- Email: [email protected]
BAS scales to measure behavioural inhibition and behavioural activation: Factor structure, validity and norms in a large community sample
PERGAMON
Personality and Individual Differences 26 (1999) 49±58
Using the BIS/BAS scales to measure behavioural inhibition and behavioural activation: Factor structure, validity and norms in a large community sample A. F. Jorm *, H. Christensen, A. S. Henderson, P. A. Jacomb, A. E. Korten, B. Rodgers NHMRC Psychiatric Epidemiology Research Centre, The Australian National University, Canberra 0200, Australia Received 15 December 1997
Abstract The Behavioural Inhibition System and Behavioural Activation System (BIS/BAS) scales of Carver and White (1994) were used in an Australian community sample of 2725 individuals aged 18±79. Factor analysis of the BIS/BAS items supported the 4-factor structure found by Carver and White, as well as a 2-factor structure re¯ecting separate behavioural inhibition and behavioural activation systems. The BIS scale was related to neuroticism and negative aectivity, while the BAS scale was related to extraversion and positive aectivity. The BIS scale was less correlated with anxiety and depression symptoms than are neuroticism and negative aectivity scales, probably because it is designed to measure predisposition to anxiety rather than the experience of anxiety. BIS scores were higher in females, while the BAS subscales showed a more complex pattern, with reward responsiveness scores higher in females and drive scores higher in males. Both BIS and BAS scores were lower in older age groups, suggesting the possibility that the behavioural inhibition and behavioural activation systems become less responsive with age. # 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction Based on ®ndings from animal learning paradigms, Gray (1987) has distinguished three emotional systems: (1) a behavioural inhibition system which mediates responses to conditioned signals of punishment; (2) a behavioural activation or approach system which mediates responses to conditioned signals of reward and (3) a ®ght±¯ight system which mediates responses to unconditioned punishing events. Gray (1987) has also argued that major * Corresponding author. Fax: +61-2-6249-0733; E-mail: [email protected] S0191-8869/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved. PII: S 0 1 9 1 - 8 8 6 9 ( 9 8 ) 0 0 1 4 3 - 3
50
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
personality dimensions may re¯ect individual dierences in the functioning of these emotional systems. In particular, the behavioural inhibition system may be responsible for individual dierences in anxiety and the behavioural activation system for individual dierences in impulsivity. Gray (1987) has speculated on how his systems might relate to Eysenck's three personality factors. He suggested that behavioural inhibition corresponds to an axis running through high neuroticism, low extraversion and low psychoticism, while behavioural activation corresponds to an axis running through high neuroticism, high extraversion and high psychoticism. For some years, a major diculty in evaluating the theory in humans was that there was no personality scale directly measuring Gray's three systems. To overcome this lack, Wilson et al. (1989) developed the Gray±Wilson Personality Questionnaire to measure the human equivalents of the animal behaviour paradigms on which Gray's theory is based. However, the pattern of scale intercorrelations and a factor analysis of the items did not support the predictions from the theory (Wilson et al., 1989, 1990). While these results are disappointing, it could be argued that the validity of a questionnaire measure of Gray's systems should rest on experimental con®rmation rather than on factor analytic criteria. Another questionnaire measure based on Gray's theory is the Generalized Reward and Punishment Expectancy Scales (GRAPES) of Ball and Zuckerman (1990). A factor analysis of this questionnaire showed two independent factors re¯ecting reward and punishment expectancies. The Reward Expectancy scale was strongly related to extraversion and had a weaker negative correlation with neuroticism, whereas Gray's theory would predict positive correlations with both extraversion and neuroticism. Similarly, the Punishment Expectancy scale was related to neuroticism, but not to extraversion, whereas Gray's theory would predict positive correlations with neuroticism and a negative correlation with extraversion. The GRAPES does not appear to have been used beyond the original study of Ball and Zuckerman (1990). There have also been two other questionnaires developed to measure behavioural inhibition, unaccompanied by a measure of behavioural activation (Torrubia and TobenÄa, 1984; MacAndrew and Steele, 1991). Both measures show strong correlations with neuroticism and smaller negative correlations with extraversion. Neither scale has come into general use. Carver and White (1994) have made the most recent attempt to develop a questionnaire measure of two of Gray's three systems. Rather than trying to write questions to re¯ect animal behaviour paradigms, they developed the Behavioural Inhibition System and Behavioural Activation System (BIS/BAS) questionnaire using Gray's general conceptualization of the two systems as a guide. The BIS scale attempts to measure concern over the possibility of a bad occurrence and sensitivity to such events when they do occur. It diers from existing anxiety scales in that it does not ask about how often the respondent experiences anxiety, but whether they would be likely to experience anxiety in certain situations. Carver and White (1994) argued that a person who is prone to anxiety might arrange his or her life so as to avoid anxiety-provoking situations, in which case questions about the experience of anxiety would not be a good measure of predisposition to anxiety. They criticized earlier behavioural inhibition scales (Torrubia and TobenÄa, 1984; MacAndrew and Steele, 1991) for focussing too much on experienced anxiety. Carver and White (1994) measured the behavioural activation system using three subscales: reward responsiveness, drive and fun seeking. Although the BIS/
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
51
BAS yields separate subscale scores for these components of behavioural activation, they are intercorrelated, as would be expected if they all re¯ected the same emotional system. Using a sample of 732 university students, Carver and White (1994) did a factor analysis of the BIS/ BAS items and found that the items from the four scales loaded on the expected factors. In various other student samples, they found that the BIS scale correlated most highly with measures of trait anxiety, negative aectivity, negative temperament, harm avoidance and (surprisingly) reward dependence, while the BAS subscales correlated most highly with measures of extraversion, positive aectivity and positive temperament. The scales were also validated using experimental situations in which a punishment or a reward were anticipated. Individuals high in BIS sensitivity were found to react with greater nervousness when punishment was anticipated, while those high in BAS sensitivity reacted with greater happiness when a reward was anticipated. Other than Carver and White's (Carver and White, 1994) initial work to develop the BIS/BAS, few other published studies appear to have used the scales so far. Harmon-Jones and Allen (1997) used the scales in a study of behavioural activation and frontal EEG asymmetry in university students. Like Carver and White (1994), they found correlations of the BIS scale with negative aect and the BAS scale with positive aect. Pickering and colleagues (Pickering, 1997; Pickering et al., 1997) used the scales to examine the eects of reward on card-sorting rate. They found a positive correlation between the Reward Responsiveness scale and the eects of reward, supporting the validity of this scale. We have used the BIS/BAS in a large community survey designed to examine molecular genetic associations with personality and psychiatric symptoms. Given that Gray has hypothesized speci®c neurophysiological underpinnings for his emotional systems, the BIS/BAS seemed a good choice in a study aimed at ®nding molecular genetic associations. Another potential advantage of the BIS/BAS for epidemiological and social surveys is its brevity (only 24 items to measure 4 scales) compared to the 100-item Tridimensional Personality Questionnaire (Cloninger et al., 1991) which is based on similar concepts. Here we report ®ndings on the BIS/BAS's factor structure, reliability, and convergent and divergent validity and we present some norms by age and gender.
2. Methods 2.1. Sample Participants were recruited from the Electoral Roll for Canberra, Australia. Enrolment to vote is compulsory for all Australian citizens aged 18 or over. Interviews were completed with 2725 individuals, representing a response rate of 67% from those who were contactable. The achieved sample was 52% female. The age breakdown was 22% aged 18±29, 23% aged 30±39, 27% aged 40±49, 16% aged 50±59, 8% aged 69±69 and 5% aged 70±79. Compared to the Canberra population, the age group 18±29 was under-represented (22% vs 27%) and the age group 40±49 over-represented (27% vs 23%), but otherwise the sample closely matched the age distribution in the population.
52
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
2.2. Questionnaire Participants were asked to complete a questionnaire which covered socio-demographic characteristics, personality, anxiety and depression symptoms, alcohol abuse, life events, social support and childhood experiences. This was done under the supervision of a professional interviewer. The participants were also asked to provide a cheek swab from which DNA could be extracted. The following components of the questionnaire are relevant to the present paper: the BIS/BAS scales (Carver and White, 1994); the Extraversion, Neuroticism and Psychoticism scales of the short form of the Eysenck Personality Questionnaire-Revised (EPQ-R) (Eysenck et al., 1985); a short form of the Positive and Negative Aect Schedule (PANAS) (Kercher, 1992) in which participants had to rate their feelings ``in general''; the anxiety and depression scales of Goldberg et al. (1988); the anxiety and depression scales of the Delusions-SymptomsStates Inventory (DSSI) (Bedford et al., 1976) and the Alcohol Use Disorders Identi®cation Test (AUDIT) (Saunders et al., 1993). 2.3. Survey procedure Persons selected at random from the Electoral Roll were sent a letter informing them about the survey and saying that an interviewer would contact them soon to see if they wanted to participate. If a person agreed to participate, the interviewer visited them at some convenient location, usually the participant's home. The interviewer took a cheek swab and then gave the participant the questionnaire to self-complete. To encourage frank responses, the questionnaire had an ID number, but not the participant's name. After completing the questionnaire, the participant sealed it in an envelope and gave it to the interviewer who never saw the answers given.
3. Results 3.1. Factor analysis of BIS/BAS items A principal component analysis was carried out on the data from the 2684 participants who completed all the BIS/BAS items. Following the procedure of Carver and White (1994), an oblique rotation was carried out on the four factors with eigenvalues greater than 1. These four factors accounted for 51% of the variance. The resulting factor pattern matrix is shown in Table 1. The ®rst factor corresponded to fun seeking, the second to behavioural inhibition, the third to drive and the fourth to reward responsiveness. The items loaded on the expected factors with some minor exceptions. One of the reward responsiveness items had low loadings on all four factors, while one of the fun seeking items loaded on both the fun seeking and reward responsiveness factors. The reward responsiveness, drive and fun seeking factors were intercorrelated 0.20±0.29 as would be expected if they all re¯ected aspects of behavioural activation. While a four-factor solution was consistent with the factor analysis of Carver and White (1994), a scree plot suggested that a two-factor solution would be more appropriate. The eigenvalues of the ®rst six principal components were 4.59. 3.19, 1.27, 1.14, 0.97 and 0.91,
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
53
Table 1 Factor loadings for BIS/BAS items 2-Factor solution
4-Factor solution Item BIS If I think something unpleasant is going to happen I usually get pretty ``worked up'' I worry about making mistakes Criticism or scolding hurts me quite a bit I feel pretty worried or upset when I think or know somebody is angry at me Even if something bad is about to happen to me, I rarely experience fear or nervousness I feel worried when I think I have done poorly at something I have very few fears compared to my friends BAS reward responsiveness When I get something I want, I feel excited and energized When I'm doing well at something, I love to keep at it When good things happen to me, it aects me strongly It would excite me to win a contest When I see an opportunity for something I like, I get excited right away BAS drive When I want something, I usually go all-out to get it I go out of my way to get things I want If I see a chance to get something I want, I move on it right away When I go after something I use a ``no holds barred'' approach BAS fun seeking I will often do things for no other reason than that they might be fun I crave excitement and new sensations I'm always willing to try something new if I think it will be fun I often act on the spur of the moment
1
2
0.08
0.75 ÿ 0.08 ÿ 0.07
ÿ 0.15 ÿ 0.09 0.01
3
0.70 ÿ 0.06 0.72 ÿ 0.05 0.75 0.03
4
1
2
0.08
0.73
0.13 0.03 0.01 ÿ 0.02 0.04 0.01
0.69 0.71 0.75
ÿ 0.29 ÿ0.43 ÿ 0.20
0.29
0.13 ÿ 0.43
ÿ 0.13 0.65 ÿ 0.06 ÿ 0.08 ÿ0.42 ÿ 0.21
0.21 0.36
0.08 0.65 0.32 ÿ 0.40
0.26 ÿ 0.12 0.13 0.32 0.34
0.50 0.69 0.47 0.41 0.23
0.55 0.36 0.34 0.33 0.62
0.20 0.12 0.42 0.23 0.28
ÿ 0.85 0.02 ÿ 0.82 ÿ 0.05 ÿ 0.69 0.06 ÿ 0.67 ÿ 0.08
0.70 0.63 0.70 0.71
ÿ 0.03 ÿ 0.12 ÿ 0.06 ÿ 0.12
0.13 ÿ 0.10 0.04 ÿ 0.06 0.35 0.01 0.14 0.14 0.28 ÿ 0.31
ÿ 0.05 0.09 ÿ 0.04 0.01 0.13 0.02 0.29 ÿ0.03
0.70 ÿ0.20 ÿ 0.04 0.09 0.63 0.04 ÿ 0.29 0.01 0.46 ÿ0.23 ÿ 0.02 0.42 0.59 0.01 ÿ 0.24 ÿ 0.07
0.56 ÿ 0.14 0.67 0.04 0.57 ÿ 0.15 0.56 ÿ 0.00
Loadings of r0.4 shown in italics.
respectively. Table 1 also shows the results of the two-factor solution. The two factors accounted for 39% of the variance. The ®rst factor corresponded to behavioural activation and the second to behavioural inhibition. All items loaded as expected except that three of the reward responsiveness items had low loadings on the behavioural activation factor. The two factors were correlated only 0.07. 3.2. Reliability Cronbach's alpha was 0.76 for BIS, 0.83 for BAS, 0.65 for reward responsiveness, 0.80 for drive and 0.70 for fun seeking.
54
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
3.3. Scale inter-correlations If participants had missing data on one item of a scale their score was pro-rated by taking the mean of the items responded to and multiplying by the number of items in the full scale. However, if they had more missing data the scale was regarded as missing. The BIS scale correlated 0.06 with BAS, ÿ0.05 with drive, ÿ0.03 with fun seeking and 0.27 with reward responsiveness. Drive correlated 0.52 with fun seeking and 0.42 with reward responsiveness, while fun seeking correlated 0.45 with reward responsiveness. 3.4. Age and gender norms Table 2 shows the means and standard deviations for the scales by age group and gender. Multiple regression analyses were carried out to see if there were age and gender dierences, with the P < 0.01 level used for statistical signi®cance. Age-by-gender interactions were tested by entering the product of age and gender after the main eects had been entered (Cohen and Cohen, 1975). It can be clearly seen in Table 2 that older groups scored lower on all the BIS/ BAS scales. Females scored higher on the BIS and reward responsiveness, while males were higher on drive. There was no gender dierence on the BAS total. There were signi®cant ageby-gender interaction eects for the BIS and drive scales, due to bigger gender dierences in younger age groups for BIS and bigger gender dierences in older age groups for drive. It can also be seen in Table 2 that the standard deviations of all the BAS scales tend to increase in older age groups and these dierences were statistically signi®cant. Table 2 Means (and SD's) for BIS/BAS scales by age and gender Group
N*
BIS
BAS
Males 18±29 30±39 40±49 50±59 60±69 70±79
277 276 336 238 112 54
19.3 20.0 20.0 19.9 19.9 18.8
(3.5) (3.4) (3.7) (3.4) (2.8) (3.3)
40.0 38.1 37.1 36.5 35.2 34.3
(4.8) (5.2) (5.5) (5.6) (6.6) (6.9)
16.8 16.5 16.1 16.3 15.5 15.0
(1.9) (2.0) (2.1) (2.0) (2.4) (2.7)
10.9 10.3 10.3 10.0 9.8 9.6
(2.3) (2.4) (2.4) (2.5) (2.8) (2.7)
12.3 11.3 10.6 10.2 9.9 9.7
(2.2) (2.3) (2.2) (2.4) (2.8) (2.6)
Females 18±29 30±39 40±49 50±59 60±69 70±79
312 341 393 208 99 71
22.0 21.7 21.4 20.9 20.1 19.8
(3.4) (3.5) (3.4) (3.4) (3.4) (3.6)
40.2 38.1 37.4 36.1 35.5 33.6
(5.1) (5.4) (5.6) (5.8) (6.1) (6.7)
17.6 17.0 16.8 16.5 16.4 16.0
(1.9) (1.9) (2.0) (2.2) (2.5) (2.6)
10.7 10.1 9.8 9.3 8.9 8.2
(2.4) (2.6) (2.7) (2.7) (2.8) (3.1)
11.9 11.1 10.8 10.3 10.2 9.5
(2.2) (2.2) (2.3) (2.5) (2.4) (2.5)
*Numbers vary slightly from scale to scale because of missing data.
Reward
Drive
Fun
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
55
Table 3 Correlations of BIS/BAS scales with personality and psychiatric symptom measures Scale
BIS
BAS
Reward
Drive
Fun
EPQ-R Extraversion Neuroticism Psychoticism
ÿ 0.19 0.64 ÿ 0.31
0.44 0.09 0.18
0.28 0.21 ÿ 0.02
0.32 0.01 0.20
0.46 0.02 0.23
PANAS Positive Aect Negative Aect
ÿ 0.12 0.48
0.46 0.04
0.33 0.13
0.39 ÿ 0.01
0.37 ÿ 0.00
DSSI Anxiety Depression
0.44 0.35
0.08 0.02
0.14 0.06
0.03 ÿ 0.01
0.03 0.01
Goldberg scales Anxiety Depression
0.42 0.36
0.10 0.04
0.16 0.10
0.04 ÿ 0.02
0.06 0.03
AUDIT
0.02
0.12
0.02
0.11
0.16
P < 0.01 for all r'sr0.05 and P < 0.001 for all r'sr0.07.
3.5. Correlations with other scales Table 3 shows the correlations with other personality and psychiatric symptom scales. It can be seen that BIS was related most strongly to Neuroticism, Negative Aect and anxiety and depressive symptoms, but there was also a small negative correlation with psychoticism. BAS was related most strongly to Extraversion and Positive Aect, but there were also small correlations with psychoticism and the AUDIT. All the BAS subscales were related to extraversion and positive aect, but the reward responsiveness scale also had a small correlation with neuroticism, while the drive and fun seeking scales had small correlations with psychoticism and the AUDIT. To get a clearer picture of how the BIS/BAS relates to the other personality scales, a principal component analysis with oblique rotation was carried out on the BIS/BAS, EPQ-R and PANAS subscales. A scree plot suggested either a 2-factor or a 3-factor solution. The 3factor solution was found to be more satisfactory because the Psychoticism scale did not load highly on either factor in the 2-factor solution. The 3-factor solution accounted for 68% of the variance. Table 4 shows the results. The BAS subscales, extraversion and positive aect loaded highly on the ®rst factor which could be labelled ``extraversion/positive aectivity''. The BIS scale, neuroticism and negative aect loaded highly on the second factor which could be labelled ``neuroticism/negative aectivity''. The third factor had a high psychoticism loading and a smaller negative loading for BIS. Factors 1 and 2 were correlated ÿ0.02, factors 1 and 3 0.10 and factors 2 and 3 ÿ0.12.
56
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58 Table 4 Loadings from factor analysis of personality scales Scale
Factor 1
Factor 2
Factor 3
BIS BAS reward BAS drive BAS fun Extraversion Neuroticism Psychoticism Positive aect Negative aect
0.02 0.72 0.71 0.76 0.69 ÿ 0.03 0.06 0.71 ÿ 0.03
0.76 0.31 0.07 0.08 ÿ 0.25 0.91 ÿ 0.00 ÿ 0.16 0.83
ÿ0.34 ÿ0.20 0.20 0.21 ÿ0.06 0.04 0.94 ÿ0.06 0.13
4. Discussion The factor analyses largely supported the factor structure reported by Carver and White (1994). The four-factor solution corresponded to the four BIS/BAS scales, while the two-factor solution gave behavioural inhibition and behavioural activation factors. However, some of the reward responsiveness items did not load unequivocally on the expected factors. In particular, the ®fth item (``When I see an opportunity for something I like, I get excited right away'') did not load speci®cally on the reward responsiveness factor, although it was adequate as a general BAS item. Other ®ndings pointed to limitations of the Reward Responsiveness scale. It had the lowest reliability of all the BIS/BAS scales and it correlated with BIS and neuroticism as well as with the other BAS subscales. The pattern of correlations of the BIS/BAS with other personality scales suggests that the BIS scale is largely a measure of a neuroticism/negative aectivity super-factor, while the BAS scale is a measure of an extraversion/positive aectivity super-factor. However, Carver and White (1994) claimed that the BIS scale is not so closely aligned with other anxiety measures because the items are worded to re¯ect propensity to anxiety given a particular situation, rather than how often anxiety is experienced. Indeed, the BIS scale had lower correlations with measures of current anxiety and depression (r's of 0.35±0.44) than either the Neuroticism or Negative Aect scales (r's of 0.56±0.62 and 0.47±0.59, respectively), supporting the interpretation that it is less a measure of experienced anxiety. There were some interesting age and gender dierences on the BIS/BAS scales. Both BIS and BAS scores were lower in older age groups, consistent with ®ndings for Neuroticism and Extraversion (Eysenck et al., 1985). Whether these age group dierences re¯ect ageing or cohort eects is impossible to determine from a cross-sectional study such as this. However, if the dierences do re¯ect ageing, they imply that the emotional systems measured by the BIS/ BAS become less responsive with age. Another interesting age group dierence was evident in the standard deviations of the BAS scale and its subscales, which were signi®cantly larger in older age groups. It is clear from an inspection of the means and the possible range of scale scores that these dierences in dispersion are not simply due to ceiling or ¯oor eects. They remain a puzzle, but are
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
57
consistent with other evidence that older age groups tend to show increased variability in a range of biological and psychological domains, including personality characteristics (Nelson and Dannefer, 1992). Gender dierences were more complex. Females had higher mean scores on the BIS scale which is consistent with the ®ndings with the neuroticism scale. Furthermore, the gender dierence on the BIS scale was greater in younger age groups than in older ones, which is also consistent with ®ndings on neuroticism (Jorm, 1987). While there was no dierence on the overall BAS score, females scored higher on the reward responsiveness subscale and males higher on the drive subscale. The gender dierence on the reward responsiveness scale may be due to the fact that it has some neuroticism/negative aectivity component (on which females tend to score higher). The higher male scores on drive were age group speci®c, being largely con®ned to the older age groups. Again, whether this is an ageing or a cohort eect is impossible to determine. Gray (1987) has speculated on the relationships of his emotional systems to Eysenck's personality traits. Assuming that Carver and White's (1994) scales are adequate measures of behavioural inhibition and activation, the pattern of correlations found here does not entirely support Gray's views. Behavioural inhibition is strongly associated with neuroticism, but only weakly with low extraversion and low psychoticism. Behavioural activation is strongly associated with extraversion, but only weakly with psychoticism and neuroticism. In short, behavioural inhibition is primarily associated with neuroticism and behavioural activation with extraversion. Ball and Zuckerman (1990) had similar ®ndings with their GRAPES questionnaire. However, the location in personality space of behavioural inhibition and behavioural activation can only be ultimately settled by linking indicators of these brain systems with personality measures. This study is the ®rst to report the psychometric properties of the BIS/BAS scales in a sample of the general population. The conclusion is that they are a valid and practical measure of two personality traits relevant for molecular genetics research.
Acknowledgements Thanks to Karen Maxwell and Michelle Powderly for assistance with managing the survey and to Simon Easteal for discussions.
References Ball, S. A., & Zuckerman, M. (1990). Sensation seeking, Eysenck's personality dimensions and reinforcement sensitivity in concept formation. Personality and Individual Dierences, 11, 343±353. Bedford, A., Foulds, G. A., & Sheeld, B. F. (1976). A new personal disturbance scale (DSSI/sAD). British Journal of Social and Clinical Psychology, 15, 387±394. Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and aective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67, 319±333. Cloninger, C. R., Przybeck, T. R., & Svrakic, D. M. (1991). The Tridimensional Personality Questionnaire: US normative data. Psychological Reports, 69, 1047±1057.
58
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
Cohen, J. & Cohen, P. (1975). Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, New Jersey: Lawrence Erlbaum. Eysenck, S. B. G., Eysenck, H. J., & Barrett, P. (1985). A revised version of the psychoticism scale. Personality and Individual Dierences, 6, 21±29. Goldberg, D., Bridges, K., Duncan-Jones, P., & Grayson, D. (1988). Detecting anxiety and depression in general medical settings. British Medical Journal, 297, 897±899. Gray, J. A. (1987). The neuropsychology of emotion and personality. In S. M. Stahl, S. D. Iversen & E. C. Goodman (Eds.), Cognitive neurochemistry (pp. 171±190). Oxford: Oxford University Press. Harmon-Jones, E., & Allen, J. J. B. (1997). Behavioral activation sensitivity and resting frontal EEG asymmetry: Covariation of putative indicators related to risk for mood disorders. Journal of Abnormal Psychology, 106, 159±163. Jorm, A. F. (1987). Sex dierences in neuroticism: A quantitative synthesis of published research. Australian and New Zealand Journal of Psychiatry, 21, 501±506. Kercher, K. (1992). Assessing subjective well-being in the old-old. Research on Aging, 14, 131±168. MacAndrew, C., & Steele, T. (1991). Gray's behavioral inhibition system: A psychometric evaluation. Personality and Individual Dierences, 12, 157±171. Nelson, E. A. , & Dannefer, D. (1992). Aged heterogeneity: Fact or ®ction? The fate of diversity in gerontological research. Gerontologist, 32, 17±23. Pickering, A. D. (1997). The conceptual nervous system and personality: From Pavlov to neural networks. European Psychologist, 2, 139±163. Pickering, A. D., Corr, P. J., Powell, J. H., Kumari, V., Thornton, J. C. & Gray, J. A. (1997). Individual dierences in reactions to reinforcing stimuli are neither black nor white: To what extent are they Gray? In H. Nyborg (Ed.), The scienti®c study of human nature: Tribute to Hans J. Eysenck at eighty (pp. 36±67). London: Elsevier Science. Saunders, J. B., Aasland, O. G., Babor, T. F., De La Fuente, J. R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identi®cation Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption. II. Addiction, 88, 791±804. Torrubia, R., & TobenÄa, A. (1984). A scale for the assessment of ``susceptibility to punishment'' as a measure of anxiety: Preliminary ®ndings. Personality and Individual Dierences, 5, 371±375. Wilson, G. D., Barrett, P. T., & Gray, J. A. (1989). Human reactions to reward and punishment: A questionnaire examination of Gray's personality theory. British Journal of Psychology, 80, 509±515. Wilson, G. D., Gray, J. A., & Barrett, P. T. (1990). A factor analysis of the Gray±Wilson Personality Questionnaire. Personality and Individual Dierences, 11, 1037±1045.
Personality and Individual Differences 26 (1999) 49±58
Using the BIS/BAS scales to measure behavioural inhibition and behavioural activation: Factor structure, validity and norms in a large community sample A. F. Jorm *, H. Christensen, A. S. Henderson, P. A. Jacomb, A. E. Korten, B. Rodgers NHMRC Psychiatric Epidemiology Research Centre, The Australian National University, Canberra 0200, Australia Received 15 December 1997
Abstract The Behavioural Inhibition System and Behavioural Activation System (BIS/BAS) scales of Carver and White (1994) were used in an Australian community sample of 2725 individuals aged 18±79. Factor analysis of the BIS/BAS items supported the 4-factor structure found by Carver and White, as well as a 2-factor structure re¯ecting separate behavioural inhibition and behavioural activation systems. The BIS scale was related to neuroticism and negative aectivity, while the BAS scale was related to extraversion and positive aectivity. The BIS scale was less correlated with anxiety and depression symptoms than are neuroticism and negative aectivity scales, probably because it is designed to measure predisposition to anxiety rather than the experience of anxiety. BIS scores were higher in females, while the BAS subscales showed a more complex pattern, with reward responsiveness scores higher in females and drive scores higher in males. Both BIS and BAS scores were lower in older age groups, suggesting the possibility that the behavioural inhibition and behavioural activation systems become less responsive with age. # 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction Based on ®ndings from animal learning paradigms, Gray (1987) has distinguished three emotional systems: (1) a behavioural inhibition system which mediates responses to conditioned signals of punishment; (2) a behavioural activation or approach system which mediates responses to conditioned signals of reward and (3) a ®ght±¯ight system which mediates responses to unconditioned punishing events. Gray (1987) has also argued that major * Corresponding author. Fax: +61-2-6249-0733; E-mail: [email protected] S0191-8869/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved. PII: S 0 1 9 1 - 8 8 6 9 ( 9 8 ) 0 0 1 4 3 - 3
50
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
personality dimensions may re¯ect individual dierences in the functioning of these emotional systems. In particular, the behavioural inhibition system may be responsible for individual dierences in anxiety and the behavioural activation system for individual dierences in impulsivity. Gray (1987) has speculated on how his systems might relate to Eysenck's three personality factors. He suggested that behavioural inhibition corresponds to an axis running through high neuroticism, low extraversion and low psychoticism, while behavioural activation corresponds to an axis running through high neuroticism, high extraversion and high psychoticism. For some years, a major diculty in evaluating the theory in humans was that there was no personality scale directly measuring Gray's three systems. To overcome this lack, Wilson et al. (1989) developed the Gray±Wilson Personality Questionnaire to measure the human equivalents of the animal behaviour paradigms on which Gray's theory is based. However, the pattern of scale intercorrelations and a factor analysis of the items did not support the predictions from the theory (Wilson et al., 1989, 1990). While these results are disappointing, it could be argued that the validity of a questionnaire measure of Gray's systems should rest on experimental con®rmation rather than on factor analytic criteria. Another questionnaire measure based on Gray's theory is the Generalized Reward and Punishment Expectancy Scales (GRAPES) of Ball and Zuckerman (1990). A factor analysis of this questionnaire showed two independent factors re¯ecting reward and punishment expectancies. The Reward Expectancy scale was strongly related to extraversion and had a weaker negative correlation with neuroticism, whereas Gray's theory would predict positive correlations with both extraversion and neuroticism. Similarly, the Punishment Expectancy scale was related to neuroticism, but not to extraversion, whereas Gray's theory would predict positive correlations with neuroticism and a negative correlation with extraversion. The GRAPES does not appear to have been used beyond the original study of Ball and Zuckerman (1990). There have also been two other questionnaires developed to measure behavioural inhibition, unaccompanied by a measure of behavioural activation (Torrubia and TobenÄa, 1984; MacAndrew and Steele, 1991). Both measures show strong correlations with neuroticism and smaller negative correlations with extraversion. Neither scale has come into general use. Carver and White (1994) have made the most recent attempt to develop a questionnaire measure of two of Gray's three systems. Rather than trying to write questions to re¯ect animal behaviour paradigms, they developed the Behavioural Inhibition System and Behavioural Activation System (BIS/BAS) questionnaire using Gray's general conceptualization of the two systems as a guide. The BIS scale attempts to measure concern over the possibility of a bad occurrence and sensitivity to such events when they do occur. It diers from existing anxiety scales in that it does not ask about how often the respondent experiences anxiety, but whether they would be likely to experience anxiety in certain situations. Carver and White (1994) argued that a person who is prone to anxiety might arrange his or her life so as to avoid anxiety-provoking situations, in which case questions about the experience of anxiety would not be a good measure of predisposition to anxiety. They criticized earlier behavioural inhibition scales (Torrubia and TobenÄa, 1984; MacAndrew and Steele, 1991) for focussing too much on experienced anxiety. Carver and White (1994) measured the behavioural activation system using three subscales: reward responsiveness, drive and fun seeking. Although the BIS/
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
51
BAS yields separate subscale scores for these components of behavioural activation, they are intercorrelated, as would be expected if they all re¯ected the same emotional system. Using a sample of 732 university students, Carver and White (1994) did a factor analysis of the BIS/ BAS items and found that the items from the four scales loaded on the expected factors. In various other student samples, they found that the BIS scale correlated most highly with measures of trait anxiety, negative aectivity, negative temperament, harm avoidance and (surprisingly) reward dependence, while the BAS subscales correlated most highly with measures of extraversion, positive aectivity and positive temperament. The scales were also validated using experimental situations in which a punishment or a reward were anticipated. Individuals high in BIS sensitivity were found to react with greater nervousness when punishment was anticipated, while those high in BAS sensitivity reacted with greater happiness when a reward was anticipated. Other than Carver and White's (Carver and White, 1994) initial work to develop the BIS/BAS, few other published studies appear to have used the scales so far. Harmon-Jones and Allen (1997) used the scales in a study of behavioural activation and frontal EEG asymmetry in university students. Like Carver and White (1994), they found correlations of the BIS scale with negative aect and the BAS scale with positive aect. Pickering and colleagues (Pickering, 1997; Pickering et al., 1997) used the scales to examine the eects of reward on card-sorting rate. They found a positive correlation between the Reward Responsiveness scale and the eects of reward, supporting the validity of this scale. We have used the BIS/BAS in a large community survey designed to examine molecular genetic associations with personality and psychiatric symptoms. Given that Gray has hypothesized speci®c neurophysiological underpinnings for his emotional systems, the BIS/BAS seemed a good choice in a study aimed at ®nding molecular genetic associations. Another potential advantage of the BIS/BAS for epidemiological and social surveys is its brevity (only 24 items to measure 4 scales) compared to the 100-item Tridimensional Personality Questionnaire (Cloninger et al., 1991) which is based on similar concepts. Here we report ®ndings on the BIS/BAS's factor structure, reliability, and convergent and divergent validity and we present some norms by age and gender.
2. Methods 2.1. Sample Participants were recruited from the Electoral Roll for Canberra, Australia. Enrolment to vote is compulsory for all Australian citizens aged 18 or over. Interviews were completed with 2725 individuals, representing a response rate of 67% from those who were contactable. The achieved sample was 52% female. The age breakdown was 22% aged 18±29, 23% aged 30±39, 27% aged 40±49, 16% aged 50±59, 8% aged 69±69 and 5% aged 70±79. Compared to the Canberra population, the age group 18±29 was under-represented (22% vs 27%) and the age group 40±49 over-represented (27% vs 23%), but otherwise the sample closely matched the age distribution in the population.
52
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
2.2. Questionnaire Participants were asked to complete a questionnaire which covered socio-demographic characteristics, personality, anxiety and depression symptoms, alcohol abuse, life events, social support and childhood experiences. This was done under the supervision of a professional interviewer. The participants were also asked to provide a cheek swab from which DNA could be extracted. The following components of the questionnaire are relevant to the present paper: the BIS/BAS scales (Carver and White, 1994); the Extraversion, Neuroticism and Psychoticism scales of the short form of the Eysenck Personality Questionnaire-Revised (EPQ-R) (Eysenck et al., 1985); a short form of the Positive and Negative Aect Schedule (PANAS) (Kercher, 1992) in which participants had to rate their feelings ``in general''; the anxiety and depression scales of Goldberg et al. (1988); the anxiety and depression scales of the Delusions-SymptomsStates Inventory (DSSI) (Bedford et al., 1976) and the Alcohol Use Disorders Identi®cation Test (AUDIT) (Saunders et al., 1993). 2.3. Survey procedure Persons selected at random from the Electoral Roll were sent a letter informing them about the survey and saying that an interviewer would contact them soon to see if they wanted to participate. If a person agreed to participate, the interviewer visited them at some convenient location, usually the participant's home. The interviewer took a cheek swab and then gave the participant the questionnaire to self-complete. To encourage frank responses, the questionnaire had an ID number, but not the participant's name. After completing the questionnaire, the participant sealed it in an envelope and gave it to the interviewer who never saw the answers given.
3. Results 3.1. Factor analysis of BIS/BAS items A principal component analysis was carried out on the data from the 2684 participants who completed all the BIS/BAS items. Following the procedure of Carver and White (1994), an oblique rotation was carried out on the four factors with eigenvalues greater than 1. These four factors accounted for 51% of the variance. The resulting factor pattern matrix is shown in Table 1. The ®rst factor corresponded to fun seeking, the second to behavioural inhibition, the third to drive and the fourth to reward responsiveness. The items loaded on the expected factors with some minor exceptions. One of the reward responsiveness items had low loadings on all four factors, while one of the fun seeking items loaded on both the fun seeking and reward responsiveness factors. The reward responsiveness, drive and fun seeking factors were intercorrelated 0.20±0.29 as would be expected if they all re¯ected aspects of behavioural activation. While a four-factor solution was consistent with the factor analysis of Carver and White (1994), a scree plot suggested that a two-factor solution would be more appropriate. The eigenvalues of the ®rst six principal components were 4.59. 3.19, 1.27, 1.14, 0.97 and 0.91,
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
53
Table 1 Factor loadings for BIS/BAS items 2-Factor solution
4-Factor solution Item BIS If I think something unpleasant is going to happen I usually get pretty ``worked up'' I worry about making mistakes Criticism or scolding hurts me quite a bit I feel pretty worried or upset when I think or know somebody is angry at me Even if something bad is about to happen to me, I rarely experience fear or nervousness I feel worried when I think I have done poorly at something I have very few fears compared to my friends BAS reward responsiveness When I get something I want, I feel excited and energized When I'm doing well at something, I love to keep at it When good things happen to me, it aects me strongly It would excite me to win a contest When I see an opportunity for something I like, I get excited right away BAS drive When I want something, I usually go all-out to get it I go out of my way to get things I want If I see a chance to get something I want, I move on it right away When I go after something I use a ``no holds barred'' approach BAS fun seeking I will often do things for no other reason than that they might be fun I crave excitement and new sensations I'm always willing to try something new if I think it will be fun I often act on the spur of the moment
1
2
0.08
0.75 ÿ 0.08 ÿ 0.07
ÿ 0.15 ÿ 0.09 0.01
3
0.70 ÿ 0.06 0.72 ÿ 0.05 0.75 0.03
4
1
2
0.08
0.73
0.13 0.03 0.01 ÿ 0.02 0.04 0.01
0.69 0.71 0.75
ÿ 0.29 ÿ0.43 ÿ 0.20
0.29
0.13 ÿ 0.43
ÿ 0.13 0.65 ÿ 0.06 ÿ 0.08 ÿ0.42 ÿ 0.21
0.21 0.36
0.08 0.65 0.32 ÿ 0.40
0.26 ÿ 0.12 0.13 0.32 0.34
0.50 0.69 0.47 0.41 0.23
0.55 0.36 0.34 0.33 0.62
0.20 0.12 0.42 0.23 0.28
ÿ 0.85 0.02 ÿ 0.82 ÿ 0.05 ÿ 0.69 0.06 ÿ 0.67 ÿ 0.08
0.70 0.63 0.70 0.71
ÿ 0.03 ÿ 0.12 ÿ 0.06 ÿ 0.12
0.13 ÿ 0.10 0.04 ÿ 0.06 0.35 0.01 0.14 0.14 0.28 ÿ 0.31
ÿ 0.05 0.09 ÿ 0.04 0.01 0.13 0.02 0.29 ÿ0.03
0.70 ÿ0.20 ÿ 0.04 0.09 0.63 0.04 ÿ 0.29 0.01 0.46 ÿ0.23 ÿ 0.02 0.42 0.59 0.01 ÿ 0.24 ÿ 0.07
0.56 ÿ 0.14 0.67 0.04 0.57 ÿ 0.15 0.56 ÿ 0.00
Loadings of r0.4 shown in italics.
respectively. Table 1 also shows the results of the two-factor solution. The two factors accounted for 39% of the variance. The ®rst factor corresponded to behavioural activation and the second to behavioural inhibition. All items loaded as expected except that three of the reward responsiveness items had low loadings on the behavioural activation factor. The two factors were correlated only 0.07. 3.2. Reliability Cronbach's alpha was 0.76 for BIS, 0.83 for BAS, 0.65 for reward responsiveness, 0.80 for drive and 0.70 for fun seeking.
54
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
3.3. Scale inter-correlations If participants had missing data on one item of a scale their score was pro-rated by taking the mean of the items responded to and multiplying by the number of items in the full scale. However, if they had more missing data the scale was regarded as missing. The BIS scale correlated 0.06 with BAS, ÿ0.05 with drive, ÿ0.03 with fun seeking and 0.27 with reward responsiveness. Drive correlated 0.52 with fun seeking and 0.42 with reward responsiveness, while fun seeking correlated 0.45 with reward responsiveness. 3.4. Age and gender norms Table 2 shows the means and standard deviations for the scales by age group and gender. Multiple regression analyses were carried out to see if there were age and gender dierences, with the P < 0.01 level used for statistical signi®cance. Age-by-gender interactions were tested by entering the product of age and gender after the main eects had been entered (Cohen and Cohen, 1975). It can be clearly seen in Table 2 that older groups scored lower on all the BIS/ BAS scales. Females scored higher on the BIS and reward responsiveness, while males were higher on drive. There was no gender dierence on the BAS total. There were signi®cant ageby-gender interaction eects for the BIS and drive scales, due to bigger gender dierences in younger age groups for BIS and bigger gender dierences in older age groups for drive. It can also be seen in Table 2 that the standard deviations of all the BAS scales tend to increase in older age groups and these dierences were statistically signi®cant. Table 2 Means (and SD's) for BIS/BAS scales by age and gender Group
N*
BIS
BAS
Males 18±29 30±39 40±49 50±59 60±69 70±79
277 276 336 238 112 54
19.3 20.0 20.0 19.9 19.9 18.8
(3.5) (3.4) (3.7) (3.4) (2.8) (3.3)
40.0 38.1 37.1 36.5 35.2 34.3
(4.8) (5.2) (5.5) (5.6) (6.6) (6.9)
16.8 16.5 16.1 16.3 15.5 15.0
(1.9) (2.0) (2.1) (2.0) (2.4) (2.7)
10.9 10.3 10.3 10.0 9.8 9.6
(2.3) (2.4) (2.4) (2.5) (2.8) (2.7)
12.3 11.3 10.6 10.2 9.9 9.7
(2.2) (2.3) (2.2) (2.4) (2.8) (2.6)
Females 18±29 30±39 40±49 50±59 60±69 70±79
312 341 393 208 99 71
22.0 21.7 21.4 20.9 20.1 19.8
(3.4) (3.5) (3.4) (3.4) (3.4) (3.6)
40.2 38.1 37.4 36.1 35.5 33.6
(5.1) (5.4) (5.6) (5.8) (6.1) (6.7)
17.6 17.0 16.8 16.5 16.4 16.0
(1.9) (1.9) (2.0) (2.2) (2.5) (2.6)
10.7 10.1 9.8 9.3 8.9 8.2
(2.4) (2.6) (2.7) (2.7) (2.8) (3.1)
11.9 11.1 10.8 10.3 10.2 9.5
(2.2) (2.2) (2.3) (2.5) (2.4) (2.5)
*Numbers vary slightly from scale to scale because of missing data.
Reward
Drive
Fun
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
55
Table 3 Correlations of BIS/BAS scales with personality and psychiatric symptom measures Scale
BIS
BAS
Reward
Drive
Fun
EPQ-R Extraversion Neuroticism Psychoticism
ÿ 0.19 0.64 ÿ 0.31
0.44 0.09 0.18
0.28 0.21 ÿ 0.02
0.32 0.01 0.20
0.46 0.02 0.23
PANAS Positive Aect Negative Aect
ÿ 0.12 0.48
0.46 0.04
0.33 0.13
0.39 ÿ 0.01
0.37 ÿ 0.00
DSSI Anxiety Depression
0.44 0.35
0.08 0.02
0.14 0.06
0.03 ÿ 0.01
0.03 0.01
Goldberg scales Anxiety Depression
0.42 0.36
0.10 0.04
0.16 0.10
0.04 ÿ 0.02
0.06 0.03
AUDIT
0.02
0.12
0.02
0.11
0.16
P < 0.01 for all r'sr0.05 and P < 0.001 for all r'sr0.07.
3.5. Correlations with other scales Table 3 shows the correlations with other personality and psychiatric symptom scales. It can be seen that BIS was related most strongly to Neuroticism, Negative Aect and anxiety and depressive symptoms, but there was also a small negative correlation with psychoticism. BAS was related most strongly to Extraversion and Positive Aect, but there were also small correlations with psychoticism and the AUDIT. All the BAS subscales were related to extraversion and positive aect, but the reward responsiveness scale also had a small correlation with neuroticism, while the drive and fun seeking scales had small correlations with psychoticism and the AUDIT. To get a clearer picture of how the BIS/BAS relates to the other personality scales, a principal component analysis with oblique rotation was carried out on the BIS/BAS, EPQ-R and PANAS subscales. A scree plot suggested either a 2-factor or a 3-factor solution. The 3factor solution was found to be more satisfactory because the Psychoticism scale did not load highly on either factor in the 2-factor solution. The 3-factor solution accounted for 68% of the variance. Table 4 shows the results. The BAS subscales, extraversion and positive aect loaded highly on the ®rst factor which could be labelled ``extraversion/positive aectivity''. The BIS scale, neuroticism and negative aect loaded highly on the second factor which could be labelled ``neuroticism/negative aectivity''. The third factor had a high psychoticism loading and a smaller negative loading for BIS. Factors 1 and 2 were correlated ÿ0.02, factors 1 and 3 0.10 and factors 2 and 3 ÿ0.12.
56
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58 Table 4 Loadings from factor analysis of personality scales Scale
Factor 1
Factor 2
Factor 3
BIS BAS reward BAS drive BAS fun Extraversion Neuroticism Psychoticism Positive aect Negative aect
0.02 0.72 0.71 0.76 0.69 ÿ 0.03 0.06 0.71 ÿ 0.03
0.76 0.31 0.07 0.08 ÿ 0.25 0.91 ÿ 0.00 ÿ 0.16 0.83
ÿ0.34 ÿ0.20 0.20 0.21 ÿ0.06 0.04 0.94 ÿ0.06 0.13
4. Discussion The factor analyses largely supported the factor structure reported by Carver and White (1994). The four-factor solution corresponded to the four BIS/BAS scales, while the two-factor solution gave behavioural inhibition and behavioural activation factors. However, some of the reward responsiveness items did not load unequivocally on the expected factors. In particular, the ®fth item (``When I see an opportunity for something I like, I get excited right away'') did not load speci®cally on the reward responsiveness factor, although it was adequate as a general BAS item. Other ®ndings pointed to limitations of the Reward Responsiveness scale. It had the lowest reliability of all the BIS/BAS scales and it correlated with BIS and neuroticism as well as with the other BAS subscales. The pattern of correlations of the BIS/BAS with other personality scales suggests that the BIS scale is largely a measure of a neuroticism/negative aectivity super-factor, while the BAS scale is a measure of an extraversion/positive aectivity super-factor. However, Carver and White (1994) claimed that the BIS scale is not so closely aligned with other anxiety measures because the items are worded to re¯ect propensity to anxiety given a particular situation, rather than how often anxiety is experienced. Indeed, the BIS scale had lower correlations with measures of current anxiety and depression (r's of 0.35±0.44) than either the Neuroticism or Negative Aect scales (r's of 0.56±0.62 and 0.47±0.59, respectively), supporting the interpretation that it is less a measure of experienced anxiety. There were some interesting age and gender dierences on the BIS/BAS scales. Both BIS and BAS scores were lower in older age groups, consistent with ®ndings for Neuroticism and Extraversion (Eysenck et al., 1985). Whether these age group dierences re¯ect ageing or cohort eects is impossible to determine from a cross-sectional study such as this. However, if the dierences do re¯ect ageing, they imply that the emotional systems measured by the BIS/ BAS become less responsive with age. Another interesting age group dierence was evident in the standard deviations of the BAS scale and its subscales, which were signi®cantly larger in older age groups. It is clear from an inspection of the means and the possible range of scale scores that these dierences in dispersion are not simply due to ceiling or ¯oor eects. They remain a puzzle, but are
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
57
consistent with other evidence that older age groups tend to show increased variability in a range of biological and psychological domains, including personality characteristics (Nelson and Dannefer, 1992). Gender dierences were more complex. Females had higher mean scores on the BIS scale which is consistent with the ®ndings with the neuroticism scale. Furthermore, the gender dierence on the BIS scale was greater in younger age groups than in older ones, which is also consistent with ®ndings on neuroticism (Jorm, 1987). While there was no dierence on the overall BAS score, females scored higher on the reward responsiveness subscale and males higher on the drive subscale. The gender dierence on the reward responsiveness scale may be due to the fact that it has some neuroticism/negative aectivity component (on which females tend to score higher). The higher male scores on drive were age group speci®c, being largely con®ned to the older age groups. Again, whether this is an ageing or a cohort eect is impossible to determine. Gray (1987) has speculated on the relationships of his emotional systems to Eysenck's personality traits. Assuming that Carver and White's (1994) scales are adequate measures of behavioural inhibition and activation, the pattern of correlations found here does not entirely support Gray's views. Behavioural inhibition is strongly associated with neuroticism, but only weakly with low extraversion and low psychoticism. Behavioural activation is strongly associated with extraversion, but only weakly with psychoticism and neuroticism. In short, behavioural inhibition is primarily associated with neuroticism and behavioural activation with extraversion. Ball and Zuckerman (1990) had similar ®ndings with their GRAPES questionnaire. However, the location in personality space of behavioural inhibition and behavioural activation can only be ultimately settled by linking indicators of these brain systems with personality measures. This study is the ®rst to report the psychometric properties of the BIS/BAS scales in a sample of the general population. The conclusion is that they are a valid and practical measure of two personality traits relevant for molecular genetics research.
Acknowledgements Thanks to Karen Maxwell and Michelle Powderly for assistance with managing the survey and to Simon Easteal for discussions.
References Ball, S. A., & Zuckerman, M. (1990). Sensation seeking, Eysenck's personality dimensions and reinforcement sensitivity in concept formation. Personality and Individual Dierences, 11, 343±353. Bedford, A., Foulds, G. A., & Sheeld, B. F. (1976). A new personal disturbance scale (DSSI/sAD). British Journal of Social and Clinical Psychology, 15, 387±394. Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and aective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67, 319±333. Cloninger, C. R., Przybeck, T. R., & Svrakic, D. M. (1991). The Tridimensional Personality Questionnaire: US normative data. Psychological Reports, 69, 1047±1057.
58
A.F. Jorm et al. / Personality and Individual Dierences 26 (1999) 49±58
Cohen, J. & Cohen, P. (1975). Applied multiple regression/correlation analysis for the behavioral sciences. Hillsdale, New Jersey: Lawrence Erlbaum. Eysenck, S. B. G., Eysenck, H. J., & Barrett, P. (1985). A revised version of the psychoticism scale. Personality and Individual Dierences, 6, 21±29. Goldberg, D., Bridges, K., Duncan-Jones, P., & Grayson, D. (1988). Detecting anxiety and depression in general medical settings. British Medical Journal, 297, 897±899. Gray, J. A. (1987). The neuropsychology of emotion and personality. In S. M. Stahl, S. D. Iversen & E. C. Goodman (Eds.), Cognitive neurochemistry (pp. 171±190). Oxford: Oxford University Press. Harmon-Jones, E., & Allen, J. J. B. (1997). Behavioral activation sensitivity and resting frontal EEG asymmetry: Covariation of putative indicators related to risk for mood disorders. Journal of Abnormal Psychology, 106, 159±163. Jorm, A. F. (1987). Sex dierences in neuroticism: A quantitative synthesis of published research. Australian and New Zealand Journal of Psychiatry, 21, 501±506. Kercher, K. (1992). Assessing subjective well-being in the old-old. Research on Aging, 14, 131±168. MacAndrew, C., & Steele, T. (1991). Gray's behavioral inhibition system: A psychometric evaluation. Personality and Individual Dierences, 12, 157±171. Nelson, E. A. , & Dannefer, D. (1992). Aged heterogeneity: Fact or ®ction? The fate of diversity in gerontological research. Gerontologist, 32, 17±23. Pickering, A. D. (1997). The conceptual nervous system and personality: From Pavlov to neural networks. European Psychologist, 2, 139±163. Pickering, A. D., Corr, P. J., Powell, J. H., Kumari, V., Thornton, J. C. & Gray, J. A. (1997). Individual dierences in reactions to reinforcing stimuli are neither black nor white: To what extent are they Gray? In H. Nyborg (Ed.), The scienti®c study of human nature: Tribute to Hans J. Eysenck at eighty (pp. 36±67). London: Elsevier Science. Saunders, J. B., Aasland, O. G., Babor, T. F., De La Fuente, J. R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identi®cation Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption. II. Addiction, 88, 791±804. Torrubia, R., & TobenÄa, A. (1984). A scale for the assessment of ``susceptibility to punishment'' as a measure of anxiety: Preliminary ®ndings. Personality and Individual Dierences, 5, 371±375. Wilson, G. D., Barrett, P. T., & Gray, J. A. (1989). Human reactions to reward and punishment: A questionnaire examination of Gray's personality theory. British Journal of Psychology, 80, 509±515. Wilson, G. D., Gray, J. A., & Barrett, P. T. (1990). A factor analysis of the Gray±Wilson Personality Questionnaire. Personality and Individual Dierences, 11, 1037±1045.