Primary Image

Rehab Measures Database

Novaco Anger Scale and Provocation Inventory

Last Updated

Purpose

The Novaco Anger Scale and Provocation Inventory (NAS-PI) is an 85-item norm-referenced self-report measure designed to describe how individuals experience anger and what types of situations provoke anger. (WPS, 2025)

Link to Instrument

Link to Instrument

Acronym NAS-PI

Area of Assessment

Executive Functioning
Cognition

Assessment Type

Patient Reported Outcomes

Cost

Not Free

Actual Cost

$238.80

Cost Description

Cost is for a kit that includes the manual plus 25 auto-score test forms. Individually, the cost for the components are: Manual-$146.40, 25 auto-score test forms-110.40, & pad of 100 answer sheets-46.80. These are available at https://www.mindresources.com/psychology/053923 (a note states the items are only available for sale within Canada)

CDE Status

Not a CDE--last searched 4/10/2025

Key Descriptions

  • 85 items; 60 on the Novaco Anger Scale (NAS) and 25 on the Provocation Inventory (PI)
  • The NAS provides four subscores (cognition, arousal, behavioral, and anger regulation) and a NAS total score
    ○ Responses for the NAS are scored on a 3-point scale (1 = Never true, 2 = Sometimes true, 3 = Always true)
    ○ Total NAS scores range from 52 to 140 and are converted to a percentile score
  • The PI provides one total PI score based on five types of situations that can lead to anger (unfairness, frustration, irritations, disrespectful treatment, and annoying traits of others)
    ○ Responses for the PI are scored on a 4-point scale from 1 = Not at all angry to 4 = Very angry
    ○ Total PI scores range from 25 to 100 and are converted to a percentile score

Number of Items

85 total items (60 in the Novaco Anger Scale and 25 in the Provocation Inventory)

Equipment Required

  • Writing utensil
  • Test forms
  • Manual

Time to Administer

25 minutes

Required Training

Training Course

Required Training Description

Level C: No training required for those with a master’s degree (e.g., MA, MS, MSW, CAGS) in psychology or a related field (e.g. occupational therapy, speech-language pathology, special education). For those with a Bachelor’s degree (e.g., BA or BS) in psychology or a related field (e.g. occupational therapy, speech-language pathology, special education), additional training in the form of a license or certification from an agency/organization that provides training and experience in assessment. (WPS, 2025)

Age Ranges

Child

9 - 12

years

Adolescent

13 - 18

years

Adults

19 - 64

years

Elderly Adults

65 +

years

Instrument Reviewers

Updated on 4/10/2025 by UIC Occupational Therapy Students Nicole Feliciano, Elizabeth Murphy, Kathryn Smith, and Joey Tepper under the direction of Sabrin Rizk, PhD, OTR/L, Department of Occupational Therapy, University of Illinois Chicago.

ICF Domain

Body Function

Measurement Domain

Emotion
Cognition

Professional Association Recommendation

None found--last searched 4/10/2025

Considerations

  • The NAS-PI has not been widely studied across populations in general, and not at all within the context of Occupational Therapy at the time of this evaluation as far as the reviewers can determine. Of the studies done, however, the validity and reliability of these assessments has been high.
  • Of the studies done exploring the validity and reliability of the NAS-PI, most were administered to people who were incarcerated or forensic participants (Moeller et al., 2015; Novaco & Taylor, 2004; Mills et al., 1998; Baker et al., 2008; Lindqvist et al., 2005; & Chilvers & Thomas, 2011), those with a psychiatric condition (Naz & Khalily, 2015; Skeem & Mulvey, 2001), or where one portion of the study was done with children (Moeller et al., 2015)
  • Although the manual states the test can be performed in 25 minutes or less, consideration should be taken if using the assessment with populations who have decreased cognition, a learning disability, or difficulty understanding their own emotions. There has not been sufficient research regarding the reliability and validity of the NAS-PI with these populations.
  • Currently, the assessment is available in English, Dutch, Swedish, German, and Spanish. 

 

Mental Health

back to Populations

Internal Consistency

Hospitalized Psychiatric Patients: (Novaco, 2012, = 142)

  • Excellent internal consistency for NAS total (α = 0.95*) and PI total (α = 0.95*)

 

Post-Traumatic Stress Disorder (PTSD): (Novaco, 2012; = 114 Vietnam combat veterans)

  • Excellent internal consistency for NAS total (α = 0.97*) and PI total (α = 0.96*)

*Scores higher than 0.9 may indicate redundancy in the scale questions.

 

Construct Validity

Convergent validity:

Male and female psychiatric patients: (Novaco, 1994; = 141)

Excellent correlations between the NAS Total score and the Buss-Durkee Hostility Inventory Total Score (r = 0.82); the Caprara scales of Irritability (r = 0.78) and Rumination (= 0.69); the Cook-Medley Hostility (r = 0.68); and the STAXI Trait Anger Scale (r = 0.84)

Male psychiatric patients: (Novaco & Renwick, 1998; = 112 male patients at a high-security forensic facility)

  • Excellent correlations between the NAS Total score and the STAXI anger disposition measures of Trait Anger (r = 0.75); Anger In (internalized anger (r = 0.55); Anger Out (externalized anger (r = 0.68); and Anger Expression Total (r = 0.73)
  • Adequate correlation between the NAS Total score and the STAXI anger disposition measure of Anger In (internalized anger) (r = 0.55)
  • Excellent correlation between the PI and the STAXI anger disposition measure of Trait Anger (r = 0.64)
  • Adequate correlation between the PI and the STAXI anger disposition measure of Anger Expression Total (r = 0.54)

     

Discriminant validity:

Post-Traumatic Stress Disorder (PTSD): (Novaco & Chemtob, 2002; = 143 Vietnam combat veterans, mean age = 50.5 (7.7) years, mean years of education = 14.3 (2.5)

  • Significant ability of the NAS measures and Provocation Index (PI) to discriminate between those diagnosed with and without PTSD

NAS Measures and Provocation Index (PI) Score by PTSD Diagnosis for Combat Veterans

Measure

No PTSD (= 40)

PTSD (= 41)

t*

Cognition

28.7 (6.8)

38.3 (5.4)

7.08

Arousal

25.9 (5.4)

38.8 (5.1)

11.10

Behavior

25.8 (6.7)

36.7 (7.2)

7.08

Total

80.4 (15.6)

113.9 (16.2)

9.49

PI

57.8 (14.7)

79.4 (13.3)

6.85

*All values < 0.001

 

Non-Patient

back to Populations

Standard Error of Measurement (SEM)

Undergraduate Students: (Culhane & Morera, 2010; = 257 U.S. Hispanics and = 246 U.S. Non-Hispanic Whites, SEM calculated)

  • SEM for non-Hispanic white (n = 246): 3.43
  • SEM for Hispanic (n = 257): 3.88

 

Minimal Detectable Change (MDC)

Undergraduate Students: (Culhane & Morera, 2010; MDC calculated)

  • MDC95 for non-Hispanic white (= 246): 9.52
  • MDC95 for Hispanic (= 257): 10.77

 

Normative Data

Standardization Sample: (Novaco, 2012; = 1,546 (= 652 males, mean age = 29 (19) and = 893 females, mean age = 32 (19))

Mean NAS-PI T-scores for Males and Females in the Standardization Samplea

NAS-PI Scores

Malesb

Femalesc

 

 

 

NAS

 

 

 

    Cognitive domain

51.0

49.0*

 

 

 

   Arousal domain

49.7

50.2

 

 

 

   Behavioral domain

51.8*

48.7*

 

 

 

   Regulation domain

49.0

50.6

 

 

 

   NAS Total

51.1

49.3

 

 

 

PI

 

 

 

   PI Total

49.8

50.1

 

 

 

a= 1,546, b= 652, c= 893

*< 0.01

 

Standardization Sample: (Novaco, 2012; = 1,546 (= 131 with African American background, mean age = 26 (18) and = 102 with Hispanic/Latino background, mean age = 26 (17))

Mean NAS-PI T-scores for Individuals from African American and Hispanic/Latino Backgrounds in the Standardization Samplea

NAS-PI Scores

African Americanb

Hispanic/Latinoc

 

 

 

NAS

 

 

 

    Cognitive domain

53.2*

48.3

 

 

 

   Arousal domain

51.6

47.7

 

 

 

    Behavioral domain

53.5*

48.4

 

 

 

    Regulation domain

47.2*

49.1

 

 

 

   NAS Total

53.1*

48.0

 

 

 

PI

 

 

 

   PI Total

52.3

49.4

 

 

 

a= 1,546, b= 131, c= 102

*< 0.01

 

Standardization Sample: (Novaco, 2012; = 1,546, age range = 9-84)

Mean NAS-PI raw scores for the Standardization Samplea

NAS-PI Scores

Mean

SD

 

 

 

NAS

 

 

 

    Cognitive domain

29.1

5.0

 

 

 

   Arousal domain

28.2

5.5

 

 

 

    Behavioral domain

26.7

6.6

 

 

 

    Regulation domain

25.3

3.7

 

 

 

   NAS Total

83.9

15.6

 

 

 

PI

 

 

 

   PI Total

63.7

14.8

 

 

 

a= 1,546

 

Internal Consistency

Standardization Sample Population from Manual: (Novaco, 2012; = 1,546)

  • Excellent internal consistency estimates for the full sample for both the NAS Total score (α = 0.94*) and PI Total score (α = 0.95*)

*Scores higher than 0.9 may indicate redundancy in the scale questions.

 

Mixed Conditions

back to Populations

Standard Error of Measurement (SEM)

Male Inmates: (Calculated from Moeller et al., 2016)

  • SEM for Provocation Inventory (PI) (n = 167): 3.960

Male Forensic Patients: (Calculated from Moeller et al., 2016)

  • SEM for NAS total (n = 64): 4.445
  • SEM for NAS-cognitive (n = 64): 2.594
  • SEM for NAS-arousal (n = 64): 2.658
  • SEM for NAS-behavioral (n = 64): 2.170
  • SEM for NAS-regulation (n = 64): 2.110

Male Offenders: (Calculated from Baker et al., 2008)

  • SEM for NAS total (n = 638): 4.003
  • SEM for NAS-cognitive (n = 638): 1.944
  • SEM for NAS-arousal (n = 638): 2.223
  • SEM for NAS-behavioral (n = 638): 1.545

Female Offenders: (Calculated from Baker et al., 2008)

  • SEM for NAS total (n = 349): 5.380
  • SEM for NAS-cognitive (n = 349): 2.066
  • SEM for NAS-arousal (n = 349): 2.420
  • SEM for NAS-behavioral (n = 349): 2.862

 

Minimal Detectable Change (MDC)

Male Inmates: (Calculated from Moeller et al., 2016)

  • MDC95 for PI (n = 167): 10.98

Male Forensic Patients: (Calculated from Moeller et al., 2016)

  • MDC95 for NAS total (n = 64): 12.32
  • MDC95 for NAS-cognitive (n = 64): 7.19
  • MDC95 for NAS-arousal (n = 64): 7.37
  • MDC95 for NAS-behavioral (n = 64): 6.02
  • MDC95 for NAS-regulation (n = 64): 5.85

Male Offenders: (Calculated from Baker et al., 2008)

  • MDC95 for NAS-total (n = 638): 11.10
  • MDC95 for NAS-cognitive (n = 638): 5.39
  • MDC95 for NAS-arousal (n = 638): 6.16
  • MDC95 for NAS-behavioral (n = 638): 4.28

Female Offenders: (Calculated from Baker et al., 2008)

  • MDC95 for NAS-total (n = 349): 14.91
  • MDC95 for NAS-cognitive (n = 349): 5.73
  • MDC95 for NAS-arousal (n = 349): 6.71
  • MDC95 for NAS-behavioral (n = 349): 7.93

 

Normative Data

Incarcerated Adult Offenders: (Baker et al., 2008; = 1,308, mean age = 32.7 (8.08))

Novaco Anger Scale (NAS) Means and Standard Deviations for Male and Female Offender Groups

NAS Scores

Male Offendersa M(SD)

Female Offendersb M(SD)

t

Part A

86.47 (15.13)

88.28 (16.22)

1.76

   Cognitive dom.

29.97 (4.86)

29.86 (5.01)

0.35

   Arousal dom.

28.55 (5.74)

30.07 (6.05)

3.92**

   Behavioral dom.

27.94 (5.84)

28.35 (6.40)

1.06

Part B

61.27 (14.68)

64.91 (15.51)

3.67**

Total score

147.74 (27.54)

153.19 (27.93)

2.98*

a= 638

b= 349

*< 0.01, **< 0.001

 

Male Offenders: (Moeller et al., 2015; = 231 (= 64 male forensic patients, 92% of whom were convicted of a violent crime and = 167 male inmates, 61% of whom were convicted of a violent crime; mean age across both groups = 32.6 (11.4), age range = 18-67)

NAS/NAS-PI Scores

M

SD

Male forensic patients

NAS total

97.1

16.8

    Cognitive

34.1

5.8

    Arousal

32.2

5.8

    Behavioral

30.9

5.8

    Regulationa

26.8

4.4

Male inmates

PI total

65.4

14.0

aNAS Regulation is not included in the NAS total.

 

Incarcerated Male Offenders: (Mills et al., 1998; = 204 (= 102 General Admissions, mean age = 33 (10.5) years, age range = 19-69 years; n = 102 Violent Admissions, mean age = 28 (7.9) years, age range = 18-55 years; all offenders sentenced to ≥ 2 years)

Means and Standard Deviations on NAS by Admission Status

 

NAS Scores

General Admission 

(= 102), M(SD)

Violent Admission 

(= 102), M(SD)

Part A

80.38 (15.78)

71.37 (13.12)

    Cognitive dom.

28.93 (5.21)

25.82 (4.48)

    Arousal dom.

26.14 (5.66)

23.31 (5.10)

    Behavioral dom.

25.31 (6.12)

22.24 (4.99)

Part B

53.01 (15.77)

48.59 (12.71)

 

 

Test/Retest Reliability

Incarcerated Male Offenders: (Mills et al., 1998; = 102 for total group and = 51 for both computer and paper retests)

  • Acceptable to Excellent test-retest reliability: (ICC’s ranged from 0.78 for the computer retest of the cognitive domain to 0.91 for the paper retest of Part A, with all other correlations between 0.82 and 0.89

     

Internal Consistency

Male Forensic patients (Scotland): (Novaco, 2012, = 119)

  • Excellent for NAS Total: Cronbach’s alpha = 0.95*
    • Excellent for Cognitive, Arousal, and Behavior Subscales: All Cronbach’s alphas ≥ 0.80

 

Incarcerated Adult Offenders: (Baker et al., 2008)

  • Excellent for NAS Total Score, Part A and domains, and Part B for male offenders (α = 0.83 – 0.93*) and female offenders (α = 0.80 – 0.89) 

 

Male Offenders: (Moeller et al., 2015; = 231)

  • Excellent for male forensic patients: Cronbach’s alpha for NAS Total = 0.93
  • Excellent for male inmates: Cronbach’s alpha for PI Total = 0.92

 

Incarcerated Male Offenders: (Mills et al., 1998)

  • Excellent for General Admissions group: Cronbach’s alpha for NAS Part A = 0.95* and for NAS Part B = 0.96*
  • Excellent for Violent Admissions group: Cronbach’s alpha for NAS Part A = 0.94* and for NAS Part B = 0.95*

 

Male Hospital Patients with Developmental Disabilities: (Novaco, 2004; = 129 (121 (94%) detained under sections of the mental health act and 8 (6%) “informal” or voluntary patients, mean age of whole group = 33.2 (11.6))

  • Excellent for NAS Total (= 110): Cronbach’s alpha = 0.92*
  • Excellent for PI Total (n = 114): Cronbach’s alpha = 0.92

 

Construct Validity

Convergent validity:

Violent Adult Male Prison Inmates: (Lindqvist et al., 2005; = 95; age range = 18-67 years; inmates with violent convictions: physical assault (= 35), rape (= 15), murder (= 41) and kidnapping (= 4); Swedish translation of NAS into NAS-1998-S)

  • Excellent correlations between NAS Total and scales of anger and aggressiveness:
    • Provocation Inventory (PI-S) (= 0.69)
    • State-Trait Anger Expression Inventory-2 (STAXI-2-S) (= 0.68-0.79)
    • Aggression Questionnaire (AQ-S) (r = 0.65-0.86)
    • Swedish Universities Scale of Personality (SSP) Aggressiveness factor scales (= 0.71-0.76)
  • Excellent correlations between NAS Behavior scale and measures of expressed anger:
    • Anger Expression Out (AX-Out) subscale of STAXI-2-S (= 0.77)
    • Physical aggression scales of AQ (r = 0.87) and SSP (= 0.80)
  • Adequate correlations between NAS Cognitive and Arousal scales and measures of cognitive distortion (AQ Hostility: r = 0.50 for both)
  • Poor correlations between NAS Cognitive and Arousal scales and internal anger (STAXI-2-S AX-In scale: = 0.30 and = 0.27, respectively).
  • Excellent correlations between NAS Regulation scale and Control scales of STAXI-2-S (Anger Control-In (= 0.65) and Anger Control-Out (= 0.68)
  • Poor to Excellent negative correlations between NAS Regulation scale and PI-S, STAXI-2-S, and AQ-S measures of anger expression (r = -0.02-0.67) 

 

Face Validity

Face validity was not statistically assessed, however independent studies have confirmed the validity of the NAS with multiple populations (Moeller et al., 2015)

 

Responsiveness

Incarcerated Male Offenders: (Mills et al.; = 204)

  • “Scores on the NAS were significantly lower in the 4th week than in the 1st week. This could be due in part to a second exposure to the test instrument. However, it may also be due to a reduction in anxiety between the two testings. Kroner and Reddon (1995) and Spielberger et al. (1983) demonstrated a significant positive correlation between anger and anxiety in two different populations. Admission to a maximum security institution is undoubtedly an anxiety provoking experience, and research among this population has shown that anxiety does decline over time (Zamble & Porporino, 1988). These results suggest the need for consistent administration of the test with respect to temporal proximity to any systemic anxiety eliciting situations. Consequently, norms established in the first week after admission to an institution may be less valid if used with test results garnered several weeks later (p. 246).”

 

Bibliography

*Baker, M. T., Van Hasselt, V. B., & Sellers, A. H. (2008). Validation of the Novaco Anger Scale in an incarcerated offender population. Criminal Justice and Behavior, 35(6), 741-754. https://doi.org/10.1177/0093854808316275

Chilvers, J., & Thomas, C. (2011). Do male and female forensic patients with learning disabilities differ on subscales of the Novaco Anger Scale and Provocation Inventory (NAS‐PI)? Journal of Learning Disabilities and Offending Behaviour : Practice, Policy and Research, 2(2), 84–97. https://doi.org/10.1108/20420921111152469

Culhane, S. E., & Morera, O. F. (2010). Reliability and validity of the Novaco Anger Scale and Provocation Inventory (NAS-PI) and State- Trait Anger Expression Inventory-2 (STAXI-2) in Hispanic and non- Hispanic White student samples. Hispanic Journal of Behavioral Sciences, 32(4), 5 86–606. https://doi.org/10.1177/0739986310381458

Hornsveld, R. H., Muris, P., & Kraaimaat, F. W. (2011). The Novaco Anger Scale-Provocation Inventory (1994 version) in Dutch forensic psychiatric patients. Psychological assessment, 23(4), 937–944. https://pubmed.ncbi.nlm.nih.gov/21668125/

*Jones, J., Thomas-Peter, B., & Gangstad, B. (2003). An investigation of the factor structure of the Novaco Anger Scale. Behavioural and Cognitive Psychotherapy, 31(4), 429–437. https://doi.org/10.1017/s1352465803004041

*Lindqvist, J. K., Dåderman, A. M., & Hellström, Å. (2005). Internal reliability and construct validity of the Novaco Anger Scale-1998-S in a sample of violent prison inmates in Sweden. Psychology, Crime & Law, 11(2), 223–237. https://doi.org/10.1080/10683160500036863

*Mills, J. F., Kroner, D. G., & Forth, A. E. (1998). Novaco Anger Scale: Reliability and validity within an adult criminal sample. Assessment, 5(3), 237–248. https://doi.org/10.1177/107319119800500304

Moeller, S. B., Novaco, R. W., Heinola-Nielsen, V., & Hougaard, H. (2015). Validation of the Novaco Anger Scale–Provocation Inventory (Danish) with nonclinical, clinical, and offender samples. Assessment, 23(5), 624-636. https://doi.org/10.1177/1073191115583713

Naz, S., & Khalily, M. T. (2015). Indigenous adoption of Novaco’s Model of Anger Management among individuals with psychiatric problems in Pakistan. Journal of Religion and Health55(2), 439–447. https://doi.org/10.1007/s10943-015-0012-y

Novaco, R. W. (1994). Anger as a Risk Factor for Violence among the Mentally Disordered. In J. Monahan and Steadman, H. J. (Ed’s), Violence and Mental Disorder, Developments in Risk Assessment (pp. 21-59). Chicago & London: The University of Chicago Press.

Novaco, R. W. (2012). The Novaco Anger Scale and Provocation Inventory (NAS-PI): Manual (W-348B). Western Psychological Services.

Novaco, R. W. & Chemtob, C. M. (2002, April). Anger and combat-related post-traumatic stress disorder. Journal of Traumatic Stress, 15(2), 123-132.

Novaco, R. W., & Renwick, S. J. (1998). Anger predictors of the assaultiveness of forensic hospital patients. In E. Sanavio (Ed.), Behavior and Cognitive Therapy Today: Essays in Honor of Hans J. Eysenck (pp. 199-208). Amsterdam: Elsevier Science.

Novaco, R. W., & Taylor, J. L. (2004). Assessment of anger and aggression in male offenders with developmental disabilities. Psychological Assessment, 16(1), 42–50. https://doi.org/10.1037/1040-3590.16.1.42

Skeem, J. L., & Mulvey, E. P. (2001). Psychopathy and community violence among civil psychiatric patients: Results from the MacArthur Violence Risk Assessment Study. Journal of Consulting and Clinical Psychology, 69(3), 358-374. https://doi.org/10.1037/0022- 006X.69.3.358

WPS. (2025). Novaco Anger Scale and Provocation Inventory (NAS- PI). WPS. https://www.wpspublish.com/nas-pi-novaco-anger-scale- and-provocation-inventory

*These studies utilized a previous version of the NAS-PI and do not reflect the most up-to-date version of the assessment.