What Works, What Doesn’t (and When)

(Target year for publication 2023)

Editor: Dilip Soman (University of Toronto)

CALL FOR CHAPTER PROPOSALS

1. About

This book will be published by the University of Toronto Press and is sponsored by the Behaviourally Informed Partnership project (biorgpartnership.com) at the BEAR centre, University of Toronto.

2. Overview

In a forthcoming edited book Behavioral Science in the Wild (Mazar and Soman, 2022, in press), we outline a framework for why well-documented results of behavioral interventions both in the field and in the lab do not necessarily translate and scale in all situations. We further propose that these differences in results occur because of a) changes in situation across the original demonstration and the new context, and b) differences in the population that is receiving the intervention.

In this proposed edited book, our goal is to develop a series of case studies and examples to illustrate and expand on this framework. In particular, we are hoping to generate a collection of interventional studies that each attempt to replicate, translate or scale a previously demonstrated intervention in a field setting. Each chapter will be based on one study (or more, if the studies are fairly similar). The format and style for each individual chapter is in Section 4 below.

3. call for chapters: who and what

Who should submit chapter proposals? You are either an academic or a practitioner (or in a team of academics and practitioners) that has recreated a previously demonstrated intervention in the field with the goal of creating behavior change, and have either been successful or not. Such studies might not always be a great fit for academic journals that might be more interested in theoretical, surprising and previously non-demonstrated findings.

What should chapters have? We seek chapters based on a field experiment that fall under any of the following categories:

1)    Direct translations (a field version of a previous lab experiment)

2)    Replications (using the same intervention from a previous demonstration in the same context and on roughly a similar sample of participants)

3)    Horizontal scaling (using a previously developed intervention, but applying it in a different domain)

4)    Vertical scaling (running a larger version of a beta/pilot experiment)

Both positive (i.e., the intervention had the expected result) and negative (i.e., the intervention “did not work” or backfired) results are welcome, as are interventions that worked but with weaker effects. Of course, negative results because of an error in implementing the intervention (e.g., a failure of randomization, or methodological confounds) will not be considered.

The table below captures the spectrum of possible contributions:

4. Audience, content and style of chapters

The audience for this book are practitioners interested in applying behavioral science to behavior change problems in society or business, and behavioral scientists interested in identifying conditions under which interventions translate and scale, and when they do not. Contributions should be written in a clear, concise manner that is easily accessible to a practitioner who might not be intricately familiar with the breadth or the nuances of the academic literature.

Each chapter will include a short report of the study, and a narrative explaining why the intervention worked/did not work and what was it about the context of your study that might shift results. Authors will also have the option of submitting additional details of their studies to be included in an online appendix.

The chapter should include the following information:

Motivation and Objective: Why the behavior change is needed, and what the specific objective of the intervention is. In particular, are you doing a replication, translation, or (horizontal or vertical) scaling of the original intervention?

Inspiration: What paper, report or study inspired your own? What was their key idea and what behavioral theory or phenomenon was it based on? What was their design, setting and participants? What were their outcome variables and the key results?

Differences: In what way was your study similar to or different from the original? Dimensions on which similarity should be reported include a) experimental design, b) setting, c) participants, d) sample size, e) measures, f) procedures and g) implementation details. Similarity and differences are best reported in a table. 

Interventions: What did the interventions in your study look like?

Results: What did you find? Were your results similar to or different from the original?

Reconciliation: If your results were different, why do you believe they were different? Were there any “supposedly irrelevant factors” (SIFs) that might have played a role? (any data or speculations are welcome). If similar, was there something you did to increase the likelihood that the intervention would “work”? Under what conditions would you have expected the results to NOT replicate?

Discussion and Prescriptive Advice: Based on your research, what have you learnt from the successful / failed translation or scaling effort? What would you have done differently? What prescriptive advice would you offer to a practitioner or a scientist looking to the literature to create positive behavior change?

Each chapter will be no longer than 3500 words (roughly 14 or 15 double-spaced pages), plus references. In addition, authors can also prepare a web appendix with additional materials that will be posted.

5. Process AND TENTATIVE TIMELINE

Questions? Please email rotmanbiorg@rotman.utoronto.ca and include “What Works Book” in the subject line.