Abstract
This paper details an experiment to migrate a long-established survey from a face-to-face to an online methodology. The survey - Ofcom's Media Tracker - has been running for more than ten years and has generated a longitudinal dataset of great value for assessing trends over time. The value of this dataset needs to be protected against any discontinuity caused by methodological change. A novel technique was developed to determine which variables in addition to demographics should be used to reweight the data from an offline survey to best replicate what would have been achieved had the traditional data collection method continued. The results helped Ofcom to make a decision about migration for this particular survey and, more generally, provide a useful addendum to existing knowledge regarding successful modal migration.
Get full access to this article
View all access options for this article.
