We tackle the important problem of understanding the accuracy of activity tracker data. To do this, we introduce the notions of daily and hourly adherence, key aspects of how consistently people wear trackers. We hypothesise that these measures provide a valuable means to address accuracy problems in population level activity tracking data. To test this, we conducted a semester-long study of 237 University students: 88 Information Technology, 149 Medical Science. We illustrate how our adherence measures provide new ways to interpret data and valuable insights that take account of tracker data accuracy. Finally, we discuss broader roles for daily and hourly adherence measures in activity tracker data.