Measuring reading time: Comparing logged and self-reported data in relation to reading skills

BACKGROUND: Children's reading time at home plays a critical role in their reading development. However, existing measures of reading time, based on self-reports, are often biased. Logged data from mobile apps may offer a more reliable alternative, as shown in studies examining screen time in digital media use. OBJECTIVES: This study compared logged and self-reported measures of reading time and examined their associations with reading skills in French primary school children. METHOD: One hundred and nine children from Grade 1 to Grade 5 and their parents participated. Parents completed a retrospective questionnaire estimating weekly reading time (self-reported measure). They then used a mobile application to record their child's reading activities in real time over a 14-day period (logged measure). All children were assessed on their reading fluency. RESULTS AND CONCLUSIONS: The self-reported measure yielded significantly higher reading time estimates (M = 6.26 hours/week) than the logged measure (M = 2.11 hours/week), with a moderate correlation between the two (r = .45). Crucially, the logged measure showed stronger predictive validity for reading fluency (r = .39) than the self-reported measure (r = .25). Regression analyses confirmed that when both measures were included simultaneously, only the logged reading time remained a significant predictor of reading performance. These findings suggest that logged measures obtained via ambulatory assessment (here, using a mobile app) provide more accurate estimates of reading time and superior predictive validity compared to traditional self-reports. This methodology offers promising avenues for future research on reading habits and literacy development.