A comparison of cross-sectional and longitudinal data in scale validation: the case of destination brand experience

Abstract

Despite the benefits of longitudinal data, most scale development research has employed cross-sectional data to validate psychometric properties of measurement. This study formally evaluates the relative strengths of cross-sectional versus longitudinal research in scale validation based on common method variance, common method bias, construct validity and reliability, and causal inferences. Our results indicate that longitudinal data exhibit diverse advantages over cross-sectional data in terms of identifying common method variance, reducing common method bias, demonstrating construct validity, and enhancing causal inferences. While longitudinal data can clearly differentiate predictor and outcome variables (i.e. via temporal ordering), cross-sectional and longitudinal data show similar evidence of coherence and covariation. These findings shed light on using longitudinal and cross-sectional data for scale validation as an important step in scale development. More importantly, we demonstrate that longitudinal data offer a valuable alternative to cross-sectional data by generating valid and reliable measures in scale validation process.