ors in non-clinical sample

1
RESULTS Mean ORS total score = 30.53 (SD = 5.83) and 29.24 (SD = 5.44) for the 1 st and 2 nd administrations, with no significant differences between males and females (t = -.51, p > .05). Cronbach’s coefficient alpha for ORS = .83 and .80. Test-retest reliability coefficient for total ORS was r = .53. Exploratory factor analysis revealed one factor (68% of variance) ORS total scores were moderately correlated with self-reported quality of life (QLS scores; r = .61, r 2 = .37; r = .66, r 2 = .44), indicating convergent validity. ORS total scores were relatively unrelated to physical activity (IPAQ scores; r = .13, r 2 = .02; r = .03, r 2 = .00) indicating discriminant validity. INTRODUCTION A growing body of research supports the utility of monitoring and using client feedback about treatment progress (Anker, Duncan, Sparks, 2009; Lambert, 2010). One increasingly popular method of soliciting client feedback is the Outcome Rating Scale (ORS), an ultra brief (4-item) measure of client functioning (Miller & Duncan, 2000). The ORS assesses psychological functioning along four dimensions: symptom distress, interpersonal relationships, social role functioning, and overall well-being. The ORS offers the reliability and validity of longer measures, but with more feasibility for use in clinical settings (Miller, Duncan, Brown, Sparks, & Claud, 2003). When used in conjunction with the Session Rating Scale, an ultra brief alliance measure, the ORS has been shown to improve therapy outcomes (Anker, et al., 2009; Miller, Duncan, Brown, Sorrell, & Chalk, 2006; Reese, Norsworthy, Rowlands, 2009). Only one study to date has investigated the reliability and validity of the ORS among a non-clinical sample (Bringhurst, Watson, Miller, and Duncan, 2006). The purpose of the present study was to investigate five psychometric characteristics of the ORS: internal consistency, test-retest reliability, factor structure, convergent validity, and discriminant validity. METHODS Participants 211 participants (50% female) Recruited from a university fitness center as part of a study concerning exercise program adherence. Mean age = 21 (SD = 6.57) 38 participants returned for follow-up at 4-weeks Measures Outcome Rating Scale (Miller & Duncan, 2000) Quality of Life Scale (QOLS; Burckhardt & Anderson, 2003) International Physical Activity DISCUSSION ORS scores for this study were comparable to previous research with a non-treatment seeking college students (Bringhurst, et. al, 2006). Internal consistency was in the good range, though below previous estimates (.97, Bringhurst, et. al., 2006; .93, Miller, et. al., 2003). ORS scores demonstrated good convergent validity with overall quality of life and discriminate validity with physical activity. Factor analysis supported a one-factor model of the ORS, as expected for a scale with only four items. Limitations: A non-clinical, student sample limits generalizability. 18% of participants returned for the follow-up, weakening test-retest estimates. Use of a more clinically relevant psychological functioning scale would have increased the Psychometric Properties of the Outcome Rating Scale in a Non-Clinical Sample Danielle Smith, B.S., Laura Crocker, B.S., Cathy Staton, M.A., Art Gillaspy, Ph.D., and Shawn Charlton, Ph.D. University of Central Arkansas Means, Standard Deviations, and Reliability Coefficients for Study Measures for Administration 1 ORS Items Factor Pattern Coefficients Symptom Distress .69 Interpersonal Relationships .62 Social Role Functioning .76 Overall Well-Being .98 Scale Mean Standard Deviation Cronbac h’s Alpha ORS 30.54 5.83 .83 QOLS 86.72 10.92 .85 IPAQ 40.27 22.23 .84 Factor Pattern Coefficients for ORS Items Bivariate Correlations between ORS, QOLS, and IPAQ for Administration 1 ORS Items ORS Total QOLS IPAQ Symptom Distress .76 .45 .17 Interpersonal Relationships .78 .50 .01 Social Role Functioning .85 .53 .15 Overall Well-Being .91 .53 .14 ORS Total .61 .14 Reliability Coefficients for ORS Test-Retest Estimates Individual Functioning .41 Interpersonal Relationships .65 Social Role Functioning .53 Overall Well-Being .40 ORS Total .53

Upload: barry-duncan

Post on 16-Nov-2014

5.824 views

Category:

Documents


0 download

DESCRIPTION

Smith, D., Crocker, L., Staton, C., Gillaspy, A., & Charlton, S. (2010). Psychometric properties of the Outcome Rating Scale in a non-clinical sample. Poster presentation at the Heart and Soul of Change Conference, New Orleans.

TRANSCRIPT

Page 1: ORS in non-clinical sample

RESULTS Mean ORS total score = 30.53 (SD = 5.83) and 29.24 (SD =

5.44) for the 1st and 2nd administrations, with no significant differences between males and females (t = -.51, p > .05).

Cronbach’s coefficient alpha for ORS = .83 and .80. Test-retest reliability coefficient for total ORS was r = .53. Exploratory factor analysis revealed one factor (68% of

variance) ORS total scores were moderately correlated with self-

reported quality of life (QLS scores; r = .61, r2 = .37; r = .66, r2 = .44), indicating convergent validity.

ORS total scores were relatively unrelated to physical activity (IPAQ scores; r = .13, r2 = .02; r = .03, r2 = .00) indicating discriminant validity.

INTRODUCTIONA growing body of research supports the utility of monitoring and using client feedback about treatment progress (Anker, Duncan, Sparks, 2009; Lambert, 2010). One increasingly popular method of soliciting client feedback is the Outcome Rating Scale (ORS), an ultra brief (4-item) measure of client functioning (Miller & Duncan, 2000). The ORS assesses psychological functioning along four dimensions: symptom distress, interpersonal relationships, social role functioning, and overall well-being. The ORS offers the reliability and validity of longer measures, but with more feasibility for use in clinical settings (Miller, Duncan, Brown, Sparks, & Claud, 2003). When used in conjunction with the Session Rating Scale, an ultra brief alliance measure, the ORS has been shown to improve therapy outcomes (Anker, et al., 2009; Miller, Duncan, Brown, Sorrell, & Chalk, 2006; Reese, Norsworthy, Rowlands, 2009). Only one study to date has investigated the reliability and validity of the ORS among a non-clinical sample (Bringhurst, Watson, Miller, and Duncan, 2006). The purpose of the present study was to investigate five psychometric characteristics of the ORS: internal consistency, test-retest reliability, factor structure, convergent validity, and discriminant validity.

METHODS Participants

211 participants (50% female) Recruited from a university fitness center as part of a

study concerning exercise program adherence. Mean age = 21 (SD = 6.57) 38 participants returned for follow-up at 4-weeks

Measures Outcome Rating Scale (Miller & Duncan, 2000) Quality of Life Scale (QOLS; Burckhardt & Anderson,

2003) International Physical Activity Questionnaire (IPAQ;

Booth, 2000)

DISCUSSION ORS scores for this study were comparable to previous

research with a non-treatment seeking college students (Bringhurst, et. al, 2006).

Internal consistency was in the good range, though below previous estimates (.97, Bringhurst, et. al., 2006; .93, Miller, et. al., 2003).

ORS scores demonstrated good convergent validity with overall quality of life and discriminate validity with physical activity.

Factor analysis supported a one-factor model of the ORS, as expected for a scale with only four items.

Limitations:A non-clinical, student sample limits generalizability.18% of participants returned for the follow-up, weakening test-retest estimates. Use of a more clinically relevant psychological functioning scale would have increased the significance of the concurrent validity estimates.

Suggestions for further research include psychometric studies that directly address or correct these limitations.

Psychometric Properties of the Outcome Rating Scale in a Non-Clinical Sample

Danielle Smith, B.S., Laura Crocker, B.S., Cathy Staton, M.A., Art Gillaspy, Ph.D., and Shawn Charlton, Ph.D.

University of Central Arkansas

Means, Standard Deviations, and Reliability Coefficients for Study Measures for Administration 1

ORS Items Factor Pattern Coefficients

Symptom Distress .69

Interpersonal Relationships .62

Social Role Functioning .76

Overall Well-Being .98

Scale MeanStandard Deviation

Cronbach’s Alpha

ORS 30.54 5.83 .83QOLS 86.72 10.92 .85IPAQ 40.27 22.23 .84

Factor Pattern Coefficients for ORS Items

Bivariate Correlations between ORS, QOLS, and IPAQ for Administration 1

ORS Items ORS Total QOLS IPAQ

Symptom Distress .76 .45 .17

Interpersonal Relationships .78 .50 .01

Social Role Functioning .85 .53 .15

Overall Well-Being .91 .53 .14

ORS Total .61 .14

Reliability Coefficients for ORS

Test-Retest Estimates

Individual Functioning .41

Interpersonal Relationships .65

Social Role Functioning .53

Overall Well-Being .40

ORS Total .53