This repository is available for researchers to provide details about the data and methods used in their publications so as to encourage others to extend their research. We also welcome comparative ads that might be used to provide further testing and development of advertising principles.
Print ads presented in Which Ad Pulled Best (WAPB) ads are shown here. These ads were provided thanks to Scott Purvis , Sandeep Patniak, and Gallup and Robinson.
Analyses of Which Ad Pulled Best (WAPB) ads
Analyses of evidence on each of the principles were reported in the original publication of Armstrong, J. Scott (2010) Persuasive Advertising in June 2010 and in Armstrong, J. Scott (2011), "Evidence-based Advertising: An Application to Persuasion".
The following link redirects to files that include the codings for the variables tested in the WAPB analyses along with their recall scores. The findings from these analyses were reported in Armstrong, J. Scott (2010), Persuasive Advertising, and in J. Scott Armstrong and Sandeep Patnaik (2009), "Using Quasi-experimental Data to Develop Principles for Persuasive Advertising", Journal of Advertising Research, 49, No.2, 170-175.
The password-protected spreadsheet provides full-disclosure for researchers who might want to replicate or extend the study. Researchers who wish to conduct research can obtain the password from J. Scott Armstrong or Sandeep Patnaik. They must provide a valid reason for use of these data and must sign a non-disclosure statement.
This spreadsheet summarizes the codings for the pairs of ads in various volumes of the WAPB series. The codings are organized by principle number, which can be accessed using the tabs at the bottom of the spreadsheet. These were 69 analyses that were used to test 58 of the principles in Persuasive Advertising. Details of the procedure were presented in Appendix B. of Persuasive Advertising. Dr. Sandeep Patnaik conducted this research.
Methods and Materials for "Testing Predictive Validity of Evidence-based Advertising"
These five batches of WAPB (Which Ads Pulled Best) print ads were used for the tests reported in our paper, “Predictive Validity of Evidence-based Advertising.” They were selected from a systematic sampling of all “final consumer” ads coded as being for high-involvement utilitarian products in Editions 5 through 9. All batches have 20 pairs, except for Batch 3 with 18. The ads are intended to provide full disclosure for our paper, but they do not include the answers. Those are only available from Scott Armstrong or Kesten Green and then only for research purposes. Requests must be on your university’s letterhead. A fee will be charged; it will be used to cover the cost of developing and maintaining these files.
Software used for the AdPrin Audits
The API scores for Batches 2 through 5 were based on Single Rater Rating 2 Ads Once. These ratings were done in 2011 and up through June 2012. This self-training module was used and the ad pair used for the training module was number 7.45 from Which Ad Pulled Best.
Unaided Judgments Used as Benchmarks
Experst and non-experts were asked to make judgments about the effectiveness of each of the WAPB ads. The same batches were used. We used the following survey materials:
The Advertising Principles site summarizes all useful knowledge about advertising so that it can be used by researchers, practitioners, and educators.