Reliability of a brief intercept survey for trail use behaviors

Purpose: This study assessed test-retest reliability of an interviewer-administered trail survey. Methods: An intercept survey was conducted with adults using 2 paved trails in Indiana and South Carolina (N = 295; mean age = 46.9 ± 18 y). The survey included items on frequency and duration of trail use for recreation and transportation, other patterns of trail use, and sociodemographic characteristics. Fiftyfive adults completed the survey twice (2-16 d apart; mean = 7.4 ± 2.6 d). Test-retest reliability was assessed with Spearman rank correlation coefficients, Kappa coefficients, and percent agreement. Results: Kappa coefficients and percent agreement for 9 categorical items ranged from 0.65 to 0.96 and from 64.0% to 98.2%, respectively. Among these items, the lowest Kappas were found for perceived safety (0.65) and reported duration of visits for recreational purposes (0.67). Spearman rank correlation coefficients for travel distance to and on the trail and frequency of trail use during the past 7 days and past 4 weeks ranged from 0.62 to 0.93. Conclusion: Though further assessments of this survey with different populations and types of trails may be warranted, its overall high reliability indicates it can be used by researchers and practitioners in its current form.