measurement error could only explain the associations if there was a systematic bias, such that children who were destined to be unhealthy adults (and their parents) overestimated viewing times, whereas those destined to be healthy underestimated their viewing