Background:
Low self-efficacy is considered physical activity recommendations. However, tools used to measure self-efficacy currently used in research and clinical practice haven’t been tested for validity and reliability in the stroke population.
Aims:
To evaluate the test-retest reliability and construct validity of three self-efficacy scales for use with stroke survivors.
Methods:
A repeated-measures reliability and construct validity study was conducted. Fifty-one community-dwelling stroke survivors (mean age = 74 years, median steps/day = 4,664) completed three self-efficacy scales on two occasions, seven days apart (Self-Efficacy for Exercise scale (SEE), Spinal Cord Injury Exercise Self-Efficacy Scale (SCI-ESES) and community management domain of the Participation Strategies Self-Efficacy Scale (PS-SES)). Construct validity was assessed using eight hypotheses determined a-priori, evaluating community participation, physical activity, comorbidities and physical function via t-tests and Pearson’s correlation coefficients. Test-retest reliability was analysed using intraclass correlation coefficients.
Results:
Test-retest reliability was established for the SEE (ICC=0.8) , PS-SES (ICC=0.8) and the SCI-ESES (ICC=0.7). The SEE and SCI-ESES achieved 3/8 hypotheses for construct validity, while the PS-SES achieved 2/8. The three measures were highly correlated with each other and with self-reported levels of physical activity, but failed to achieve target levels (r≥0.5) for association with objective physical activity levels, comorbidity or function.
Conclusion:
All three measures (SEE, PS-SES and SCI-ESES) showed high test-retest reliability. However, it appears that construct validity is insufficient for clinical or research settings. Measurement of a complex phenomenon such as self-efficacy poses a challenge for the stroke research and clinical communities. Development of valid and reliable tools to better understand and measure post stroke self-efficacy for physical activity appears warranted.