Word Test: A manual for clinical and experimental uses. Wood Dale, IL: Stoelting Co.

Downloaded by [University of Otago] at 09:23 15 July 2015

BDS-II STRUCTURE AND VALIDITY

99

Green, P. (2005). Green’s Word Memory Test for Windows user’s manual. Edmonton, Canada: Green’s Publishing Inc. Greve, K. W. (2001). The WCST-64: A standardized short-form of the Wisconsin Card Sorting Test. The Clinical Neuropsychologist, 15, 228–234. Greve, K. W., Love, J. M., Sherwin, E., Mathias, C. W., Houston, R. J., & Brennan, A. (2002). Temporal stability of the Wisconsin Card Sorting Test in a chronic traumatic brain injury sample. Assessment, 9, 271–277. Grigsby, J., & Kaye, K. (1996). The Behavioral Dyscontrol Scale: Manual (2nd ed.). Denver, CO: Authors. Grigsby, J., Kaye, K., Kowalsky, J., & Kramer, A. M. (2002a). Association of behavioral selfregulation with concurrent functional capacity among stroke rehabilitation patients. Journal of Clinical Geropsychology, 8, 25–33. Grigsby, J., Kaye, K., Kowalsky, J., & Kramer, A. M. (2002b). Relationship between functional status and the capacity to regulate behavior among elderly persons following hip fracture. Rehabilitation Psychology, 47, 291–307. Grigsby, J., Kaye, K., & Robbins, L. J. (1992). Reliabilities, norms and factor structure of the Behavioral Dyscontrol Scale. Perceptual and Motor Skills, 74, 883–-892. Gronwall, D. M. A. (1977). Paced Auditory Serial Addition Task: A measure of recovery from concussion. Perceptual and Motor Skills, 44, 367–373. Hall, J. R., & Harvey, M. B. (2008). Behavioral regulation: Factor analysis and application of the Behavioral Dyscontrol Scale in dementia and mild cognitive impairment. International Journal of Geriatric Psychiatry, 23, 314–318. Heaton, R. K. (1981). Wisconsin Card Sorting Test Manual. Odessa, FL: Psychological Assessment Resources Inc. Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 31–36. Kaye, K., Grigsby, J., Robbins, L. J., & Korzun, B. (1990). Prediction of independent functioning and behavior problems in geriatric patients. Journal of the American Geriatrics Society, 38, 1304–1310. Lansbergen, M. M., Kenemans, J. L., & van Engeland, H. (2007). Stroop interference and attention-deficit/hyperactivity disorder: A review and meta-analysis. Neuropsychology, 21, 251–262. Leahy, B., Suchy, Y., Sweet, J. J., & Lam, C. S. (2003). Behavioral Dyscontrol Scale deficits among traumatic brain injury patients, Part I: Validation with nongeriatric patients. The Clinical Neuropsychologist, 17, 474–491. Lezak, M. D., Howieson, D. B., Bigler, E. D., & Tranel, D. (2012). Neuropsychological assessment (5th ed.). Oxford, UK: Oxford University Press Inc. Luria, A. R. (1980). Higher cortical functions in man (2nd ed.). New York, NY: Consultants Bureau, Enterprises Inc. Meyers, J. E., & Meyers, K. R. (1995). Rey Complex Figure Test and Recognition Trial: Professional manual. Lutz, FL: Psychological Assessment Resources Inc. Oosterman, J. M., Vogels, R. L. C., van Harten, B., Gouw, A. A., Poggesi, A., Scheltens, P., … Scherder, E. J. A. (2010). Assessing mental flexibility: Neuroanatomical and neuropsychological correlates of the Trail Making Test in elderly people. The Clinical Neuropsychologist, 24, 203–219. Rabin, L. A., Barr, W. B., & Burton, L. A. (2005). Assessment practices of clinical neuropsychologists in the United States and Canada: A survey of INS, NAN, and APA Division 40 members. Archives of Clinical Neuropsychology, 20, 33–65. Reitan, R. M., & Davison, L. A. (Eds.). (1974). Clinical neuropsychology: Current status and applications. Washington, DC: V. H. Winston & Sons. Reitan, R. M., & Wolfson, D. (1985). The Halstead-Reitan Neuropsychological Test Battery: Theory and clinical interpretation. Tuscon, AZ: Neuropsychology Press.

Downloaded by [University of Otago] at 09:23 15 July 2015

100

ROBERT D. SHURA ET AL.

Ruff, R. M., Light, R. H., Parker, S. B., & Levin, H. S. (1996). Benton Controlled Oral Word Association Test: Reliability and updated norms. Archives of Clinical Neuropsychology, 11, 329–338. Shura, R. D., Rowland, J. A., & Yoash-Gantz, R. E. (2014). The Behavioral Dyscontrol Scale-II with non-elderly Veterans. Archives of Clinical Neuropsychology, 29, 409–414. Strauss, E., Sherman, E. M. S., & Spreen, O. (2006). A compendium of neuropsychological tests: Administration, norms, and commentary (3rd ed.). New York, NY: Oxford University Press. Suchy, Y., Blint, A., & Osmon, D. C. (1997). Behavioral Dyscontrol Scale: Criterion and predictive validity in an inpatient rehabilitation unit population. The Clinical Neuropsychologist, 11, 258–265. Suchy, Y., Derbidge, C., & Cope, C. (2005). Behavioral Dyscontrol Scale-Electronic Version: First examination of reliability, validity, and incremental utility. The Clinical Neuropsychologist, 19, 4–26. Suchy, Y., Eastvold, A., Whittaker, W. J., & Strassberg, D. (2007). Validation of the Behavioral Dyscontrol Scale-Electronic Version: Sensitivity and subtle sequelae of mild traumatic brain injury. Brain Injury, 21, 69–80. Suchy, Y., Leahy, B., Sweet, J. J., & Lam, C. S. (2003). Behavioral Dyscontrol Scale deficits among traumatic brain injury patients, Part II: Comparison to other measures of executive functioning. The Clinical Neuropsychologist, 17, 492–506. Tombaugh, T. N. (2006). A comprehensive review of the Paced Auditory Serial Addition Test (PASAT). Archives of Clinical Neuropsychology, 21, 53–76. Watkins, M. W. (2000). Monte Carlo PCA for Parallel Analysis. Retrieved from: http://edpsychas sociates.com/Watkins3.html Wechsler, D. (1997). Wechsler Adult Intelligence Scale – III: Administration and scoring manual. San Antonio, TX: The Psychological Corporation.

Factor structure and construct validity of the Behavioral Dyscontrol Scale-II.

The Behavioral Dyscontrol Scale-II (BDS-II) was developed as an improved scoring method to the original BDS, which was designed to evaluate the capaci...
194KB Sizes 0 Downloads 7 Views