Table 2.

Variables used in the Physical Activity at Work process evaluation.

Variable				Explanation		Scale
Associations between movement break participation and daily Fitbit sedentary time and step count data during the intervention period
	Outcome variables
		Sedentary time	Daily sedentary time data from Fitbit		Continuous (min)
		Steps	Daily step count data from Fitbit		Continuous (step count)
	Exposure variables
		Movement break participation	Daily movement break participation		Ordinal; 1-4 (sessions)
Effects of intervention components on weekly movement break participation percentage
	Outcome variable
		Movement break participation percentage	The frequency of an individual’s participation in movement break sessions within a specific week compared to the maximum number of sessions available during that week		Continuous (percentage)
	Exposure variables
		Individual reward	We designated the weeks following individual reward wins as {1}, while all other weeks were coded as {0}, using the web monitoring data		1=won a reward last week; 0=others
		Cluster reward	We assigned {1} to the weeks following office colleagues’ reward wins and {0} to all other weeks, using the web monitoring data		1=a colleague in the same office cluster won a reward last week; 0=others
		Fitbit wear time	We categorized the average weekly Fitbit wear time for each individual into a binary variable using the median		1=wore Fitbit at or above the median wear time; 0=wore Fitbit below the median wear time
		Fitbit wear time (alternative)	How often Fitbit led individuals to engage in physical activity in the office in the last 2 weeks of the intervention		1=at least 4 d/wk; 0=<4 d/wk
		Leadership support	How often the directors’ support led individuals to engage in physical activity in the office in the last 2 weeks of the intervention		1=at least 4 d/wk; 0=<4 d/wk
		Movement break leaders	How the movement break leaders’ (1) encouragement and (2) enthusiasm contributed to movement break participation (We asked the movement break leaders and used the data as cluster-representative data)		3=both factors contributed somewhat or a lot; 2=only the encouragement contributed somewhat or a lot; 1=only the enthusiasm contributed somewhat or a lot; 0=both factors contributed not at all, very little, or a little
		Posters	How often the posters led individuals to engage in physical activity in the office in the last 2 weeks of the intervention		1=at least 4 d/wk; 0=<4 d/wk
		Posters (alternative)	How accurate individuals were in reporting the number of different poster designs in their office		1=3-4 designs; 0=0-2 designs