10 minutes before the initiation of the report. Further, we identified participants who were noncompliant with other aspects of the study. It might be expected that these participants would also be noncompliant with the drink reports. Poor compliance was operationally defined as (a) completing fewer than 50% of audibly prompted assessments per week; (b) using the ED sleep function, which allowed participants to turn the device off while sleeping, for 13 or more hours on 4 or more days in a week; and (c) suspending audible prompting for more than 14 hours in a week. If any of these criteria were met, the participant’s data were not used in the analyses. In the entire sample, compliance with the protocol was very high (e.g., approximately 80% compliance with audible prompts; Tidey et al., 2008).
