Testing the Validity of Campaign Ad Exposure Measures: A Family Planning Media Campaign in Pakistan
Beaudoin, C.E.; Stephenson, M.T.; Agha, S.
Journal of Health Communication 21(7): 773-781
2016
ISSN/ISBN: 1081-0730 PMID: 27337154 DOI: 10.1080/10810730.2016.1157655Document Number: 373665
Although prior research has tested the nomological validity of media campaign exposure, including the related comparative validity of some measures, it has not well studied predictive validity or made extensions to other types of media campaign exposure. To help build on research in this area, the current study tested the nomological and predictive validity of 5 ad recall and recognition measures specific to the Touch condom media campaign in Pakistan. Between-effects regression of panel survey data confirmed the nomological validity of each of the 5 measures of Touch ad exposure. In addition, 2 sets of panel regression models (i.e., fixed-effects models and fixed-effects with lag models) confirmed the predictive validity of each of the 5 ad exposure measures. Results on comparative validity were quite similar for nomological and predictive validity, indicating that confirmed ad recall and recognition measures tend to have greater validity than unconfirmed measures.