Read doi:10.1016/j.paid.2008.04.012 text version

Personality and Individual Differences 45 (2008) 286­292

Contents lists available at ScienceDirect

Personality and Individual Differences

journal homepage: www.elsevier.com/locate/paid

Impulsivity among adult prisoners: A confirmatory factor analysis study of the Barratt Impulsivity Scale

Jane L. Ireland a,b,*, John Archer a

a b

Department of Psychology, University of Central Lancashire, Preston, UK Psychology Department, Ashworth Hospital, Mersey Care NHS Trust, Liverpool, UK

a r t i c l e

i n f o

a b s t r a c t

The current study explored the structure of a commonly applied impulsivity measure, the Barratt Impulsivity Scale [BIS-11], in a sample likely to include elevated levels of impulsivity (i.e. a prison sample). The main aim was to assess whether the original factor structure of the BIS could be confirmed in a large sample (n = 1103), comprising two subsamples of adult men (n = 383 and 250), and adult women (n = 250 and 220) prisoners. It was predicted that the three factor solution originally proposed (i.e. motor; attentional; non-planning) would be replicated; and produce a multi-dimensional structure consistent across sex. Although exploratory analysis indicated a three factor solution, the nature of the factors was different to those originally proposed for the BIS-11 although there was some convergence for `behavioural [motor] impulsivity'. A different factor structure was preferred for women. The importance of assessing in detail the structure of commonly applied research measures using confirmatory approaches is outlined, with implications for research noted. Ó 2008 Elsevier Ltd. All rights reserved.

Article history: Received 18 January 2008 Received in revised form 8 April 2008 Accepted 14 April 2008 Available online 23 May 2008 Keywords: Impulsivity Prisoners BIS Impulsivity factor structure Sex differences

1. Introduction Impulsivity plays a prominent role in understanding and diagnosing various forms of psychopathology and, after subjective distress, represents the most common diagnostic criteria in the Diagnostic and Statistical Manual for Mental Disorders (DSM-IV). Despite its pervasive importance as a concept across psychology, there remains a lack of clarity both on its definition and its component structure (Whiteside & Lynam, 2001), with researchers questioning the consequent utility of existing methods of measurement (e.g. Carrillo-de-la-Peña, Otero, & Romero, 1993). With regards to definition, some link impulsivity with concepts such as `sensation-seeking' and use the terms synonymously (e.g. Dom, De Wilde, Hulstijn, van den Brink, & Sabbe, 2006); others describe it as a closely related personality trait (Lejoyeux, Feuché, Loi, Solomon, & Adès, 1998); others consider it a concept closely related to `executive functioning', with a basis in neuropsychological functioning (Cohen, Rosenbaum, Kane, Warnken, & Benjamin, 1999); whereas others argue it is synonymous with `self-control' (e.g. Baumeister & Vohs, 2007). Despite difficulties in definition, there has been a considerable amount of research exploring how impulsivity relates to a range of externalising problems. These include substance misuse (e.g. Dom et al., 2006), personality disorders (e.g. Fossati et al., 2007),

* Corresponding author. Address: Department of Psychology, University of Central Lancashire, Preston, UK. Tel.: +44 01772 894471. E-mail address: [email protected] (J.L. Ireland). 0191-8869/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2008.04.012

dyslexia (e.g. Baker & Ireland, 2007) aggression (e.g. Fossati et al., 2007; Smith, Waterman, & Ward, 2006), self-injurious behaviour (e.g. Raust et al., 2007), compulsive buying (e.g. Billieux, Rochat, Ribetez, & Van der Linden, 2008), major mental illness (e.g. Enticott, Ogloff, Bradshaw, & Fitzgerald, 2008), and anger and psychopathy (e.g. Jackson, Neumann, & Vitacco, 2007). It is clear that impulsivity is considered relevant to a range of social and individual problems. There are number of impulsivity measures available ranging from psychometric instruments to neuropsychological/cognitive testing, although they do not necessarily all correlate with one another (Campbell, 2006). The Barratt Impulsivity Scale (BIS) is the most commonly applied psychometric measure of impulsivity (Spinella, 2007). It has been used across a range of contexts and countries, and translated into various languages (e.g. Baylé et al., 2000; Fossati, Di Ceglie, Acquarini, & Barratt, 2001; von Diemen, Szobot, Kessler, & Pechansky, 2007). The BIS has also been subject to a number of revisions. The original 1959 (Barratt, 1959) version portrayed impulsivity as unidimensional, but subsequent revisions moved to a multidimensional construct. A final revised version was produced in 1995 (version 11) comprising three factors: motor (behaviour), attentional (cognitive) and non-planning. Evidence for these factors was found in samples of undergraduates, psychiatric inpatients and adult male prisoners (Patton, Stanford, & Barratt, 1995). However, the sample size for inpatients and prisoners in this study was small for a factor analysis, with no use of formal confirmatory factor analysis (CFA).

J.L. Ireland, J. Archer / Personality and Individual Differences 45 (2008) 286­292

287

Although some studies, including those using a shorter version of the BIS (e.g. Spinella, 2007), have retained the three factor structure, research has far from confirmed it (Whiteside & Lynam, 2001). Studies employing exploratory factor analysis have produced similar but not identical factors (Baylé et al., 2000), although studies specifying factors a priori in exploratory analyses have reported more convergence (Spinella, 2007). However, this is still not as rigorous an assessment as confirmatory factor analysis. There have, however, been differences noted across studies; for example, a smaller number of factors have been reported with Italian adolescents (Fossati et al., 2001); and a translated version for Portuguese adolescents failed to produce a three factor structure (von Diemen et al., 2007). Thus, although there appears agreement that the BIS can be used as a unidimensional [single factor] measure, the originally proposed three factor structure, remains unconfirmed across a range of samples. There have been few large scale studies exploring the multidimensional nature of impulsivity in populations where impulsivity is expected to be elevated. The majority of research has focused on student samples where, arguably, impulsivity is not a problem of clinical levels. Impulsivity is associated with a range of externalising problems, such as personality difficulties, attention-hyperactivity disorder, aggression and substance use (Whiteside & Lynam, 2001). Accessing populations with increased evidence of such problems, and therefore an expectation of higher risk for elevated impulsivity, is of value in determining the nature and structure of impulsivity. Indeed, the failure to explore the factor structure of impulsivity and other psychometric properties beyond undergraduate student populations has been a noted criticism by those attempting to define and conceptualise impulsivity (Whiteside & Lynam, 2001). The current study aims to explore the nature and structure of impulsivity in a sample where elevated levels of impulsivity are expected, in this instance a prisoner sample. It will employ both parallel and confirmatory factor analysis, across four adult samples of men and women prisoners. It was predicted that: (1) The three factor solution proposed by Patton et al. (1995) would be replicated. (2) That the multi-dimensional structure of the BIS would be consistent between sex. 2. Method Four samples were employed as follows1: 2.1. Study I participants Adult men (n = 442) were sampled from two medium secure prisons in the UK. Fifty-nine failed to complete the questionnaire, leaving a final sample of 383 (86.7% completion rate). The mean age was 32.9 years (SD = 11.1). Eighty-nine percent were of White ethnic origin. Thirty-one percent were serving for violent offences, 22% for acquisitive offences, 22% for sex offences, 17% for other indictable offences and 8% for drug offences. 2.2. Study II participants Adult women (n = 287) were sampled from two medium secure prisons. Thirty-seven failed to complete the questionnaire, leaving

a final sample of 250 (87% completion rate). The mean age was 32.0 years (SD = 8.8). Ninety-two percent were of White ethnic origin. Thirty-five percent were serving for violent offences, 30% for drug offences, 25% for acquisitive offences and 10% for other indictable offences. 2.3. Study III participants Adult men (n = 295) were sampled from four medium secure prisons. Forty-five were excluded for failing to complete the questionnaire, leaving 250 (84.7% completion rate). The mean age was 32.5 years (SD = 9.8). Ninety percent were of White ethnic origin. Thirty-seven percent were serving for violent offences, 25% for acquisitive offences, 10% for sex offences, 11% for other indictable offences and 17% for drug offences. 2.4. Study IV participants Adult women (n = 236) were sampled from two medium secure prisons. Sixteen failed to complete the measures, leaving a final sample of 220 (93.2% completion rate). The mean age was 30.7 years (SD = 8.6). Eighty-nine percent were of White ethnic origin. Twenty-seven percent were serving for violent offences, 36% for acquisitive offences, 1% for sex offences, 8% for other indictable offences and 28% for drug offences. 2.5. Measures All prisoners completed the Barratt Impulsivity Scale, Version 11 (BIS-11; Patton et al., 1995). Two items were removed for the purposes of the current study since they were not applicable to a prison sample (i.e. `I change where I live' and `I plan for job insecurity'). The BIS-11 comprised 28 statements. Participants were asked to respond on a 4-point Likert scale (1 = rarely/never to 4 = almost always/always). 2.6. Procedure All questionnaires were administered during a lunchtime period when prisoners are in their cells on their own. They were given to each prisoner personally by a research assistant. All were informed of the nature, purpose and anonymity of the study. Analysis was conducted using SPSS and AMOS for the confirmatory factor analyses. 3. Results Initially, the structure of the BIS was examined with regards to its unidimensional structure, followed by an attempt to confirm the published three factor structure. Following this are exploratory and confirmatory stages examining a proposed revised factor structure. All models were recursive and identified and standardized with variances set to 1.00. Samples from studies I and II were used to explore the factor structure, with samples from studies III and IV used to confirm the revised structure. 3.1. STEP I: Is the BIS-11 unidimensional? Two confirmatory factor analyses were conducted, one on adult men (Study I) and one on adult women (Study II), with a single latent variable. The models failed to fit for the men (v2 (350) = 1668.99 [P = .00]; RMSEA = .10 [.09­.10]; GFI = .63; CFI = .51; ECVI = 4.66 [4.33­5.00]) and the women (v2 (250) = 1287.33 [P = .00]; RMSEA = .10 [.09­.11]; GFI = .61; CFI = .50; ECVI = 5.62 [5.19­6.08]).

The data was derived from a larger dataset exploring aggression and prison bullying. The results of these studies can be found in Ireland, Archer, and Power (2007) and Archer, Ireland, and Power (2007). The present study, however, is wholly unconnected with the issues addressed in these previous articles. The first paper makes no mention of impulsivity and the second mentions level of impulsivity across the variables of interest i.e. aggression and bullying. No exploration of the BIS data is conducted.

1

288

J.L. Ireland, J. Archer / Personality and Individual Differences 45 (2008) 286­292

3.2. STEP II: Confirming the published three factor structure of Patton et al. (1995) Following demonstration of a non-unidimensional structure, the next step was to attempt to confirm the BIS-11 structure proposed by Patton et al. (1995), using samples from studies I and II. The three factors were non-planning (10 items); motor (10 items); and attentional (8 items). This was explored separately between men and women using confirmatory factor analysis. Covariances were added between the factors to reflect their correlation. The models produced are presented in Figs. 1 and 2. The three factor model failed to fit the observed data well for men (v2 (347) = 1342.7 [P = .00]; RMSEA = .09 [.08­.09]; GFI = .77; CFI = .63; ECVI = 3.82 [3.53­4.13]); i.e. although the RMSEA value was below .10, neither GFI nor CFI exceeded .90. The three factor model also failed to fit the observed data well for women (v2 (357) = 1065.3 [P = .00]; RMSEA = .09 [.08­.09]; GFI = .75; CFI = .62; ECVI = 4.75 [4.37­5.16]). For both models an attempt was made to improve fit by examining modification indexes and covariances. Adding item covariances still ensured the models failed to fit the observed data. 3.3. STEP III: Exploring the factor structure of the BIS-11 The next step focused on exploring the structure of the BIS-11. The data was found suitable for exploratory factor analysis for both

men and women (using Studies I and II), i.e. inspection of the correlation matrix revealed the presence of many coefficients .4 and above, a Kaiser­Meyer­Olkin value of .87 for men and .83 for women, and a Bartlett's test of Sphericity of statistical significance. Two factor analyses were conducted in the first instance, separately for men and women. In order to more strictly identify the number of factors evident, Parallel Analysis (PA), was employed. An examination of the initial factor outlines and explained variance suggested similarities between the sexes and thus men and women from Studies I and II were combined. Parallel Analysis indicated a three factor solution. A Principal Components Analysis with Varimax Rotation was thus employed (to make the use of future item parceling more appropriate and to ensure no assumptions about the model were made). Results are indicated in Table 1, along with the original positioning in the BIS-11 factors and reliability analyses. Two items (22 and 3), `I solve problems by trial and error' and `I am happy-go-lucky' failed to load. Factor 1 was best described as Cognitive Planning Skills; Factor 2 as Behavioural Impulsivity; and Factor 3 as Distractibility. 3.4. STEP IV: Confirming the revised structure The next step focused on confirming the revised three factor structure on two different samples (250 men: Study III and 220 women: Study IV). Reliability analyses for each factor were conducted, separately for men and women. These produced alphas

.64

.94

.61

.75

.92

1.03

.62

1.04

1.11

.76

b1

b3

b5

b8

b13

b16

b20

b23

b26

b28

-56

-.18

.68

.60

.08

.36

.55

.15

.08

.59

Non-Planning

.48

.81

1.04

.63

.48

.37

.67

.94

.84

.83

b2

b6

b9

b11

b14

b17

b19

b21

b24

b27

.65

.32

.35

.56

.66

.77

.51

.23

.27

.51

Motor

1.13

56

.40

.68

.94

.37

.88

1.03

b4

b7

b10

b12

b15

b18

b22

b25

.06

.65

.73

.59

.28

.73

.18

.06

Attentional

Fig. 1. Confirmatory factor analysis of the BIS-11 factor structure using adult men prisoners (n = 383: Study I: RMSEA = .09; GFI = .77; CFI = .63; ECVI = 3.82).

J.L. Ireland, J. Archer / Personality and Individual Differences 45 (2008) 286­292

289

.67

.90

.77

.75

.95

.99

.73

1.19

1.10

.80

b1

b3

b5

b8

b13

b16

b20

b23

b26

b28

.60

-.18

.59

.61

.06

.46

.55

.09

-.02

.59

Non-Planning

.53

.94

1.11

.64

.62

.37

.75

1.00

1.02

.79

b2

b6

b9

b11

b14

b17

b19

b21

b24

b27

.66

.28

.30

.59

.64

.79

.50

.16

.39

.49

Motor

1.07

70

.43

.81

1.02

.25

.86

1.14

b4

b7

b10

b12

b15

b18

b22

b25

-.007

.15

.59

.75

.45

.28

.85

.23

Attentional

Fig. 2. Confirmatory factor analysis of the BIS-11 factor structure using adult women prisoners (n = 250: Study II: RMSEA = .09; GFI = .75; CFI = .62; ECVI = 4.75).

ranging from moderate to good, accounting for the number of items (see Table 1), with all item-to-total correlations positive (Men: Factor 1, a = .85; Factor 2, a = .79; Factor 3, a = .67; Women: Factor 1, a = .85; Factor 2, a = .78; Factor 3, a = .61). Item parcels were employed to examine the model structure in order to reduce error rates (Anderson, Benjamin, Wood, & Bonacci, 2006). They were calculated by using factor totals and correlation magnitude (Table 1 indicates the parcel into which each item was placed).2 Two confirmatory factor analyses were conducted, one for men and one for women. Covariances were added between factors to reflect their correlation. For men this produced a model of good fit (v2 (24) = 37.2 [P = .04]; RMSEA = .047 [.009­.08]; GFI = .97; CFI = .98; ECVI = .32 [.27­.90]), and an improvement over a model not including covariances (RMSEA = .15 [.13­.17]; GFI = .88; CFI = .81; ECVI = .83 [.68­

1.02]). This model is presented in Fig. 3.3 The three factor solution did not fit the data for women, with fit indices poor (v2 (24) = 108.4 [P = .00]; RMSEA = .12 [.10­.15]; GFI = .92; CFI = .88; ECVI = .69 [.56­.85]). Removing covariances worsened the model fit. Attempts to improve the model fit for women were made by examining modification indexes and item covariances. The factor Distractibility appeared to fit poorly overall. Removal of this factor brought model fit to an acceptable level, most notably with regards to GFI and CFI, with RMSEA brought below .10 (RMSEA = .09 [.05­ .14]; GFI = .96; CFI = .96; ECVI = .23 [.18­.32]). The model is presented in Fig. 4.3 4. Discussion The current study did not support the original three factor BIS11 structure proposed by Patton et al. (1995). The prediction was

2 For completeness, models were also conducted without the use of item parcels, i.e. with individual BIS-11 items instead. For men and women, resulting model fit converged with those employing item parcels (i.e., Men; RMSEA = .05; GFI = .91; CFI = .93; ECVI = 1.20; Women; RMSEA = .065; GFI = .91; CFI = .92; ECVI = .93).

3 The three factor model for men was also tested as a second order model [hierarchical confirmatory factor analysis] to determine if the three factors proposed were underneath a single order factor. The resulting model was a good fit, virtually identical to the first (n = 250: Study III: v2 = 32.2, df = 24, p = .04; RMSEA = .047; GFI = .96; CFI = .98; ECVI = .32). A second order model could not be explored for women as there are only two first order factors, with a minimum of three required for identification.

290

J.L. Ireland, J. Archer / Personality and Individual Differences 45 (2008) 286­292

Table 1 Exploratory factor analysis of BIS-11 for men and women prisoners (n = 633) (NB: Factor loadings are rounded to two decimal places) Original BIS factor Factor 1 (19.9% variance). Cognitive Planning Skills 11 items a = .85; I am a careful thinker I am a steady thinker I concentrate easily I am self-controlled I plan tasks carefully I finish what I start I plan trips well ahead of time I save regularly I like to think about complex problems I plan for the future I have regular medical/dental check ups 10 items a = .86 Attentional Attentional Attentional Motor Non-planning Non-planning Non-planning Non-planning Attentional Non-planning Non-planning

a

Factor loading .76 .76 .71 .66 .66 .64 .62 .59 .56 .52 .41 .74 .70 .61 .54 .49 .44 .43 .43 .41 .60 .60 .53 .51 .49 .49

Parcel factor/current study factor F1/1 F1/1 F1/1 F1/2 F1/1 F1/2 F1/2 F1/2 F1/3 F1/3 F1/3 F2/1 F2/1 F2/1 F2/2 F2/2 F2/2 F2/3 F2/3 F2/3 F3/2 F3/1 F3/3 F3/1 F3/2 F3/3

Factor 2 (14.1% variance) Behavioural Motor Impulsivity 9 items a = .76; 4 items a = .74a I act on the spur of the moment Motor I act ``on impulse" Motor I do things without thinking Motor I buy things on impulse Motor I spend or charge more than I earn Non-planning I say things without thinking Motor I am more interested in the present than the future Non-planning I change jobs Non-planning I have ``racing" thoughts Attentional Factor 3 (4.6% variance) Distractibility 6 items a = .66; 4 items a = .60a I find it hard to sit still for long periods of time I have outside thoughts when thinking I talk fast I am restless in class/groups I get easily bored when solving thought problems I walk and move fast Motor Attentional Motor Motor Attentional Motor

Items in bold are least likely to form part of the specific factor with loadings below .50. a All item-to-total correlations were positive.

2.01

2.20

2.38

3.12

2.18

3.41

F1P1 F1P2 F1P3 F1P1

1

.87 .84 .68

F1P2

1

.86

F1P3

cog

.82

.66

cog

1.49 1.47 1.43 1.62

1.98

1.91

F2P1 F2P2 F2P3 F3P1 F3P2 F3P3

1 1

.66 .70 .48 .28 1.25 .81 .74 .67

1.90

1.93

3.05

F2P1

1.91

F2P2

1

.76

F2P3

distract

beh

.51

beh

1.51

*Cog = Cognitive Planning Skills; Beh = Behavioural Impulsivity. Variances are shown above parcel items and standardized regression weights below.

Fig. 4. Confirmatory factor analysis of the BIS factor structure using adult women prisoners3 (n = 220: Study IV: RMSEA = .09; GFI = .96; CFI = .96; ECVI = .23).

*Cog = Cognitive Planning Skills; Distract = Distractibility; Beh = Behavioural Impulsivity. Variances are shown above parcel items and standardized regression weights below.

Fig. 3. Confirmatory factor analysis of the BIS-11 factor structure using adult men prisoners3 (n = 250: Study III: RMSEA = .047; GFI = .97; CFI = .98; ECVI = .32).

thus not supported. The current findings are consistent with previous research that has also failed to replicate the factors (e.g. Fossati

et al., 2001; von Diemen et al., 2007). It seems, therefore, that there remains difficulty in confirming the true structure of the concept of impulsivity (Whiteside & Lynam, 2001); with the current study demonstrating that this extends to samples where elevated levels

J.L. Ireland, J. Archer / Personality and Individual Differences 45 (2008) 286­292

291

of impulsivity are expected. The failure to confirm the original three factor solution is not surprising when it is considered that a confirmatory analysis was never undertaken with the original BIS-11 study. Parallel analysis in the current study (Study I and II) demonstrated, however, that a three factor solution was an optimal fit. The revised factors in the current study described Cognitive Planning Skills, Behavioural Impulsivity and Distractibility. All factors were easily interpretable and although the themes were similar to those proposed by Patton et al. (1995), the placement of items were markedly different. The cognitive planning skills factor, for example, appeared to comprise attention and non-planning items from the original factor structure of Patton et al. (1995), suggesting that a failure to plan incorporates attentional ability. Only the factor `behavioural impulsivity' demonstrated any clear convergence with the original BIS-11 solution with this comprised largely of motor items. The ensuing confirmatory analysis showed two interesting features: (1) the data fitted the proposed three factor model for men only when covariance between factors was allowed; and (2) for women, a two-factor solution was preferable. The first feature indicated that the three factors are not distinct, although they have been described as such in factor analytical studies (e.g. Spinella, 2007). The current study questions this and suggests instead that the factors simply represent related components of a higher order construct (i.e. impulsivity). Indeed, the second order model reported2 indicated that the three factors were underneath a single higher order factor of `impulsivity'. Such a model could not be tested for women since there were only two first order factors (with a minimum of three factors required). However, the results suggest that using the higher order factor of `impulsivity' may be preferable to adopting an individual factor approach. This is an important point in that there does appear to be distinction between some schools of study, with those focusing on aggression per se preferring to utilise the BIS-11 as a unidimensional measure whereas those focusing on general delinquency (including substance use, and gambling etc.) focus on factors. The current study suggests there may be value in using the BIS-11 both as a unidimensional and a multi-dimensional measure although it cannot be assumed that the three factor solution proposed by Patton et al. (1995) will automatically transfer. This appears to particularly be the case for women. Evidence for a different structure for men and women was not consistent with the prediction that the multi-dimensional structure of the BIS-11 would be replicated between sex. Although a three factor solution was a good fit for men, it was not for women. For women, the factor `distractibility' required removal before an acceptable fit could be indicated. Thus for men, although there are components of behaviour, cognition and planning skills evident, for women impulsivity is better described as comprising behaviour and planning skills. This does query the potential value in using impulsivity as a mediator between variables, such as using it to attempt to explain differences in aggression and sex (Campbell, 2006), and delinquency and sex (Rowe, Flannery, & Flannery, 1995). The current indicates that impulsivity does not present in a uniform fashion between sex. The failure to reliably confirm a factor structure for women in the current study may indicate that the BIS-11 has failed to adequately account for the possibility of sex differences in impulsivity. The current study is not, however, without its limitations. First, although the analyses included both exploratory and testing (confirmatory) analyses, it was correlational in nature with a selfselecting sample that had opted-in to complete the measure with no incentives for engagement offered. Second, although there was evidence of internal reliability with the BIS-11, the extent to which it is ecologically valid is an issue which falls outside the

scope of the current study, but would be worthy of further investigation. The current study does illustrate, however, that the BIS-II is an internally reliable measure when applied to a sample where impulsivity is expected to be at elevated levels. However, the current study questions the structure of impulsivity. This finding has important implications beyond the population used here to illustrate the problems in concept structure. Impulsivity has been used as a mediating variable to describe sex differences in aggression (Campbell, 2006), to describe delinquency, eating disorders, psychopathy, distress, substance use (Whiteside & Lynam, 2001) and dyslexia (Baker & Ireland, 2007), to name but a few. The argued components of impulsivity have also been used to describe the processes underlying various maladaptive behaviours, such as compulsive buying (Billieux et al., 2008). What the current study suggests, is that the concept of impulsivity presents with structural difficulties in measurement, and that although there is some evidence of heterogeneous qualities (i.e. a multi-dimensional structure), the actual components are unclear, making it difficult to find a common nosology across samples (Whiteside & Lynam, 2001). The current study also illustrates the importance of exploring and confirming measure structure, rather than assuming that a published structure will automatically apply (e.g. Spinella, 2007). This is a criticism which can be leveled at a range of impulsivity measures, and not just the BIS (e.g. Billieux et al., 2008). Indeed, impulsivity as a concept is poorly defined and thus it cannot be assumed that the factors proposed in any published measures are fixed, particularly when there has been almost a complete absence of attempts to rigidly explore and confirm structures using more refined analyses, such as confirmatory factor analysis. In essence this is perhaps the wider scientific relevance of the current study: there is a need for researchers to assure themselves more of the structure of the measures that they are employing, particularly if these are argued to comprise factors, and not simply assume factor presence following a simple exploratory structure analysis. A valuable focus for future research, therefore, would be the further development and refinement of measures of impulsivity, extending this to focus on assessing the structural validity of measures overall and any suggested components of such measures. Acknowledgements Thanks are extended to Christina Power for carrying out data collection and to Steve Lane for assistance with SEM queries. This research was financed by a grant from the Economic and Social Research Council (ESRC: No. RES: 000-22-0268). The views contained within this article are those of the researchers and not of the Grant Authority. References

Anderson, C. A., Benjamin, A. J., Wood, P. K., & Bonacci, A. M. (2006). Development and testing of the Velicer attitudes toward violence scale. Aggressive Behavior, 32, 122­136. Archer, J. A., Ireland, J. L., & Power, C. P. (2007). Differences between bullies and victims, and men and women, on aggression-related variables among prisoners. British Journal of Social Psychology, 46, 299­322. Baker, S. F., & Ireland, J. L. (2007). The link between dyslexic traits, executive functioning, impulsivity and social self-esteem among an offender and nonoffender sample. International Journal of Law and Psychiatry, 30, 492­503. Barratt, E. S. (1959). Anxiety and impulsiveness relative to psychomotor efficiency. Perceptual and Motor Skills, 9, 191­198. Baumeister, R. F., & Vohs, K. D. (2007). Handbook of self-regulation: Research, theory, and applications. New York: Guilford Press. Baylé, F. J., Bourdel, M. C., Caci, H., Gorwood, P., Chignon, J. M., Adés, J., et al. (2000). Factor analysis of French translation of the Barratt impulsivity scale (BIS-10). Canadian Journal of Psychiatry, 45, 156­165. Billieux, J., Rochat, L., Ribetez, M. M. L., & Van der Linden, M. (2008). Are all facets of impulsivity related to self-reported compulsive buying behaviour? Personality and Individual Differences, 44, 1432­1442.

292

J.L. Ireland, J. Archer / Personality and Individual Differences 45 (2008) 286­292 Jackson, R. L., Neumann, C. S., & Vitacco, M. J. (2007). Impulsivity, anger, and psychopathy: The moderating effect of ethnicity. Journal of Personality Disorder, 21, 289­304. Lejoyeux, M., Feuché, N., Loi, S., Solomon, J., & Adès, J. (1998). Impulse-control disorders in alcoholics are related to sensation seeking and not to impulsivity. Psychiatry Research, 81, 149­155. Patton, J. H., Stanford, M. S., & Barratt, E. S. (1995). Factor structure of the Barratt Impulsivity Scale. Journal of Clinical Psychology, 51, 768­774. Raust, A., Slama, F., Mathieu, F., Roy, I., Chenu, A., Koncke, D., et al. (2007). Prefrontal cortex dysfunction in patients with suicidal behavior. Psychological Medicine, 37, 411­419. Rowe, D. C., Flannery, A. T., & Flannery, D. J. (1995). Sex differences in crime: Do means and within-sex variation have similar causes? Journal of Research in Crime and Delinquency, 32, 84­100. Smith, P., Waterman, M., & Ward, N. (2006). Driving aggression in forensic and nonforensic populations: Relationships to self-reported levels of aggression, anger and impulsivity. British Journal of Psychology, 97, 387­403. Spinella, M. (2007). Normative data and a short form of the Barratt Impulsiveness Scale. International Journal of Neuroscience, 117, 359­368. von Diemen, L., Szobot, C. M., Kessler, F., & Pechansky, F. (2007). Adaptation and construct validation of the Barratt Impulsiveness Scale (BIS-11) to Brazilian Portuguese for use in adolescents. Revista Brasileira de Psiquiatria, 29, 153­156. Whiteside, S. P., & Lynam, D. R. (2001). The five factor model and impulsivity: Using a structural model of personality to understand impulsivity. Personality and Individual Differences, 30, 669­689.

Campbell, A. (2006). Sex differences in direct aggression: What are the psychological mediators? Aggression and Violent Behaviour: A Review Journal, 11, 237­264. Carrillo-de-la-Peña, M. T., Otero, J. M., & Romero, E. (1993). Comparison among various methods of assessment of impulsiveness. Perception Motor Skills, 77, 567­575. Cohen, R. A., Rosenbaum, A., Kane, R. L., Warnken, W. J., & Benjamin, S. (1999). Neuropsychological correlates of domestic violence. Violence and Victims, 14, 397­411. Dom, G., De Wilde, B., Hulstijn, W., van den Brink, W., & Sabbe, B. (2006). Behavioural aspects of impulsivity in alcoholics with and without a cluster-B personality disorder. Alcohol and Alcoholism, 41, 412­420. Enticott, P. G., Ogloff, J. R., Bradshaw, J. L., & Fitzgerald, P. B. (2008). Cognitive inhibitory control and self-reported impulsivity among violent offenders with schizophrenia. Journal of Clinical and Experimental Neuropsychology, 30, 157­162. Fossati, A., Barratt, E. S., Borroni, S., Villa, D., Grazioli, F., & Maffei, C. (2007). Impulsivity, aggressiveness, and DSM-IV personality disorders. Psychiatry Research, 149, 157­167. Fossati, A., Di Ceglie, A., Acquarini, E., & Barratt, E. S. (2001). Psychometric properties of an Italian version of the Barratt Impulsiveness Scale-11 (BIS-11) in nonclinical subjects. Journal of Clinical Psychology, 57, 815­828. Ireland, J. L., Archer, J., & Power, C. L. (2007). Characteristics of male and female prisoners involved in bullying behavior. Aggressive Behavior, 33, 220­229.

Information

doi:10.1016/j.paid.2008.04.012

7 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

189078


You might also be interested in

BETA
doi:10.1016/j.paid.2008.04.012