Вы находитесь на странице: 1из 14

Perceptualand Motor Skills, 2007, 104, 1169-1182.

O Perceptual and Motor Skills 2007

AN EXAMINATION O F THE PSYCHOMETRIC PROPERTIES O F


T H E CHINESE VERSION O F THE BARRATT IMPULSIVENESS
SCALE, 1~ T H
VERSION I N A SAMPLE O F
CHINESE ADOLESCENTS ' ' '

SHUQIAO YAO, HUIQIN YANG, AND XIONGZHAO ZHU


Medical Psychological Research Center
Central South Urzzversity

RANDY P . AUERBACH, J O H N R. Z. ABELA.


RYAN W. PULLEYBLANK, AND XI T O N G

Summary.-The current study examined the psychometric properties of the Chi-


nese translation of the Barratt Impulsiveness Scale 11th version in a sample of adoles-
cents from Hunan province, mainland China. During an initial assessment, 396 sec-
ondary school students (Grades 10-12) completed the scale and self-report measures
assessing problem behaviors and alcohol use. The scale was re-administered 1 mo. la-
ter. Analysis gave Cronbach alpha of .80 and test-retest reliability of .70. Confirmatory
factor analysis indicated a model containing six first-order factors and two second-or-
der factors best fit the data. Girls reported higher Total scores than boys as well as
higher scores on the motor impulsiveness, self-control, and cognitive instability sub-
scales. Scores were associated in the predicted direction with a wide variety of self-re-
ported problem behaviors including alcohol use, gambling, and academic misconduct.
Current findings indicate that the translated scale is a promising tool with some fur-
ther development for assessing impulsiveness with Chinese adolescents.

Impulsiveness is defined as the tendency to react spontaneously in re-


sponse to internal and external stimulation without regard for subsequent
consequences (Moeller, Barratt, Dougherty, Schmitz, & Swann, 2001). Re-
search has shown that high impulsiveness is associated with a wide range of
psychiatric disorders including attention-deficit/hyperactivity disorder, con-
duct disorder, antisocial personality disorder, and borderline personality
disorder (e.g., Barratt & Patton, 1983; Barratt, 1994; McCarty, Weisz, Wani-
tromanee, Eastman, Suwanlert, Chaiyasit, & Band, 1999; Fossati, Di Ceglie,
Acquarini, & Barratt, 2001). High impulsiveness is also a strong predictor of
problem behaviors such as alcohol and drug use, gambling, unsafe sex, and
academic misconduct (e.g., Pedersen & Skrondal, 1999; Hulse, 2001; Daniel-

'Address correspondence to Shuqiao Yao, Medical Psychological Research Center, 2nd Xiang-
ya Hospital, Central South University, Changsha, Hunan 410011, P.R. China.
De artment of Psychology, McGill University, Stewart Biological Sciences Building, 1205 Dr.
Penkeld Avenue, Montrcal, Quebec, Canada H3A IB1.
This research was supported by a Canadian Psychiatric Research Foundation (CPRP) Award
and a McGill University Sabbatic Leave Research Grant awarded to John R. Z. Abela.
S. YAO, ET AL.

son, Overholver, & Butt, 2003; Spinella & Miley, 2003). Given the influen-
tial role impulsiveness appears to play in the etiology or maintenance of
various forms of psychopathology and problem behaviors, researchers have
devoted substantial resources to developing a reliable and valid means of
assessing impulsiveness.
While a number of researchers have focused on purely biological mod-
els of impulsivity (e.g., Linnoila, Virkkunen, Scheinin, Nuutila, R m o n , &
Goodwin, 1983; Apter, Van Praag, Plutchik, Seavy, Korn, & Brown, 1990),
Barratt and Stanford (1995) attempted to differentiate the biological, social,
cognitive, and behavioral components of impulsivity as well as to understand
the relationship among these. One of the primary goals of the Barratt Impul-
siveness Scale-1 (Barratt, 1959) was to assess impulsiveness and anxious
symptomology as orthogonal constructs. In line with the Hull-Spence learn-
ing theory (Hull, 1943; Spence, 1956), Taylor (1958) proposed that anxiety
and impulsivity operate within the same motivational system, with anxiety
being related to habit strength and impulsiveness to behavioral oscillation.
Taylor and Spence posited that habit strength and behavioral oscillation are
connected to different neural pathways (cf. also, Taylor & Spence, 1952;
Taylor, 1958; Patton, Stanford, & Barratt, 1995). To define the nature and
structure of impulsiveness more clearly, Barratt (1959) sought to differentiate
these biological and behavioral constructs. Over the years, the construct
validity of the Barratt Impulsiveness Scale has been revised and improved,
with the aims of version 11 being increased efficacy in differentiating impul-
siveness from trait anxiety, personality dimensions (e.g., extraversion and
psychoticism) and "action-oriented dimensions" such as sensation-seeking
(Barratt & Patton, 1983; Patton, Stanford, & Barratt, 1995).
In an initial study examining the reliability and validity of the Barratt
Impulsiveness Scale-11, Patton and colleagues (1995) administered the scale
to three different American samples, i.e., university undergraduates, psychi-
atric inpatients, and male prison inmates. Analysis indicated that the 30-item
measure possessed strong internal consistency and contained six first-order
factors of attention, motor impulsiveness, self-control, cognitive complexity,
perseverance, and cognitive instability, and three second-order factors of at-
tentional impulsiveness, motor impulsiveness, and nonplanning impulsive-
ness. As both the six first-order factors and the three second-order factors
significantly intercorrelated with each other (.I5 to .42, p < .0001; .46 to .53,
p < ,0001, respectively), the authors argued that the 30 items of the scale as-
sess the general personality trait of impulsivity. At the same time, as the
magnitude of intercorrelation among the subfactors was only moderate, the
authors proposed that each subfactor assesses a relatively separate compo-
nent of impulsivity.
The Barratt Impulsiveness Scale-11 has not only exhibited strong psy-
BARRATT IMPULSIVENESS SCALE-11: PSYCHOMETRICS 1171

chometric properties in American samples but has also been a reliable and
valid means of assessing impulsiveness in a number of international settings
(Patton, et al., 1995; Fossati, et al., 2001; Someya, Sakado, Seki, Kojima,
Reist, Tang, & Takahashi, 2001). For example, the reliability and validity of
translated versions have been examined in adults in both Japan (Someya, et
al., 2001) and Italy (Fossati, et al., 2001). Both the Japanese (Cronbach al-
pha = .79; ICC= .71) and Italian (Cronbach alpha = .79; ICC= 3 9 ) versions
possessed moderately strong internal consistency and moderately strong to
strong test-retest reliability. With regard to validity, both the Japanese and
Italian versions had factor structures similar to those for the American sam-
ples, i.e., six first-order and three second-order factors. More specifically,
confirmatory factor analyses of the Japanese sample indicate a three-factor
structure gave the best fit to the data, with each factor corresponding to one
of the three American second-order factors. Exploratory factor analysis of
the Italian sample also supported a six first-order factor and three second-
order factor structure. However, some of the subfactor item loadings differ-
ed from those for the American sample. More specifically, whereas Factor I
in the Italian sample was loaded by attention and motor impulsiveness,
Factor I in the American sample was loaded by attention and cognitive in-
stability. In addition, whereas Factor I1 in the Italian sample was loaded by
perseverance and delay in obtaining gatification, Factor I1 in the American
sample was loaded by perseverance and motor impulsiveness. In both the
American and Italian samples, Factor I11 was loaded by cognitive complexity
and self-control.
As there is currently no measure to assess impulsiveness of adolescents
in mainland China, the objectives of the current study were four-fold: (1) to
provide descriptive data for the Chinese version of the Barratt Impulsiveness
Scale-11 in a sample of secondary school students (Grades 10-12) from H u -
nan province, China; (2) to examine internal consistency and 1-mo. test-re-
test reliability; (3) to conduct confirmatory factor analyses with which were
assessed whether the six first-order and three second-order factor structure of
past research (Patton, et al., 1995; Fossati, et al., 2001; Someya, et al., 2001)
provided the best fit of the data; (4) also examined were associations of
scores on the Barratt Impulsiveness Scale-11-C with measures of problem
behaviors (e.g., alcohol use, drug use, and gambling), expecting high impul-
siveness to be positively associated with engagement in a greater number of
problem behaviors.
METHOD
Participants
Participants were recruited from an urban high school in Yue Yang,
Hunan (China). The final sample consisted of 396 adolescents (200 girls and
S. YAO. ET AL.

196 boys). Participants' ages ranged from 14 to 19 years (M= 16.2, SD= .95).
The sample was 97.1% Han, 1.7 ethnic minority, and 1.2% of participants
did not report their ethnicity. With regard to the family structure of the
household, participants' reported the following: 70.1% two parent families,
19.7% single parent families, 6.6% grandparent and two parent families,
and 3.6% did not respond.
Procedure
Prior to the initial assessment, letters of informed consent were sent to
parents of students detailing the project aims which included developing a
Chinese version of the Barratt Impulsiveness Scale-11 and examining its psy-
chometric properties. Students were only permitted to participate if they
both received parental consent and gave personal consent. Consent rates
were higher than 95% in all classes. During the initial assessment, students
completed a demographic information form and Chinese versions of the fol-
lowing measures: (1) Barratt Impulsiveness Scale (Patton, et al., 1995), (2)
the Risky Behavior Questionnaire-Adolescents (Auerbach & Abela, 2006),
(3) the Alcohol Use Disorders Identification Test (Saunders, Aasland, Babor,
De La Fuente, & Grant, 1993), (4) the Rutgers Alcohol Problem Index
(White & Labouvie, 1989), and (5) the Quantity-Frequency Measure (Skitch,
2005). One month later, the Barratt Impulsiveness Scale-11 was re-adminis-
tered.
Measures
The Chinese versions of all measures were developed using the back-
translation method. First, the original versions were translated into Chinese
by one bilingual translator from the Psychology Department at Central South
University (Changsha, Hunan). Next, the Chinese versions were back-trans-
lated into English by another bilingual translator from the Psychology De-
partment at McGill University. Finally, the original versions were compared
with the back-translations. If discrepancies arose in the back-translations,
translators worked cooperatively to make corrections to the Chinese ver-
sions.
Chinese Version of the Barratt Impulsiveness Scale, 1 l t h verszon (Patton,
et al., 1995).-The Barratt Impulsiveness Scale-11-C is a 30-item self-report
questionnaire designed to assess characteristics of impulsiveness. The mea-
sure contains six subscales which correspond to the six first-order factors.
The first-order factor labels and definitions include (1) attention, "focusing
on the task at hand"; (2) motor impulsiveness, "acting on the spur of the
moment"; (3) self-control, "planning and thinking carefullyn; (4) cognitive
complexity, "enjoy challenging mental tasks"; (5) perseverance, "a consistent
life styleN; and (6) cognitive instability, "thought insertions and racing
thoughts." In addition, the six first-order factors converge into three sec-
BARRATT IMPULSIVENESS SCALE-11: PSYCHOMETRICS 1173

ond-order factors: (1) attentional impulsiveness (attention and cognitive in-


stability), (2) motor impulsiveness (motor impulsiveness and perseverance),
and (3) non~lannedimpulsiveness (self-control and cognitive complexity).
Items on the scale range from 1 (rarely or never) to 4 (almost always or al-
ways), and the self-control, cognitive complexity, and perseverance subscales
are reverse scored. Higher scores reflect greater impulsiveness. Past research
has indicated strong validity and reliability across samples from at least three
countries (e.g., Italy, United States, and Japan; Patton, et al., 1995; Fossati,
et al., 2001; Someya, et al., 2001).
Risky Behavior Questionnaire for Adolescents (Auerbach & Abela,
2006).-This questionnaire is a 20-item self-report measure created to assess
frequency of engagement in risky behaviors. Respondents are asked to report
their engagement in such behaviors over the past month using identifications
of never, 1 time per month, 2 to 4 times per month, 2 to 3 times per week,
4 or more times per week. Item scores range from O to 4 with higher scores
reflecting greater engagement in risky behaviors. Examples of questions in-
clude "Have you bullied or threatened a peer(s)," "Have you destroyed
property (other than your own)," and "Have you used illegal drugs?" Sub-
scales assess engagement in six subgroups of behaviors: (1) unsafe sexual
practices; (2) aggressive and/or violent behaviors; (3) rule-breaking; (4) dan-
gerous, destructive, or illegal behaviors; ( 5 ) self-injurious behaviors; and (6)
alcohol and/or drug use.
The Alcohol Use Disorders Identifcation Test (Saunders, et al., 1993).-
This test is a 10-item self-report measure developed by a research group af-
filiated with the World Health Organization which assesses the frequency
and effects of alcohol use. Scores range from O to 4 with higher scores indi-
cating elevated consumption of alcohol. Examples of questions include
"How often do you have a drink containing alcohol" and "How often dur-
ing the last year have you had a feeling of guilt or remorse after drinking."
Past research indicates that this test is reliable and valid across several cul-
tural samples and item pools (Babor, Higgins-Biddle, Saunders, & Monteiro,
2001).
The Rutgers Alcohol Problem Index (White & Labouvie, 1989).-This
test is a 23-item measure which assesses problems youth attribute to their
substance use (e.g., missing school, changes in personality, or getting into
fights). Scores are based on O (never) and 4 (more than 10 times per month),
so higher scores reflect a greater number of problems as a result of alcohol
use. Participants are asked how many times during the last four weeks a
given event occurred because of one's use of alcohol. Examples of questions
include "Not able to do your homework or study for a test" and "Suddenly
found yourself in a place that you could not remember getting to?" Past re-
search suggests that this index is a strong marker of problem drinking and
S. YAO, ET AL

correlates moderately well with other problem-drinking indicators (White &


Labouvie, 1989).
Quantz'ty-Frequency Measure (Shtch, 2005).--The Q-F Measure has four
items which assess an individual's amount and frequency of drinking over
the past month. Scores are based on O (never) and 4 (8 or more times over
the past month), for which higher scores reflect greater amounts of alcohol
use or more frequent alcohol use. As individuals may consume any number
of dfferent alcoholic beverages (e.g., beer, wine, coolers, and hard liquor),
respondents are asked to specify their answers in terms of standard drinks,
e.g., 1.5 ounces of hard liquor (one shot) or 5 ounces of wine (one glass).
Examples of questions include "During the past month, how often have you
had a drink containing alcohol" and "During the past month, how often
have you had five or more standard drinks on one occasion?"
Statistical Analysis
Analyses were conducted using SPSS 13.0 and AMOS 7.0 software. To
examine sex differences both independent-samples t tests and a one-way
multivariate analysis of variance were performed. Cronbach coefficients al-
pha were calculated to estimate the internal consistency. Pearson correlations
were applied to examine test-retest reliability.
Maximum likelihood confirmatory factor analysis was used to specify
the factor structure which gave the best fit for the data. In general, confir-
matory factor analysis is a more stringent procedure for testing factor struc-
ture than exploratory factor analysis because it provides tests for significance
as well as multiple goodness-of-fit indices. To evaluate model fit, eight indi-
ces were analyzed (parentheses indicate goodness-of-fit standards): ( I ) the
goodness of fit index (GFI>.90); (2) the adjusted goodness of fit index
(AGFI > 20); (3) the Tucker-Lewis, nonnormed fit index (NNFI > .90); (4)
the root means square error of approximation (RMSEA < .lo); (5) the com-
parative fit index (CFI> .90); ( 6 ) the x2 divided by its degrees of freedom
(x2/df values below 2.0 usually indicate a good fit); (7) the Akaike Informa-
tion Criterion (preferred models have lower AIC values); and (8) Schwartz's
Bayesian Information Criterion (preferred models have lower BIC values)
(Akaike, 1974; Schwarz, 1978; Bollen, 1989; Bentler, 1990; Browne & Cud-
eck, 1993).

Descriptive Data
Means and standard deviations for all measures are presented in Table
1 and those by sex in Table 2. Total scores on the Barratt Impulsiveness
Scale-11-C ranged from 44 to 97. An independent-samples t test with Barratt
Impulsiveness Scale-11-C total scores indicated the girls scored significantly
BARRATT IMPULSIVENESS SCALE-11: PSYCHOMETRICS 1175

TABLE 1
DESCRIPTIVE
STATISTICS
AND INTERNAL
CONSISTENCY
OF SELF-REPORT
MEASURES
---
Measure M SD Cronbach a
Barratt Impulsiveness Scale-1l-C 68.55 9.27 .80
&sky Behavior Questionnaire-Adolescent 9.07 6.42 .81
Alcohol Use Disorders Identification Test 1.66 2.56 .65
Rurgers Alcohol Problem Index 5.58 7.93 .92
Quality-Frequency Measure 1.07 2.07 .62
-

higher than the boys (t=2.07, p < .05). When multivariate analysis of vari-
ance was utilized to analyze sex differences for scores on the Barratt Impul-
siveness Scale-11-C subscales, significant sex differences were obtained for
the following subscales: Motor Impulsiveness ( F ,,,,= 4.86, p < ,051, Self-con-
trol ( F ,,,, =4.76, p < .05), and Cognitive Instability (F,,3,, =4.35, p < .05). Sex
differences on the remaining subscales were not statistically significant. Fur-
ther, there were no significant interactions for sex by subscale.

TABLE 2
CHINESE
VERSION
OF BARRATT
IMPULSIVENESS
SCALE-1
1-C: MEANSAND
STANDARD
DEVIATIONS
FORBOYSAND GIRLSAND TOTALGROUP
-- -
-

Scale Total Subjects Girls Boys


-
(N=396)
---
( n= 200) - -
(n= 196) -
M SD M SD M SD
Total Barratt Impulsiveness Score 68.55 9.27 69.50 9.09 67.58 9.39
Attention 10.85 2.47 10.82 2.26 10.88 2.67
Motor Impulsiveness 15.44 3.60 15.83 3.59 15.04 3.58
Self-control 15.31 3.33 15.67 3.24 14.94 3.38
Cognitive Complexity 12.76 2.21 12.82 2.21 12.70 2.21
Perseverance 7.18 1.38 7.19 1.39 7.17 1.38
Cognitive Instability 7.02 1.60 7.19 1.54 - 6.85 1.64
--

Independent-sample t tests, conducted to examine sex differences on


measures assessing problem behaviors, indicated the boys scored significantly
higher than the girls on the &sky Behavior Questionnaire for Adolescents
(t=4.16, p<.001), The Alcohol Use Disorders Identification Test (t=2.55,
p < .05), and the Quantity-Frequency Measure it =3.83, p < .001), but no dif-
ferences on The Rutgers Alcohol Problem Index were statistically significant.
Item Analysis and Internal Consistency Reliability
The internal consistency for each measure is reported in Table I. Val-
ues of Cronbach alpha for all scales indicate strong internal consistency,
except for the Alcohol Use Disorders Identification Test and Quantity-Fre-
quency Measure which fell below the usual cutoff value of .70. Evaluation of
items which might underlie this discrepancy is needed.
S. YAO, ET AL.

Test-Retest Reliability
Test-retest reliability results of the Total score and Barratt Impulsive-
ness Scale-11-C subscales are presented in Table 3 . Pearson correlation for
the initial Total scores on the Barratt Impulsiveness Scale-11-C and the 1-
mo. follow-up scores was moderately strong (.70) as was that for Self-control
scores (.74). Additional evaluation is desirable.
TABLE 3
CHINESE
VERSION
OF BARRATT
IMPULSIVENESS
SCALE-1
1-C: TEST-RETEST
RELIABILITY
-.
Scale Y P
Total Barratt Impulsiveness Score .70 < ,001
Attention .48 < .001
Motor Impulsiveness .52 < ,001
Self-control .74 < ,001
Cognitive Complexity .53 < ,001
Perseverance .40 < ,001
Cognitive Instability .5 1 < ,001

Factor Structure
Although the majority of past research examining the factor structure of
the Barratt Impulsiveness Scale-11 has yielded a six first-order and three sec-
ond-order factor structure, some studies have obtained divergent structures
such as a six first-order and two second-order factor structure (Patton, et al.,
1995; Fossati, et al., 2001; Fossati, Barratt, Acquarini, & Di Ceglie, 2002).
Table 4 contains values for the eight indices of model fit when comparing a
TABLE 4
CONFIRMATORY
FACTOR
ANALYSIS: AND INCREMENTAL
GOODNESS-OF-FIT FIT STATISTICS
FORCHINESE OF BARRATT
VERSION IMPULSIVENESS
SCALE-1
1-C
-- -
Measure Rule Six Correlated Interpretation
First-order Factors
Three Two
Second-order Second-order
GFI > .90 .89 .91 Borderline: two second-order
factors model
AGFI > .80 .84 .86 Both adequate
NNFI > .90 .70 .73 Both inadequate
CFI > .90 .77 .78 Both inadequate
RMSEA < .10 .10 .09 Borderline: two second-order
factors model
x2/df <2.0 2.02 1.35 Two second-order factors model
AIC Lower value 3 14.28 299.61 Two second-order factors model
preferred
BIC Lower value 421.78 399.14 Two second-order factors model
preferred
Note.-GFI = Goodness of Fit Index; AGFI = Adjusted Goodness of Fit Index; NNFI = Non-
normed Fit Index; CFI=Comparative Fit Index; RMSEA=Root Mean Square Error of Ap-
proximation; AIC = Akaike Information Criterion; BIC = Schwartz Bayesian Information Crite-
rion.
BARRATT IMPULSIVENESS SCALE-11: PSYCHOMETRICS 1177

model with six first-order factors and three second-order factors and a mod-
el with six first-order factors and two second-order factors. The results in-
dicated that (a) the six first-order factors in both models positively intercor-
relate (r = .18 to .49), (b) the two second-order factors positively intercorre-
late (r = .39), and (c) the three second-order factors positively intercorrelate
(r = .31 to .45). The AGFI, NNFI, and CFI do not identify a preferred mod-
el. The GFI and RMSEA values indicate borderline preference for the two
second-order factors model. The ~ ' l d f AIC,
, and BIC values indicate prefer-
ence for the two second-order factors model. These results tentatively indi-
cate that the two second-order factors model is a better fit than the three
second-order factors model.
The factor loadings of the best-fit model are shown in Table 5. The sec-
ond-order Factor 1 included Attention (first-order Factor I ) , Motor Impul-
siveness (first-order Factor 2), and Cognitive Instability (first-order Factor
6). The second-order Factor 2 was composed of the Nonplanned Impulsive-
ness factor, defined by Self-control (first-order Factor 3 ) , Cognitive Complex-
ity (first-order Factor 4), and Perseverance (first-order Factor 5 ) .

TABLE 5

----
Factor I Factor I1
General Nonplanning
Imvulsiveness Imvulsiveness
Attention .52
Motor Impulsiveness .74
Self-control .78
Cognitive Complexity .72
Perseverance .65
Cognitive Instability .86 - --
Note.-Only factor loadings which were not fixed to zero are shown. The correlation between
second-order factors was significant ( r = 3 9 , p < ,001).

Relationship Between Impulsiveness and Problem Behaviors


The correlation coefficients between the Chinese translation of the Bar-
ratt Impulsiveness Scale, 11th version total score and the (a) Risky Behavior
Questionnaire for Adolescents, (b) Alcohol Use Disorders Identification
Test, (c) Rutgers Alcohol Problem Index, and (d) Quantity-Frequency Mea-
sure are resented in Table 6. The Chinese translation of the Barratt Impul-
siveness Scale, 11th version was significantly associated with both the Risky
Behavior Questionnaire for Adolescents (r = .3 1, p < .01) and the Rutgers Al-
cohol Problem Index (r = .18, p < .01) but not the Alcohol Use Disorders
Identification Test or Quantity-Frequency Measure. Analyses examining the
S. YAO, ET AL.

TABLE 6
CORRELATIONSOF BARRATT
IMPULSIVENESS
SCALE-1
1-C TOTALSCOREWITH SELF-REPORT
MEASURES
OF RISKYBEHAVIORQUESTIONNAIRE-ADOLESCENTS,
ALCOHOLUSF DISORDERSIDENTIFICATION
TEST,RUTGERS
ALCOHOLPROBLEMINDEX,AND QUANTITY-FREQUENCY
MEASURE
--
- - --pppp-
ppp

Measure Barratt Impulsiveness Scale-11 C


f'
- ?P
Risky Behavior Questionnaire-Adolescent .31t .30t
Alcohol Use Disorders Identification Test .08 .08
Rutgers Alcohol Problem Index ,181- .I6t
Quantity-Frequency Measure .07 .07
Age at which subjects started to use alcohol .13" 13 9:-
-

Note.-vp = partial correlation coefficient controlling for the effect of age. "p < .05. t p < .01.

correlation of each of the Risky Behavior Questionnaire for Adolescents sub-


scales and the Chinese translation of the Barratt Impulsiveness Scale, 11th
version total scores found that all of the &sky Behavior Questionnaire for
Adolescents subscales are significantly positively associated with the Chinese
translation of the Barratt Impulsiveness Scale, 11th version total scores sug-
gesting higher impulsiveness was associated with greater engagement in a
broad set of problem behavior types (unsafe sexual practices, r = .16; aggres-
sive and/or violent behavior, r = .23; rule-breaking, r = 3 3 ; dangerous, de-
structive, and/or illegal behaviors, r = 2 0 ; self-injurious behaviors, r = .16; al-
cohol and/or drug use, r = .17).Finally, analysis of covariance was used to
examine whether engagement in problem behaviors significantly differed
amongst the top and bottom ten percentile group of the Chinese translation
of the Barratt Impulsiveness Scale, 11th version total scores after controlling
for sex differences. Results indicated that there were statistically significant
differences in the following measures: Risky Behavior Questionnaire for Ad-
olescents, Alcohol Use Disorders Identification Test, Rutgers Alcohol Prob-
lem Index, Quantity-Frequency Measure (see Table 7).
TABLE 7
COMPARING
ENGAGEMENTIN PROBLEMBEHAVIORSAMONGST TOPAND BOTTOM
PFRCENTILE
10 GROUPSOF BARRATT
~MPULSIVENESSSCALE-]
1-C (TOTAL
SCORE)AFTERCONTROLLINGFORSEXDIFFERENCES
------- ~
--
Measure Top 10'%, Bottom 10% - MS F"
--- --
M SD M SD --

Risky Behavior Questionnaire-Adolescent 6.07 4.38 12.12 8.14 7310.07 358.66


Alcohol Use Disorders Identification Test 1.28 1.67 1.61 1.76 1080.74 169.28
Rutgers Alcohol Problem Index 3.26 3.96 8.10 11.55 12008.73 307.76
Ouantitv-Freuuencv Measure 0.43 1.06 1.15 2.01 713.90 324.80

D~scussro~
The present study examined the reliability and the validity of a Chinese
BARRATT IMPULSIVENESS SCALE-I 1: PSYCHOMETRICS 1179

translation of the Barratt Impulsiveness Scale, 11th version. To do this sex


differences, internal consistency, test-retest reliability, factorial validity, and
convergent validity were examined.
In the current study, a number of sex differences emerged. More specif-
ically, the girls reported significantly higher Total scores on the Barratt Im-
pulsiveness Scale-11-C than the boys. Also, the girls scored significantly high-
er than the boys on Motor Impulsiveness, Self-control, and Cognitive Insta-
bility subscales. Such findings contrast with those of Stanford, Greve, Bou-
dreaux, and Mathias (1996) who reported that American high school boys
exhibited significantly higher Total scores than girls. Such findings also con-
trast with lack of sex differences on either total or subscale scores as re-
ported by Patton, et al. (1995) for three independent adult samples. Thus,
the pattern of findings across studies suggests that differing social, develop-
mental, and cultural contexts may lead to different patterns in sex differ-
ences in impulsiveness. One possible explanation for higher impulsiveness in
the Chinese adolescent girls may involve social pressures on boys in China
(Gu & Roy, 1995). More specifically, there is still a cultural preference in
China for having boys as they are perceived to be better equipped than girls
to provide expected economic support for their immediate and extended
families. As boys are raised knowing this familial responsibility, it is possible
that they learn more quickly to inhibit their impulsiveness. At the same time,
as this is the first study to report such a difference, further research is re-
quired to assess stability across samples necessary for understanding cultural
factors that may be associated with such differences.
The current study provides preliminary support for the reliability of the
Barratt Impulsiveness Scale-11-C. The coefficients alpha for the total scores
and the subscale scores were weak to excellent. As the test-retest correla-
tions ranged from low moderate (i.e., perseverance) to moderately strong
(i.e., self-control), findings suggest that impulsiveness is a relatively stable
personality trait over 1 mo. Finally, the current pattern of findings is similar
to those obtained for the Italian, American, and Japanese samples (Patton,
et al., 1995; Fossati, et al., 2001; Someya, et al., 2001).
Results on the validity of the translated scale are comparable to those
obtained using previous translations into Italian and Japanese. Confirmatory
factor analyses yielded six first-order factors that converged into two second-
order factors. Although this factor structure differs slightly- .
from that ob-
tained in the American samples, such differences may reflect normal devel-
opmental change and requires study. More specifically, the factor structures
obtained in the United States and Italy were obtained with adult samples,
but the current Chinese work was with a sample of adolescents. Research by
Patton, et al. (1995) suggests that differentiation of impulsiveness may occur
S. YAO. ET A L

across ages. Thus the three second-order factors obtained in past research
with adult samples may emerge in young adulthood. Nevertheless, as factor
structure for the Chinese adolescents was similar to that for prior adult sam-
ples, the translated scale may be a valid means of assessing impulsiveness in
adolescents in China. This requires replication and extension across age
groups.
Finally, there may be statistically significant associations of the Barratt
Impulsiveness Scale-11-C with scores on multiple measures of problem be-
haviors including unsafe sex, alcohol use, drug use, and gambling, such that
individuals who reported higher impulsiveness may show greater engagement
in problem behaviors and larger number of problems with drinking alcohol.
Such findings are consistent with past studies of adolescents (e.g., Pedersen
& Skrondal, 1999; Danielson, et al., 2003); and suggest construct validity for
this translation. Further, as research on the Barratt Impulsiveness Scale-11
has consistently identified associations of impulsiveness with engagement in
problem behaviors (e.g., Patton, et al., 1995; Hulse, 2001; Spinella & Miley,
2003), this translated scale may be useful in the early detection of adolescents
who may be at risk in mainland China.
In sum, the current results suggest that the Barratt Impulsiveness Scale-
11-C, with some further psychometric tuning, is appropriate in assessing im-
pulsiveness Chinese adolescents and that the reliability and validity is good
across cultures. As the current study was based on a sample of middle- to
late-adolescents, additional research is needed on psychometric properties of
the scale in other age groups, i.e., children, early adolescents, adults, etc.,
and preferably with larger samples. Such research would allow a robust ex-
amination of whether the factor structure obtained for each of these age
groups is similar to prior studies with participants of similar ages. Research
is required to examine the convergent and divergent validity of the Barratt
Impulsiveness Scale-11-C with indigenously developed scales as well as from
examining the associations of scores on the Barratt Impulsiveness Scale-11-C
with measures assessing anxious symptoms and personality traits. Such re-
search should lead to better understanding of impulsiveness among Chinese
adolescents, and consequently, provide a basis for their training.
REFERENCES
AKAIKE,H. (1974) A new look at the statistical model identification. IEEE Transactions on
f l u t O m ~ t 2COntrf~l,
~ 19, 7 16-723.
APTER,A., VAN PRAAG,H.M., PLUTCEIIK, R., SEAVY,S., KORN,M., &BROWN,S. L. (1990) Inter-
relationships among anxiety, aggression, impulsivity, and mood: a ~erotonergicall~ liked
cluster? Psychzhtry Rrscarch, 32, 191-199.
AUERBACH, R. P., &ABELA,J. R. Z. (2006) Risky Behavior Questionnaire-Adolescents. Unpub-
lished test, McGill Univer.
BABOR,T. F,, HIGGINS-BIDDLE, J. C., SAUNDERS, 1. B., & MONTEIRO,M. G. (2001) The Alcohol
Ure Disorders Idrntzficution Tect: guidelines for use in primary health care. Geneva, Switz.:
World Health Organization.
BAKRATT IMPULSIVENESS SCALE-11: PSYCHOMETRICS 1181

BARRATT, E. S. (1959) Anxiety and impulsiveness related to psychomotor efficiency. Perceptual


and Motor Skzlls, 9, 191-198.
BARRATT, E. S. (1994) Im ulsiveness and aggression. In J. Monahan & H . J. Steadman (Eds.),
Violence and mentar dtsordev: developments in risk assessment. Chicago, IL: Univer. of
Chicago Press. Pp. 61-79.
BARRATT, E. S., &PATTON, J. H. (1983) Impulsivity: cognitive, behavioral, and psychophysiologi-
cal correlates. In M. Zuckerman (Ed.), Biological base, of sensation seektng, impulsivity
and anxtety. Hillsdale, NJ: Erlbaum. Pp. 77-116.
BARRATT, E. S., &STANFORD, M. S. (1995) Im ulsiveness. In C. G . Costello (Ed.), Per.\onality
characteristics of the personality dirorcleredP~ewYork: Wiley. Pp. 91-1 19.
BENTLER, P. M. (1990) Comparative fit indexes in structural models. Prychological Bulletin,
107, 238-246.
BOLLEN, K. A. (1989) A new incremental fit index for general structural equation models. So-
czological Methods and Research, 17, 303-316.
BROWWE, M. W., &CUDECK, R. (1993) Alternative ways of assessing model fit. In K. A. Bollen
& J. S. Long (Eds.), Testing structural equatzon models. Newbury Park, CA: Sage. Pp.
136-162.
DANIELSON, C. K., OVERHOLVER, J. C., &BUTT,A. 2.(2003) Association of substance abuse and
depression among adolescent psychiatric inpatients. Canadian Journal of l'sychiatry, 48,
761-765.
FOSSATI,A., BARRATT, E. S., ACQUARINI, E., & DE CEGLIE,A. (2002) Psychometric roperties of
an adolescent version of the Barratt Impulsiveness Scale-ll for a sample oP1talian high
school students. Perceptual and Motor Skills, 95, 621-635.
FOSSATI,A,, DI CEGLIE, A,, ACQUARINI, E., & BARRATT, E. S. (2001) Psychometric properties of
an Italian version of the Barratt Impulsiveness Scale (BIS-11) in nonclinical subjects. Jour-
nal of Clinical P.sychology, 57, 815-828.
Gu, B., &ROY,K. (1995) Sex ratio at birth in China with reference to other areas in East Asia:
what we know. Asia-Pacz5c Population Journal, 10, 17-42.
HULL,C . L. (1943) Principles ofhehavior. New York: Wiley Interscience.
HULSE,G. K. (2001) Impediments to screening for hazardous alcohol use and dependence in
general hospital psychiatric inpatients. Australian and New Zealand Journal of Psychiatty,
35, 606-612.
LINNOILA, M., VIRKKUNEN, M., SCHEININ, M., NUUTILA, A,, RIMON,R., &GOODWIN, F. K. (1983)
Low cerebrospinal fluid 5-hydr~x~indoleacetic acid concentration differentiates impulsive
from nonimpulsive violent behavior. Life Sciencec, 33, 2609-2614.
MCCARTY, C. A., WEISZ,J. R., WANITROMANEE, K., EASTMAN, K. L., SIJWANLERT,S., CHAIYASIT, W.,
& BAND.E. B (19991 Culture, coping, and context: frimary and secondary control
among Thai and American youth. Journal of Child Psycho ogy and Psychiatry, 40, 809-818.
MOELLER, F. G., BARRATT, E. S., DOUGHERTY, D. M., SCHMITZ,J. M., & SWANN,A. C. (2001)
Psychiatric aspects of impulsivity. American Journal ofPsychiatry, 158, 1783.1793.
PATTON, J. H., STANFORD, M. S., &BARRATT, E. S. (1995) Factor structure of the Barratt Impul-
siveness Scale. louvnal of Clinical Psychology, 5 1, 768-774.
PEDERSEN, W., & SKRONDAL, A. (1999) Ecstasy and new patterns of drug use: a normal popula-
tion study. Addiction, 94, 1695-1706.
SAUNDERS, J. B., AASLAND, 0.G., BABOR,T. F., DE LA FUENTE,J. R., &GRANT,M. (1993) Devel-
opment of the Alcohol Use Disorders Identification Test (AUDIT): W H O collaborative
project on early detection of persons with harmful alcohol consumption-11. Aildiction, 88,
791-804.
SCHWARZ, G. (1978) Estimating the dimension of a model. Annals of Statirtzcs, 6 , 461-464.
SKITCH,S. A. (2005) Quantity and Frequency Measure. Unpublished test, McGill Univer.
SOMEYA, T, SAKADO, K., SEKI,T., KOJIMA,M., REIST,C., TANG,S. W, & TAKAHASHI, S. (2001)
The Japanese version of the Barratt Impulsiveness Scale, 11th version (BIS-11): its reli-
ability and validity. Psychiatry and Clinical Neurosciences, 55, 111-114.
SPENCE, K. W. (1956) Behavior theory and conditioning. New Haven, CT: Yale Univer. Press.
SPINELLA, M., & MILEY,W. M. (2003) Impulsivity and academic achievement in college stu-
dents. College Student Journal, 37, 545-549.
STANFORD, M. S., GREVE,K. V., BOUDREAUX, J. K., &MATHIAS,C. W (1996) I~npuIsivenessand
S . YAO, ET A L .

risk-taking behavior: comparison of hi h school and college students using the Barratt
Impulsiveness Scale. Perc.onality and lnlbiilzlal Differences, 21, 1073-1075.
TA~OR J. ,A. (1958) The effects of anxiety level and psychological stress on verbal learning
journal of Abnormal and Social I'sychology, 57, 55-60.
TAYLOR,J. A,, &SPENCE, K. W. (1952) The relationship of anxiety level of performance in serial
learning. j0urnal of Experimental Psychology, 44, 6164.
WHITE,H. R, &LABOWVIE, E. \X! (1989) Towards the assessment of adolescent problem drink-
ing. journal of Studies oon Alcohol, 50, 30-37.

Accepted Aprd 17, 2007.

Вам также может понравиться