Illustrative dataset for the article: Vieira, R., McDonald, S., Araujo-Soares, V., Sniehotta, F., Henderson, R. (2017) "Dynamic modelling of n-of-1 data: Powerful and flexible data analytics applied to individualised studies"

This dataset is supplementary material of the manuscript "Dynamic modelling of n-of-1 data: Powerful and flexible data analytics applied to individualised studies. McDonald et al. (2016) presents a series of novel n-of-1 studies that intended to explore the relationship between physical activity change during the retirement transition. The file contains the data of one participant. The column names correspond to the following variables:

time: duration of follow-up (minutes);
minute: time of day (hours and minutes);
day_num: day since beginning of follow-up (the first two days were considered as adaptation phase and therefore removed); 
PAscore: accelerometer raw score; 
startBout: 1 (a bout of PA was initiated in this minute) or 0 (a bout of PA wasn't 
initiated in this minute); 
nPAbouts_day: number of PA bouts per day; 
nPAbouts_day.l1: number of PA bouts in previous day (lag 1); 
nPAbouts_day.l2: number of PA bouts two day before (lag 2); 
nBoutsLast2hours: number of PA bouts in previous 2 hours; 
retirement: 0 (before retirement) or 1 (after retirement)
weekday: 0 (workday) or 1 (weekend)
sleepLength: number of hours of sleep last night
sleepLength.l1: number of hours of sleep the night before
sleepLength.l2: number of hours of sleep two nights before
pers: personalised measure of partner's influence (scale 0-1)
periodDay: morning, evening or afternoon

McDonald, S., Vieira, R., O'Brien, N., White, M., & Sniehotta, F. F. (2016). Does physical activity and sedentary behavior change during the retirement transition? Findings from a series of novel n-of-1 natural experiments. International Journal of Behavioral Medicine, 23, S261-S261.