BACKGROUND: Ecological momentary assessment (EMA) is a method of collecting behavioral data in real time. The purpose of this study was to examine EMA compliance, identify factors predicting compliance, assess criterion validity of, and reactivity to, using EMA in a workplace intervention study. METHODS: Forty-five adults (91.1% female, 39.7 [9.6] y) were recruited for a workplace standing desk intervention. Participants received 5 surveys each day for 5 workdays via smartphone application. EMA items assessed current position (sitting/standing/stepping). EMA responses were time matched to objectively measured time in each position before and after each prompt. Multilevel logistic regression models estimated factors influencing EMA response. Cohen kappa measured interrater agreement between EMA-reported and device-measured position. Reactivity was assessed by comparing objectively measured sitting/standing/stepping in the 15 minutes before and after each EMA prompt using multilevel repeated-measures models. RESULTS: Participants answered 81.4% of EMA prompts. Differences in compliance differed by position. There was substantial agreement between EMA-reported and device-measured position (κ = .713; P < .001). Following the EMA prompt, participants sat 0.87 minutes more than before the prompt (P < .01). CONCLUSION: The use of EMA is a valid assessment of position when used in an intervention to reduce occupational sitting and did not appear to disrupt sitting in favor of the targeted outcome.
BACKGROUND: Ecological momentary assessment (EMA) is a method of collecting behavioral data in real time. The purpose of this study was to examine EMA compliance, identify factors predicting compliance, assess criterion validity of, and reactivity to, using EMA in a workplace intervention study. METHODS: Forty-five adults (91.1% female, 39.7 [9.6] y) were recruited for a workplace standing desk intervention. Participants received 5 surveys each day for 5 workdays via smartphone application. EMA items assessed current position (sitting/standing/stepping). EMA responses were time matched to objectively measured time in each position before and after each prompt. Multilevel logistic regression models estimated factors influencing EMA response. Cohen kappa measured interrater agreement between EMA-reported and device-measured position. Reactivity was assessed by comparing objectively measured sitting/standing/stepping in the 15 minutes before and after each EMA prompt using multilevel repeated-measures models. RESULTS:Participants answered 81.4% of EMA prompts. Differences in compliance differed by position. There was substantial agreement between EMA-reported and device-measured position (κ = .713; P < .001). Following the EMA prompt, participants sat 0.87 minutes more than before the prompt (P < .01). CONCLUSION: The use of EMA is a valid assessment of position when used in an intervention to reduce occupational sitting and did not appear to disrupt sitting in favor of the targeted outcome.