Megan McMahon 2021
Study information: Preregistration plan in progress on OSF
Actigraphy data collected for 10-14 days for each participant. Some participants had their actigraphy periods split into multiple periods since they were unable to come into lab for their initial Session 2 visit due to bad weather.
Actigraphy data was preprocessed using the preproc function in the wearables package on GitHub. This involved truncating the data to include just 7 days in order to retain more subjects, interpolating up to 10 min periods of missing data, and excluding subjects who had less than 7 days of data or > 10% missing data.
Rest-activity rhythm measures were calculated using an in-house R script based on the logistic-transform cosinor method (Marler et al., 2006)
fMRI data was preprocessed using fMRIprep.
Subjects with mean FD > 0.50 mm or with > 20% of TRs with mean FD > 0.50 mm will be excluded from analyses.
Based on Sherman et al. (2015), which reported the beta weight for memory ~ CR stability was 0.20. Converting this to r using psychometria, this is equivalent to an r value of 0.25. r2 is then 0.0625. This value is used to obtain Cohen's f2 = 0.06 for G*Power.
Total number of predictors: age group, amplitude, acrophase, DMN modularity, DMN participation coefficient, FPN modularity, FPN participation coefficient, sleep efficiency, physical activity, response time or lapses, age (11) Number of tested predictors: 8 Corrected alpha level: 0.05/4 = 0.0125