Mechano-content analysis as an aid to predicting effects of advertising microcommunications

Abstract

Performed a regression analysis of 204 posters to predict the percentage of respondents who remembered seeing an outdoor advertising message. The independent variables were 21 mechanical and 6 content elements; the dependent variables were recognition scores. The predictive validity of 20 equations was too low to be of value to advertising communicators. The correlations between predicted and actual scores were low (.30-.41, p < .05) because the independent variables varied only slightly as a result of the outdoor medium's simplicity. (PsycInfo Database Record (c) 2025 APA, all rights reserved)