Judged originals (inaccurately) labeled as revisions to be superior to revisions (inaccurately) labeled as originals, or when revisions were trivial, incidental, non-existent, & even objectively worse than the original

Garcia-Rada, Ximena and John, Leslie K. and O'Brien, Ed and Norton, Michael I., The Revision Bias (February 4, 2019). Harvard Business School NOM Unit Working Paper No. 19-087. http://dx.doi.org/10.2139/ssrn.3329340

Abstract: Things change. Things also get changed—often. Why? The obvious reason is that revising things makes them better. In the current research, we document a less obvious reason: Revising things makes people think they are better, absent objective improvement. We refer to this phenomenon as the revision bias. Nine studies document this effect and provide insight into its psychological underpinnings. In Study 1, MBA students perceived their revised resumes to be of higher quality the more they differed from their original versions, but this perception was not justified: observers judged originals (inaccurately) labeled as revisions to be superior to revisions (inaccurately) labeled as originals. Study 2 pinpoints the direction of the effect: revisions are appealing, as opposed to originals being unappealing. Moreover, the revision bias holds in a variety of settings in which the revision is devoid of objective improvement: when revisions are trivial (Study 3A), incidental (Study 3B), non-existent (Study 3C), and even objectively worse than the original (Study 3D). Study 4 directly tests the self-fulfilling nature of the revision bias, testing whether mere revision framing leads people to become less critical of the experience—in this study, less sensitive to possible bugs while playing an otherwise identical “revised” video game—and whether this mediates the effect of revision framing on positive evaluations. Studies 5A and 5B offer further support by testing whether the revision bias is accentuated when people engage in a holistic processing style, whether measured as an individual difference (Study 5A) or experimentally induced (Study 5B).

Keywords: change, heuristics and biases, framing, sequences, judgment