Replication Data for: Automated Versus Human Agents: A Meta-Analysis of Customer Responses to Robots, Chatbots, and Algorithms and Their Contingencies

This dataset is used for our meta-analysis. It contains 1,167 effect sizes from 327 studies. Of these, 943 are effect sizes between AAs versus HAs and customer responses; another 224 are effect sizes between customer responses. For the meta-analysis, we used the 943 effects between AAs versus HAs and customer responses. For each effect size, the dataset includes a comprehensive set of information about the studies (e.g., study name, time, sample size, etc.), the respective independent and dependent variable, contingency variables, and additionally calculated variables, such as Fisher’s z-transformed effect sizes, which are needed for the meta-analysis. It also contains information on whether a study refers to robot, chatbot, or algorithm. We appreciate JM’s data transparency. We have uploaded all data, codes, and information needed to replicate the results presented in the tables and figures of the submitted manuscript. We propose an alternative disclosure plan as the reasons provided in item 5 of Appendix 2 apply. Specifically, we propose a delay in the release of data and codes of two years. We have invested considerable time and effort to retrieve the data for our meta-analysis, with 1,167 effect sizes retrieved from 327 studies and all variables being coded manually. We would like to harvest our extensive investment from building the database.