Psychometric properties of the Perceived Stress Scale (PSS): measurement invariance between athletes and non-athletes and construct validity

Background. Although Perceived Stress Scale (PSS, Cohen, rack & Mer 1983) has been validated arid widely used in many domains, there is still no validation in sports by comparing athletes and non-athletes and examining related psythometric indices. Purpose. The purpose of this study was to examine the measurement invariance of PSS between athletes and non-athletes, and examine construct validity and reliability in the sports contexts. Methods. Study 1 sampled 359 college student-athletes (males = 233; females = 126) and 242 non-athletes (males = 124; females = 118) and examined factorial structure, measurement invariance and internal consistency. Study 2 sampled 196 student-athletes (males = 139 females = 57, M-age = 19.88 yrs, SD = 1.35) and examined discriminant validity and Convergent validity of PSS. Study 3 sampled 37 student-athletes to assess test-retest reliability of PSS. Results. Results found that 2-factor PSS-10 fitted the model the best and had appropriate reliability. Also, there was a measurement invariance between athletes and non-athletes; and PSS positively correlated with athletic burnout and life stress but negatively correlated with coping efficacy provided evidence of discriminant validity and convergent validity. Further, the test-retest reliability for PSS subscales was significant (r = .66 and r = .50). Discussion. It is suggested that 2-factor PSS-10 can be a useful tool in assessing perceived stress either in sports or non-sports settings. We suggest future study may use 2-factor PSS-10 in examining the effects of stress on the athletic injury, burnout, and psychiatry disorders.