State-Trait Anger Expression Inventory - 2 (Staxi - 2) [PDF]

Citation preview

STATE-TRAIT ANGER EXPRESSION INVENTORY – 2 (STAXI -2) OVERVIEW Disclaimer. Investigators must follow all STAXI-2 copyright laws, and use due diligence in verifying the accuracy of all information provided in this report. Do not depend on this brief summary for any clinical, research, or other purposes. Use of this material is intended as background only for further reading, and there may be factual or other errors in this document. Domain and Purpose. Anger. The 57-item Staxi-2 (Spielberger, 1996) -II is a self-report instrument designed to evaluate anger in a psychological context, and to measure how the multiple dimensions of anger affect physical health. The STAXI-2 “assesses either the intensity of [individuals’] anger at a particular time or how frequently anger is experienced, expressed, suppressed, or controlled” (Spielberger, 1996, page 4). Administration. Written response form requires 6th grade reading level, or material may be read to test-takers. May be given to in Individual or group settings. Limited training required for administering or scoring, but interpretations requires professional training. The STAXI-2 takes about 12-15 minutes to complete. It is available in several languages. Background. Like the STAXI, the STAXI-2 measures State Anger (a psychological and physiological response to feeling unfairly treated, feeling frustrated, or being attacked) and Trait Anger (a person’s disposition or tendency to perceive situations as unfair, frustrating, or as an attack, and to respond to this with increased state anger. Spielberger (1996, page 2) describes one aspect of the state/trait anger relationship as follows: “Individuals with high trait anger scores experience state anger more often and with greater intensity than individuals who are low in trait anger.” State anger items how one feels “right now”; trait anger items ask about how the respondent “generally feels.” The STAXI and the STAXI-2 also measure aspects of anger expression and anger control. The STAXI-2 revises the original 44-item STAXI (Spielberger, 1988) to 57 items. Selection was based on an eight-factor solution accommodating gender differences. The primary changes were to construct subscales for three newly identified components of State Anger, to revise the former Anger Control into Anger Control-Out and the new Anger Control-In. anger items ask about how the respondent “generally feels.” Copyright. 1988, 1996, 1999 by Psychological Assessment Resources, Inc. 16204 N. Florida Ave., Lutz, Florida, 33549. 1.800.331.8378. www.parinc.com. DESCRIPTION The 57-item STAXI-II measures multiple dimensions of anger with six scales. In addition to State Anger and Trait Anger; the STAXI measures Anger Expression-In and Anger Expression Out (expressing or suppressing angry feelings, respectively), and Anger Control-Out (control of angry feelings by preventing their expression. The STAXI-2 introduces a new scale, Anger Control-In (controlling suppressed anger through calming oneself when angry), as well as new subscales. State Anger has three subscales and Trait Anger has two. A composite score, the Anger Expression Index, combines scores from the two Anger Expression and two Anger Control scales. Plots of transformed scale and subscale scores provide trained clinicians with a profile quantifying describing state and trait aspects of anger, and how anger is expressed or controlled, relative to age, gender, and psychiatric status norms. Because of the complex structure and somewhat nuanced concepts measured by these scales and subscales, exemplar items will be listed later. Scales and subscales. Adopted from Spielberger STAXI-2 Manual (1996) identifies scales and subscales, and the items that comprise them. Factor loadings are reported in tables 8- 11. Scale abbreviation Items Scale/subscale description Sample Items(s) “AF”= Angry feelings (n of Items) Subscale abbreviation State Anger



S-Ang



1-15 (15)



Intensity of AF, how much person feels like expressing it “at a particular time.”



8. feel like hitting someone (but any of items 1-15 count).



Feeling angry S-Ang/F Feel like expressing anger verbally S-Ang/V Feel like expressing anger physically S-Ang/P Trait Anger T-Ang



1-3, 6, 10 (5) 4, 9, 12, 13, 15 (5) 5, 7, 8, 11, 14 (5) 16-25 (10)



Angry Temperament T-Ang/T



16-18, 21 (4)



Angry Reaction



19, 20, 23, 25 (4)



T-Ang/R



Anger Expression-Out AX-O Anger Expression-In



AX-I



Anger Control-Out



AC-O



Anger Control-In



AC-I



Anger Expression Index AX Index



27, 31, 35, 39, 43, 47, 51, 55 (8) 29, 33, 37, 41, 45, 49, 53, 57 (8) 26, 30, 34, 38, 42, 46, 50, 54 (8) 28, 32, 36, 40, 44, 48, 52, 56 (8) AX index=AX-0+AXI – (ACO+AC-I) +48



Intensity of AF person is currently experiencing. Intensity of current anger related to the verbal expression of anger. Intensity of current anger related to the physical expression of anger. How often AF are experienced over time. Disposition to experience anger without specific provocation. Freq., experiencing AF in settings involving frustration/ negative eval. Freq., AF expressed in verbally or physically aggressive behavior. Freq., AF experienced but not expressed (i.e., suppressed). Freq., person controls outward expression of AF. Freq., person tries to control AF by “calming down” or “cooling off.” General index of anger expression based on AX-), AX-I, AC-0, AC-I.



3. feel angry 9. feel like swearing 11. feel like kicking someone 16. Am quick-tempered (but any of items 16-25 count) 17. Have a fiery temperament 25. Feel infuriated when do good job and get poor evaluation 27. Express anger. 41. Boil inside, but don’t show it. 42. Control behavior 40. Try to soothe [own] angry feelings



Subgroups. Spielberger (1996) presents STAXI-2 normative scores by gender, psychiatric status, and age group (see tables 2-5). We highlight here major normative results only, using the format (x, X vs. y,Y), where x and X represent male and female scores in the non-psychiatric sample and y and Y represent male and female scores in the psychiatric sample. Compared to the non-psychiatric sample, the psychiatric sample scored sig. higher on S-Ang. (17.90, 19.25 vs. 24.05, 22.71) and T-ang (17.89, 18.40 vs. 19.79, 20.14) and the 5 related subscales, as well as Anger Expression-In (15.86, 16.35 vs. 18.79, 18.26) and the Anger Expression Index (32.04, 33.68 vs. 36.80, 39.58). Their scores were sig. lower on the two Anger Control-In (23.28, 22.60 vs. 22.69, 21.39) and Anger Control Out (23.21, 23.53 vs. 22.31, 21.06); F ranged from 13.98 to 169.58, all at p < .001. Across both sample, women scored significantly lower on Anger Expression Out, Anger ControlIn, and the Anger Expression Index( F ranged from 5.02 to 6.24, all at p < .05). There were 5 group by gender interactions. Age group differences (16-19 years, 20-29 years, 30 and older) are reported for the non-psychiatric sample only. There were no age by gender interactions. In general, anger scale and subscales scores monotonically increased from older to younger age group. Taken together, the significant score differences by psychiatric status, gender, and age highlight the importance of using normative scales (Appendices A and B) to interpret scores. SCORING AND INTERPRETATION Scoring. The 15 State Anger items (items 1-15), respondents rate how angry they feel “right now” on a four point scale, ranging from “not at all” (1) to very much so (4). For the 10 Trait Anger items (items 16-25), respondents rate how they “generally” feel on a four-point scale, ranging from almost never (1) to Almost always (4). For the 32 items that make up Anger Expression-In and -Out, and Anger Control-In and –Out, respondents “rate how often they generally react or behave in certain ways when they feel angry or furious” (Spielberger, 1996, page 4) using the same 4-point scale used for Trait Anger. Scale and subscale scores are obtained by summing the values (always between 1 and 4, inclusive) for the corresponding items. Thus, the 15-item (items 1-15) State Anger scale score ranges from 15 to 60, and the 4-item (items 19, 20, 23, 25) Trait Anger-Reaction subscale State Anger-Physical subscale score ranges from 4 to 16. Using the normative tables (Appendix A and B), raw scores are converted into percentiles or T-scores (linearly transformed scores with a mean of 50 and an SD of 10). As reported above, norms differ by gender, psychiatric status, and age group, although only one age group (18 and older) is available for psychiatric patients. Raw scores and percentiles may be plotted on the STAXI-2 Profile Form for a graphic representation of the STAXI profile.



Missing Data. To minimize the occurrence of missing data in individual testing situations, test administrators should try to examine score sheets, and probe respondents about any items not answered. Otherwise, the Manual presents rules about discarding scale and subscale scores due to missing data; there are no guidelines for imputing values. For example, the measures is considered invalid if 10 or more items are not answered. The same disposition is true for any of the six major scales with > 2 missing items, or > 1 item on the S-Ang or T-Ang subscales. If a total of 8 or fewer items are missing, weights based on mean scores are applied: Missing values for S-Ang and T-Ang are assigned a 1; missing items on the 4 anger expression and anger control scales are assigned a 2. (See Manual, page 5.) Interpretation of Scores. Interpretation is based not only on scale and subscale scores, but on the profile of scores. For example, a clinician may evaluate an individual based on high scores on one or more scales and medium or low scores on others. Interpretation (e.g., a “high” score) will depend on norms based that differ by age group, gender, and psychiatric status. The Manual advises that anger scores between the 25th and 75th percentiles, inclusive, “may be considered to fall within the normal range” (page 13). Although scores within this range may differentiate anger intensity, frequency, expression, and control, “such differences are generally not sufficient to detect anger problems that may predispose individuals to develop psychological or psychological disorders.” Table 7 (page 14) presents some interpretation guidelines for selected clinical profiles. Partial Literature Review Article



Men



Women



Veterans



PTSD



Setting



Age



Other Info



Spielberger Normative data based on two samples. No information reported for veteran or PTSD status. et al, 1996, Sample 1. “Normal” adults” across several occupational levels, undergrad and grad students. N=1,644; 977 female, 667 male. Age approx. 27 (16 to 68)i Manual. Sample 2. Adults admitted for treatment of psychiatric and SUDS. N=276; 105 females, 171 males



Psychometrics Article



Spielberger et al, 1996, Manual.



Reliability



Validity



Cronbach’s Alpha; Item/Remainder (IR)



Construct, Convergent



By subgroups, acceptable (.72) to excellent (.95). Trait and state higher than expression scales, see tables 2-5. See table 12-13 for IR correlations.



Due to positively skewed S-Ang and T-Ang/Temperament distributions, these scales do not discriminate well among low scorers; however, consistently low cross-scale scores may be evidence of denial and/or repression. Scores > 75th percentile may interfere with functioning (Spielberger, 1988, 1996), and high AX-I scores (especially with low AX-O scores and high anxiety) are associated with high blood pressure and hypertension problems (e.g., Vogele et al., 1997).



LIMITATIONS Cut-offs suggested by Beck appear to over-identify person with depressive disorder; more research is needed to identify appropriate cutoffs and statistical properties for different populations. Future psychometric evaluations would be improved with structured interviews, multiple assessments, and other analyses to test-retest reliability, construct validity. Predictive validity, and the utility of the instrument in detecting change.



REFERENCES Arbisi, P. A., & Farmer, R. F. (2001). Review of the BDI-II. The fourteenth mental measurements yearbook. Retrieved July 2017 at http://blog.naver.com/mistyeyed73/220427762670. Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck Depression Inventory manual (2nd ed.). San Antonio. TX: Psychological Corporation. Beck, A. T., Rush, A. J., Shaw B. F., Emery G. (1979). Cognitive therapy of depression. New York: Guilford.



Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of general psychiatry, 4(6), 561-571. Buckley, T. C., Parker, J. D., & Heggie, J. (2001). A psychometric evaluation of the BDI-II in treatment-seeking substance abusers. Journal of substance abuse treatment, 20(3), 197-204. Dozois, D. J., Dobson, K. S., & Ahnberg, J. L. (1998). A psychometric evaluation of the Beck Depression Inventory– II. Psychological assessment, 10(2), 83. Palmer, G. A., Happe, M. C., Paxson, J. M., Jurek, B. K., Graca, J. J., & Olson, S. A. (2014). Psychometric properties of the Beck Depression Inventory-II for OEF/OIF veterans in a polytrauma sample. Military medicine, 179(8), 879-884. https://doi.org/10.7205/MILMED-D-14-00048 [NOTE: Sample composed of veterans with TBI, but focus almost exclusively on BDI-II factor structure] Santor, D. A., Ramsay, J. O., & Zuroff, D. C. (1994). Nonparametric item analyses of the Beck Depression Inventory: Evaluating gender item bias and response option weights. Psychological Assessment, 6, 255-270.