INTRODUCTION

Aggression is a multidimensional construct that develops within a complex interaction of biological, psychological, social, and cultural factors (1). While aggressive acts are state phenomena, trait aggressiveness refers to a disposition to behave aggressively across various situations and over repeated occasions (2). Aggression is a potential predictive factor of later criminal activity during adolescence (3) and a psychological topic of particular relevance in forensic and clinical contexts (4). Aggression continues to play a prominent role in clinical disorders, in conjunction with maladjusted and socially inappropriate behavior, in the case of criminal offenders, or even in conflicts at school or in a partnership (5). Also, previous studies explored the associations between trait aggression and suicide (6), particularly among dependent populations such as alcohol dependents (7) and cocaine dependents (8). Nevertheless, to predict if somebody will behave aggressively is not only important for the patient within the institutions himself, but also in some cases, it can be useful for the staff as well (4,9).

The association between substance abuse and aggressiveness is undoubtedly complex, but a series of investigations with first graders revealed that boys who were identified by teachers/peers as more aggressive are more likely to use drugs in the future (10). Early aggressive behavior was found predictive of later substance abuse (11) and conduct disorder symptoms have been observed to begin some years before regular drug use (12). White et al. (13) found that early aggressive behavior leads to an increase in alcohol use, and that alcohol use does not lead to later aggressive behavior. Leonard and Senchak (14) found a significant relationship between heavy alcohol use and premarital aggression. Naturalistic studies have also shown a link between severity of aggression and level of alcohol consumption (15-17). Finally, the trait-aggression level of patients with alcohol dependence was compared with the trait aggression level in a population sample, and a significantly higher level of aggression in the alcoholics was found (18). In conclusion, there is evidence suggesting that aggressive personality traits may predate addictive behavior.

Self-rated aggression measures remain a popular method to assess trait aggression. To reduce the time, effort, and resources involved in measuring aggression, Buss and Durkee (19) developed a 75-item self-report instrument with false-true format, the Buss-Durkee Hostility Inventory, which has quickly become the gold-standard for the measurement of aggression. The inventory is divided into seven scales across two main dimensions of overt aggression and hostility: Assault, Indirect Aggression, Irritability, Negativism, Resentment, Suspicion, and Verbal Aggression. Subsequent factor analyses revealed similar, but not identical, two-factor structures, named either aggressiveness and hostility (19), overt and covert hostility (20) or expressive and neurotic hostility (21). A meta-analysis of factor analyses (22) confirmed this two-factor structure.

To improve the psychometric properties of this instrument, Buss and Perry (23), more recently developed a 29-item self-report questionnaire, the Aggression Questionnaire (AQ) (23,24), which represents an updated and psychometrically improved version of the Buss-Durkee Hostility Inventory (19). In constructing this questionnaire, Buss and Perry (23) borrowed some items intact from the earlier Hostility Inventory, revised other Buss-Durkee items to improve clarity, and added many new items to generate an initial pool of 52 questions. Four correlated factors emerged —Physical Aggression, Verbal Aggression, Anger, and Hostility— on which a core set of 29 items loaded, and this four-factor structure appeared to replicate across all three samples (25). Thus, they designed the AQ to measure four dispositional subtraits of aggression, which they defined as follows: ‘‘Physical and Verbal Aggression (these two corresponding to the former Assault and Verbal Aggression), which involve hurting or harming others, represent the instrumental or motor component of behavior. Anger (made up only of Irritability items because Indirect Aggression items did not replicate), which involves physiological arousal and preparation for aggression, represents the emotional or affective component of behavior. Hostility (based on Resentment and Suspicion items), which consists of feelings of ill will and injustice, represents the cognitive component of behavior (5,23). Buss and Perry (23) also suggested that “after anger has cooled down”’, hostility might be “a cognitive residual of ill will, resentment and perhaps suspicion of others’ motives”.

Further, Williams et al. (26), Morren and Meesters (27) and, Diamond and Magaletta (28) assessed the properties of the AQ in offender samples. Two further proposals for the latent structure of the AQ are proposed in the literature, namely: a) a one-factor model, which assumes that all items load on one first order factor, and b) a second order factor implicated in the initial four-factor model (1). A more recent version of the AQ with an additional scale consisting of indirect aggression is available commercially (24).

As discussed previously, the AQ has been widely used and cited as a self-report measure of aggression. The 29-item AQ has been validated with North American (29-31) as well as British (32), Dutch (33), Swedish (34), Japanese (35), Spanish (36), Italian (37), German (38), Greek (1,39), Chinese (40), and Hungarian (41) populations.

Expression of cognitive and emotional components of aggressiveness might be different in different cultures. This could be an effect attributable to the different social pressure upon individual in different cultures (36). Cultures encouraging the expression of emotions may be different from those which lead to the inhibition of most of them (4). The aim of this study was to determine the validity and reliability of the Buss-Perry’s AQ Turkish version in male substance dependent inpatients.

METHODS

Participants

The study was conducted in Bakirköy Training and Research Hospital for Psychiatry, Neurology and Neurosurgery, Alcohol and Drug Research, Treatment and Training Center (AMATEM) in Istanbul between May 2009 and December 2009. AMATEM is a specialized center for substance use disorders with 84 inpatient beds, and accepts patients from all over Turkey. The ethical committee of the hospital approved the study protocol. Patients’ written informed consent was obtained after the study protocol was thoroughly explained.

One hundred consecutively admitted male alcohol dependent inpatients (without other substance use disorders) and 100 consecutively admitted male drug dependent inpatients (without other substance dependence) participated in the study. Alcohol and drug dependent inpatients were included in the study separately, because the study was conducted at the end of the detoxification processes of alcohol and drug dependent inpatients. In the drug dependent group (n=100); 38 (38.0%) patients were cannabis dependents, 32 (32.0%) were opiate-dependent, 7 (7.0%) were cocaine-dependent, and 23 (23.0%) were inhalant-dependent.

Translation

The original AQ was independently translated from English into Turkish by two experts in psychiatry. Consensus was reached on a common draft by these experts. This Turkish version was back translated into English by an independent translator.

Procedure

The final Turkish version of the AQ was applied to 200 substance dependent inpatients. All patients were assessed by using a semi-structured socio-demographic form. The diagnosis of alcohol dependence was based on the clinical examination, a screening interview based on the Structured Clinical Interview for DSM-IV (SCID-I) (42), Turkish version (43), conducted by a trained interviewer (CE). First interviews with the study group were conducted after detoxification periods (i.e. 3-4 weeks after the last day of alcohol or substance use). Second administration was repeated again after 2 weeks to 166 (83.0%) of these patients (76 from alcohol, 90 from drug dependents) in a test-retest procedure to assess the test-retest reliability, because the rest were drop-out from treatment before filling the scale second time.

Assessment Instruments

Buss-Perry’s Aggression Questionnaire (AQ): Trait aggression was measured by the total score of the AQ and scores of the subscales including Physical Aggression (items 1 to 9), Verbal Aggression (items 10 to 14), Anger (items 15 to 21), and Hostility (items 22 to 29). The AQ comprises 29 items in a 5-point Likert format from 0 (extremely uncharacteristic of me) to 4 (extremely characteristic of me). Evidence for the scale’s construct validity is available elsewhere (23). Internal consistency reliability is shown on Table 3.

State-Trait Anger Expression Inventory (STAXI): The original STAXI provides a self-reported measure of the experience and expression of anger in 44 items (44). Although original version designed to measure anger as a situational emotional response (state) and as a predispositional quality (trait), the Turkish version with 34 items that we used does not include state anger (45). Trait anger assesses the tendency to experience anger (i.e., anger proneness) so that higher scores indicate more frequent and intense anger. Individuals answered on a 4-point Likert scale (score range: 0-136). The STAXI also contains three scales designed to assess three different dimensions of the expression of anger: (a) Anger-In (anger suppression), (b) Anger-Out (aggressive anger expression), and (c) Anger-Control (inability to diminish the occurrence of angry feelings).

The Barratt Impulsiveness Scale (BIS-11): The measure of impulsiveness as a personality trait was the Barratt Impulsiveness Scale (BIS-11) (46,47). It includes 3 first order factors; Attentional, Motor, and Non-planning impulsiveness. This is a 30-item self-report questionnaire that has been validated to assess impulsiveness as long term patterns of behavior in various populations, including substance-dependent individuals and violent suicide attempters (47,48). Others have found evidence for the validity of this measure; BIS-11 scores are higher in patients with substance abuse problems (49-51). The Turkish version of BIS-11 was used in the present study (52).

State-Trait Anxiety Inventory (STAI): As a measure of state and trait anxieties, the Spielberger’s State-Trait Anxiety Inventory (STAI), a 40-item self-report instrument was used (53). Participants indicated their agreement with each item on a Likert scale ranging from 1=”not at all” to 4=”very much so”. The Turkish version of the STAI has been shown to have good reliability and validity (54). The Cronbach’s alpha was 0.94 for state anxiety and 0.91 for trait anxiety in the present study.

Beck Depression Inventory (BDI: Symptoms and severity of depression were evaluated by using the Beck Depression Inventory (BDI) (55), Turkish version (56). The Cronbach’s alpha was 0.90 for BDI in the present study.

Statistical Analysis

The statistical package SPSS 17.0 for Windows was used for all the analyses. Categorical variables were compared by means of the chi-square statistics. We used Student’s t-test to compare the groups on continuous variables. Factor analysis for AQ items was performed (Extraction Method: Principal Component Analysis, Rotation Method: Varimax with Kaiser Normalization). Correlation analyses (Pearson, bivariate) between the AQ items and their subscales, test-retest, AQ total and subscales of AQ and the STAXI and the BIS-11 were performed. Cronbach’s alphas were evaluated for internal consistency. For all statistical analyses, p values were two-tailed and differences were considered significant at p<0.05.

RESULTS

General Results

A total of 200 consecutive substance dependent male inpatients were included in the statistical analyses. The mean age of the participants was 35.39 (SD=10.54, range=18-64). Sociodemographic variables are shown on Table 1 (Table 1).

Factor Structure

Item 7, which is one of the two reversed items in AQ, showed low corrected item-total score (lower than 0.15) and test-retest correlations, and was deleted from the questionnaire. Table 2 presents the factor loadings of the AQ items (Table 2). An inspection of the eigenvalue curve reveals a clear 4-factor structure according to the scree-test. On Table 2, the 28 items of the AQ are listed according to the order of magnitude of their loadings obtained on the four factors.

The four factors account for 52.9% of the variance, the contribution of the single factors being 36.3%, 7.0%, 5.3%, and 4.3%. The corresponding eigenvalues were 16.77, 3.29, 2.53, and 2.03.

When taking a closer look at which items load on the separate factors, one may notice that all 9 items (items 1 to 9) of the Physical Aggression scale load together on one single factor (Factor 1). Additional to these items, 16th and 20th items which originally included in Anger scale loaded in Factor 1. The 8 Hostility items all belong to Factor 2. However, 4 items (items 17, 18, 19, and 21) from Anger scale and 3 items (items 11, 13, and 14) from Verbal Aggression scale have their primary loadings on F3. Similarly, 1 item (item 15) from Anger scale and 2 items (items 10 and 12) from Verbal Aggression scale have their primary loadings on Factor 4. It may therefore be concluded that the four dimensional structure of the AQ was confirmed, although the original assignment of the items to their subscales could not be reproduced without exception. Nevertheless, since Factor 3 and Factor 4 comprises items of the both Anger and Verbal Aggression scales, we decided to evaluate psychometric properties of Factor 3 and Factor 4, both separately and together as a single factor.

Item-Subscale and Corrected Item-Total

Correlations

Item-subscale correlations ranged between 0.54 and 0.80 (p<0.001) (Table 3). When Factor 3 and Factor 4 were evaluated together, item-subscale (items 10, 11, 12, 13, 14, 15, 17, 18, 19, and 21) correlations ranged between 0.45 and 0.73 (not shown). For each of the items, the corrected item-total correlation values were between 0.30 and 0.77 (p<0.001) (Table 3).

Reliability and Validity

Internal Consistency Coefficient

In substance dependents, the internal consistency coefficient (Cronbach’s alpha) was 0.89 for Factor 1 (Physical Aggression), 0.84 for Factor 2 (Hostility), 0.82 for Factor 3, 0.59 for Factor 4, and 0.93 for AQ. Cronbach’s alphas’ are shown on Table 2. Mean scores of subscales and AQ total score and correlations between them are shown on Table 3. Also Cronbach’s alphas’ of Physical Aggression, Hostility, and AQ from studies of Buss and Perry (23) and Vitoratou et al. (1), are shown on this table to compare with the results of the present study (Table 4).

When Factor 3 and Factor 4 were evaluated together, correlation coefficient was 0.77 with Factor 1, 0.63 with Factor 2, and 0.91 with AQ. Also Cronbach’s alpha was 0.81 (not shown).

Test-Retest Reliability

Test-retest correlations were 0.80 for Factor 1 (Physical Aggression), 0.77 for Factor 2 (Hostility), 0.78 for Factor 3, 0.54 for Factor 4, and 0.84 for AQ total score (Table 3).

When Factor 3 and Factor 4 was evaluated together test-retest value was 0.77 (not shown).

Concurrent Validity

All four factors (and when Factor 3 and Factor 4 evaluated together) and total score of AQ were correlated significantly in the degree of p<0.001 with total and subscale scores of the STAXI and the BIS-11, BDI, STAI-S, and STAI-T (Table 4).

Discriminating Power for Specific Group’s

Validity

Turkish version of Buss-Perry AQ total and subscale (not shown, p<0.001) scores were higher among those with history of suicide attempt (n=66, 86.77±21.47, t=-5.25, p<0.001), among those with history of fights (n=61, 91.93±20.34, t=-7.95, p<0.001) and among those with history of trouble with law (n=103, 83.32±20.70, t=-5.17, p<0.001) than those without these histories (n=134, 71.06±19.09, n=139, 69.36±17.63 and n=97, 68.73±19.09 respectively) (not shown).

DISCUSSION

The Aggression Questionnaire (AQ) (23) is a widely used self-reported measure of trait aggression. In the present study, the Turkish version of this questionnaire with 28-item and four factor solution was found to be compatible with the original scale among substance dependent inpatients. The results suggest that each subscales and AQ had an adequate reliability in terms of internal consistency. The item-subscale and the corrected item-total correlation coefficient values were significant at moderate to high degrees and were stable over two weeks of testing. Finding Buss-Perry’s AQ and its subscales correlated with related constructs such as anger (measured with STAXI) and impulsivity (measured with BIS-11) showed concurrent validity. The appropriateness of the 4-factor model posited by Buss and Perry (23) was further examined by other studies and was supported (35,38).

A number of researchers have investigated the reliability and validity of this instrument, and the findings generally point in the direction of good psychometric properties with the qualification that the elimination of a few items would improve the instrument (57). Specifically, Harris (30) proposed to omit two items from the Hostility subscale (H6 and H8; items 27 and 29); Meesters et al. (33) further omit one more item (first indicator of the Verbal aggression subscale); in addition, Bryant and Smith (25) proposed a modified 12-item version of the AQ. The seventh item of the Physical Aggression (7th item) and fourth item of the Anger (18th item) from the Buss and Perry’s AQ items, which are the only negatively worded and the reversed scored items in AQ, were found to have relatively low factor loadings in the Japanese version of the AQ (35). Authors suggested that the results of Japanese version of the AQ may be improved psychometrically if these two reversed scored items were removed from the scale for cross-cultural use (35). Gerevich et al. (41) also suggested excluding the two inverse items from the AQ, because of the low reliability of these items with regard to their hypothesized constructs. Similarly, when evaluating the Turkish version of the scale, the 7th item showed low corrected item-total and test-retest correlations and was removed from the questionnaire. It was suggested that responses to affirmatively worded items differ from those to negatively worded items in inventories (58). The difference in responses to those two types of items may be much more remarkable for some cultures. The results, however, might be due to a translation artifact rather than the reflection of cross-cultural differences in the dimensionality of aggression. Thus item, “I can think of no good reason for ever hitting a person” might be difficult to understand when translated to Turkish. In summary, the Turkish version of the scale with 28-item solution was found to be compatible with the original scale.

Several studies on adaptations of the AQ in different languages and cultures (29-41), although being able to confirm the four-factor structure, found some disagreement with the original model of Buss and Perry (23) on an item level and with respect to the interpretation and denomination of the factors (38). In the present sample, Physical Aggression (Factor 1) and Hostility (Factor 2) were consistent with the original questionnaire, whereas Factor 3 and Factor 4 computed from items that were originally from Anger and Verbal Aggression subscales. Thus, although the expected four-dimensional structure of the AQ was confirmed in the present study, the assignment of the items from Anger and Verbal Aggression subscales did not agree with the original scale, which showed considerable overlap with each other computing Factor 3 and Factor 4. Because a great many of the previous studies on the AQ were conducted with student samples with a preponderance of female participants, or with individuals from a general adult population including both genders, it is difficult to compare the results of our sample with these studies. Although these studies had possible limitation of themselves to generalize the results to a broader population, it is evident, however, that our sample was severely biased, too: patients included in this study were all male and the study group was restricted to a treatment population. Men were found to be more aggressive in PA than women in study of Buss and Perry (23), whereas some cultural differences were found for other subscales when considering gender (36). With respect to the gender effect, the well known results with men having higher Physical Aggression scores were consistently observed in samples of general population (1,38) and drug dependent individuals (1), whereas higher scores in the Anger subscale for female participants observed in general population (38). It is evident that male participants were more likely to score higher in the motor component of aggressive behavior whereas females showed a stronger tendency to express their anger instead (38). Cultures encouraging the expressing of emotions may be different from those which result inhibiting most of them (4). Disinhibiting (59) effect of substances and their relationship with alexithymia (60) should also be taken into consideration when discussing the results. Thus, although the possibility of an inaccurate translation of the item wording cannot be excluded, there might be semantic overlap between statements of verbal dispute (being argumentative, disagreeing openly) or Physical Aggression and an accompanying emotion of anger that might be responsible for this result specific to our sample. Nevertheless, the expression of cognitive and emotional components of aggressiveness might be different in different context (4,36).

Berkowitz’s theory (61) proposes that negative affect activates ideas, memories, angry feelings, and expressive-motor reactions. The Turkish version of the AQ was correlated with negative affect, such as depression, state and trait anxiety, which supports this theory. Based on correlations between AQ scales, Buss and Perry (23) suggested anger was the affective component that bridges the cognitive component of hostility and the instrumental components of verbal and physical aggression. They explained that correlations between anger and aggression made sense because anger often precedes aggression and aggression is more likely to occur when people are angry. In fact, many researchers accept the view that hostile aggression is motivated by anger (62). Others pointed out that although anger may motivate aggression, it often leads to responses other than aggression (63), which may depend on such cognitive factors as attributions and expectations (61). In our sample, it may have resulted in using substances as a mean of coping with anger, since sample population had severe substance dependents enough to seek treatment as an inpatient. Substance use may also cause verbal or physical aggression by the disinhibiting role of the substance (59). Consistent with this in a previous study at lower severity of alcohol use (total AUDIT score <8), the clinical sample showed a substantially higher severity of PA than the control sample, but at higher severity of alcohol use the control sample displayed a higher score of PA compared to the clinical sample (64). Thus, clinical populations in different severity may need different risk management strategies in order to address their needs in rationalistic way.

In creating the AQ, Buss and Perry (23) found that the aggression scales were positively related to measures of impulsiveness, as well as to other measures of aggression (31). Result of present study supported these findings such that AQ showed high correlations with anger (measured with STAXI) and impulsivity (measured with BIS-11) suggesting concurrent validity.

Our findings must be understood in the light of several limitations. First, patients included in this study were all male and the study group was restricted to a treatment population. Therefore it is not possible to generalize the findings to female substance dependent patients and non-treatment groups. A second limitation was that although participants were not assessed during withdrawal, patients might still have some cognitive problems to evaluate themselves correctly at the time of the interview, since we used self-report instruments. Despite of these limitations, results which were obtained in this study suggests that the Turkish version of the Buss-Perry’s AQ 28-item solution could be used as reliable and valid tool for substance dependent inpatients. Although the results supported the four factor scale, psychometric evaluation showed that Factor 3 and Factor 4 can also be scored together as a single factor.



REFERENCES

1. Vitoratou S, Ntzoufras I, Smyrnis N, Stefanis NC. Factorial composition of the Aggression Questionnaire: a multi-sample study in Greek adults. Psychiatry Res 2009; 168:32-39.

2. Tremblay PF, Dozois DJA. Another perspective on trait aggressiveness: overlap with early maladaptive schemas. Pers Individ Dif 2009; 46:569-574.

3. Pulkkinen L, Pitkanen T. Continuities in aggressive behaviour from childhood to adulthood. Aggress Behav 1993; 19:249-263.

4. Gallardo-Pujol D, Kramp U, García-Forero C, Pérez-Ramírez M, Andrés-Pueyo A. Assessing aggressiveness quickly and efficiently: the Spanish adaptation of Aggression Questionnaire-refined version. Eur Psychiatry 2006; 2:487-494.

5. Grumm M, von Collani G. Personality types and self-reported aggressiveness. Pers Individ Dif 2009; 47:845-850.

6. Giegling I, Olgiati P, Hartmann AM, Calati R, Möller HJ, Rujescu D, Serretti A. Personality and attempted suicide. Analysis of anger, aggression and impulsivity. J Psychiatr Res 2009; 43:1262-1271.

7. Bácskai E, Czobor P, Gerevich J. Suicidality and trait aggression related to childhood victimization in patients with alcoholism. Psychiatry Res 2009; 165:103-110.

8. Roy A. Characteristics of cocaine dependent patients who attempt suicide. Arch Suicide Res 2009; 13:46-51.

9. Sjöström N, Eder DN, Malm U, Beskow J. Violence and its prediction at a psychiatric hospital. Eur Psychiatry 2001; 16:459-465.

10. Kellam SG, Ialongo N, Brown H, Laudolff J, Mirsky A, Anthony B, Ahearn M, Anthony J, Edelsohn G, Dolan L. Attention problems in first grade and shy and aggressive behaviours as antecedents to late heavy or inhibited substance use. NIDA Res Monogr 1989; 95:368-369.

11. Moffitt TE. Adolescence-limited and life-course-persistent antisocial behavior: a developmental taxonomy. Psychol Rev 1993; 100:674-701.

12. Young SE, Mikulich SK, Goodwin MB, Hardy J, Martin CL, Zoccolillo MS, Crowley TJ. Treated delinquent boys’ substance use: onset, pattern, relationship to conduct and mood disorders. Drug Alcohol Depend 1995; 37:149-162.

13. White HR, Brick J, Hansell S. A longitudinal investigation of alcohol use and aggression in adolescence. J Stud Alcohol 1993; 11:62-77.

14. Leonard K, Senchak M. Alcohol and premarital aggression among newlywed couples. J Stud Alcohol Suppl 1993; 11:96-108.

15. Leonard KE, Collins RL, Quigley BM. Alcohol consumption and the occurrence and severity of aggression: an event-based analysis of male to male barroom violence. Aggress Behav 2003; 29:346-365.

16. Wells S, Graham K. Aggression involving alcohol: relationship to drinking patterns and social context. Addiction 2003; 98:33-42.

17. Graham K, Osgood DW, Wells S, Stockwell T. To what extent is intoxication associated with aggression in bars? A multilevel analysis. J Stud Alcohol 2006; 67:382-390.

18. Gerevich J, Bácskai E, Czobor P. Aggression levels in treatment seeking inpatients with alcohol-related problems compared to levels in the general population in Hungary. J Nerv Ment Dis 2007; 195:669-672.

19. Buss AH, Durkee A. An inventory for assessing different kinds of hostility. J Consult Psychol 1957; 21:343-349.

20. Bendig AW. Factor analytic scales of covert and overt hostility. J Consult Psychol 1962; 26: 200.

21. Siegman AW, Dembroski TM, Ringel N. Components of hostility and the severity of coronary artery disease. Psychosom Med 1987; 49:127-135.

22. Bushman BJ, Cooper H M, Lemke KM. Meta-analysis of factor analyses: an illustration using the Buss-Durkee Hostility Inventory. Pers Soc Psychol Bull 1991; 17:344-349.

23. Buss AH, Perry M. The aggression questionnaire. J Pers Soc Psychol 1992; 63:452-459.

24. Buss AH, Warren WL. The Aggression Questionnaire manual. Los Angeles, CA: Western Psychological Services, 2000.

25. Bryant FB, Smith BD. Refining the architecture of aggression: a measurement model for the Buss–Perry Aggression Questionnaire. J Res Pers 2001; 35:138-167.

26. Williams TY, Boyd JC, Cascardi MA, Poythress N. Factor structure and convergent validity of the aggression questionnaire in an offender population. Psychol Assess 1996; 8:398-403.

27. Morren M, Meesters C. Validation of the Dutch version of the aggression questionnaire in adolescent male offenders. Aggress Behav 2002; 28:87-96.

28. Diamond PM, Magaletta PR. The short-form Buss-Perry Aggression Questionnaire (BPAQ-SF): a validation study with federal offenders. Assessment 2006; 13:227-240.

29. Bernstein IH, Gesn PR. On the dimensionality of the Buss/Perry Aggression Questionnaire. Behav Res Ther 1997; 35:563-568.

30. Harris JA. Confirmatory factor analysis of The Aggression Questionnaire. Behav Res Ther 1995; 33:991-993.

31. Harris JA. A further evaluation of the Aggression Questionnaire: issues of validity and reliability. Behav Res Ther 1997; 35:1047-1053.

32. Archer J, Kilpatrick G, Barmwell R. Comparison of two aggression inventories. Aggress Behav 1995; 21:371-380.

33. Meesters C, Muris P, Bosma H, Schouten E, Beuving S. Psychometric evaluation of the Dutch version of the Aggression Questionnaire. Behav Res Ther 1996; 34:839-843.

34. Prochazka H, Agren H. Aggression in the general Swedish population, measured with a new self-rating inventory: the Aggression Questionnaire--revised Swedish version (AQ-RSV). Nord J Psychiatry 2001; 55:17-23.

35. Nakano K. Psychometric evaluation on the Japanese adaptation of the Aggression Questionnaire. Behav Res Ther 2001; 39:853-858.

36. García-León A, Reyes GA, Vila J, Pérez N, Robles H, Ramos MM. The Aggression Questionnaire: a validation study in student samples. Span J Psychol 2002; 5:45-53.

37. Fossati A, Maffei C, Acquarini E, Di Ceglie A. Multigroup confirmatory component and factor analyses of the Italian version of the aggression questionnaire. Eur J Psychol Assess 2003; 19:54-65.

38. von Collani G, Werner R. Self-related and motivational constructs as determinants of aggression. An analysis and validation of a German version of the Buss–Perry aggression questionnaire. Pers Individ Dif 2005; 38:1631-1643.

39. Tsorbatzoudis H. Psychometric evaluation of the Greek version of the Aggression Questionnaire. Percept Mot Skills 2006; 102:703-718.

40. Maxwell JP. Development and preliminary validation of a Chinese version of the Buss-Perry Aggression Questionnaire in a population of Hong Kong Chinese. J Pers Assess 2007; 88:284-294.

41. Gerevich J, Bácskai E, Czobor P. The generalizability of the Buss-Perry Aggression Questionnaire. Int J Methods Psychiatr Res 2007; 16:124-136.

42. First MB, Spitzer RL, Gibbon M, Williams JBW. Stuructured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), Clinical Version. Washington D.C. and London: American Psychiatric Press, Inc., 1997.

43. Çorapçıoğlu A, Aydemir O, Yıldız M, Esen A, Köroğlu E. DSM-IV Eksen I Bozuklukları (SCID-I) için Yapılandırılmış Klinik Görüşme, Klinik Versiyon. Hekimler Yayın Birliği, Ankara, 1999 (Article in Turkish).

44. Spielberger CD. State-trait anger expression inventory. Odessa, FL: Psychological Assessment Resources, Inc., 1988.

45. Özer AK. Sürekli öfke ve öfke ifade tarzı ölçekleri ön çalışması. Türk Psikoloji Dergisi 1994; 9:26-35 (Article in Turkish).

46. Barratt ES. Anxiety and impulsiveness related to psychomotor efficiency. Percept Mot Skills 1959; 9:191-198.

47. Patton JH, Stanford MS, Barratt ES. Factor structure of the Barratt impulsiveness scale. J Clin Psychol 1995; 51:768-774.

48. Allen TJ, Moeller FG, Rhoades HM, Cherek DR. Impulsivity and history of drug dependence. Drug Alcohol Depend 1998; 50:137-145.

49. Moeller FG, Dougherty DM, Barratt ES, Schmitz JM, Swann AC, Grabowski J. The impact of impulsivity on cocaine use and retention in treatment. J Subst Abuse Treat 2001; 21:193-198.

50. Moeller FG, Dougherty DM, Barratt ES, Oderinde V, Mathias CW, Harper RA, Swann AC. Increased impulsivity in cocaine dependent subjects independent of antisocial personality disorder and aggression. Drug Alcohol Depend 2002; 68:105-111.

51. Patkar AA, Murray HW, Mannelli P, Gottheil E, Weinstein SP, Vergare MJ. Pre-treatment measures of impulsivity, aggression and sensation seeking are associated with treatment outcome for African-American cocaine-dependent patients. J Addict Dis 2004; 23:109-122.

52. Güleç H, Tamam L, Güleç MY, Turhan M, Karakuş G, Zengin M, Stanford MS. Psychometric properties of the Turkish version of the Barratt Impulsiveness Scale-11. Klinik Psikofarmakoloji Bülteni 2008; 18:251-258.

53. Spielberger C, Gorsuch R, Lushene R. Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologist Press, 1970.

54. Öner N, Compte AL. Süreksiz durumluk/Sürekli Kaygı Envanteri El kitabı. 2. Basım, Boğaziçi Üniversitesi Yayınevi, İstanbul, 1998 (Article in Turkish).

55. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4:561-571.

56. Hisli N. Reliability and validity of Beck Depression Inventory among university students. Journal of Turkish Psychology 1989; 7:3-13.

57. Tremblay PF, Ewart LA. The Buss and Perry Aggression Questionnaire and its relations to values, the Big Five, provoking hypothetical situations, alcohol consumption patterns, and alcohol expectancies. Pers Individ Dif 2005; 38:337-346.

58. Mook J, Kleijn WC, van der Ploeg HM. Symptom-positively and negatively worded items in two popular self-report inventories of anxiety and depression. Psychol Rep 1991; 69:551-560.

59. Hawton K, Harriss L, Simkin S, Bale E, Bond A. Self-cutting: patient characteristics compared with self-poisoners. Suicide Life Threat Behav 2004; 34:199-208.

60. Evren C, Kose S, Sayar K, Ozcelik B, Borckardt JP, Elhai JD, Cloninger CR. Alexithymia and temperament and character model of personality in alcohol-dependent Turkish men. Psychiatry Clin Neurosci 2008; 62:371-378.

61. Berkowitz L. On the formation and regulation of anger and aggression: A cognitive-neoassociationistic analysis. Am Psychol 1990; 45:494-503.

62. Spielberger CD, Reheiser EC, Sydeman SJ. Measuring the experience, expression, and control of anger. Issues Compr Pediatr Nurs 1995; 18:207-232.

63. Averill JR. Studies on anger and aggression: implications for theories of emotion. Am Psychol 1983; 38:1145-1160.

64. Gerevich J, Bácskai E, Czobor P. Heavy drinking as a differential predictor of physical aggression in clinical and general populations. Prog Neuropsychopharmacol Biol Psychiatry 2008; 32:668-672.