1. The original paper has to be substantive in nature, as well as clearly relevant to and well cited in the marketing literature. In this regard, we require authors to provide the following information about the original paper to demonstrate impact relative to other papers published in the same journal that year and in top marketing journals in that year using at least one of these three sources:
a) ISI citation counts to-date on the Web of Knowledge;
b) SCOPUS citation counts to date (http://www.scopus.com
), typically available through your organization's library or information department, or
c) Google Scholar citations.
2. Having had experience with the Replication Corner for over a year, we find that the most interesting replications extend the original paper in significant ways. As a consequence, a bit more space is sometimes needed to adequately represent what was done and what it means. Therefore, we have decided to increase the word limit to 2000 words.
NOTE: 2000 is a STRICT limit and not a suggested length. Authors are responsible for providing a word count (in addition to the citation count and including in the paper a table comparing their results to those of the paper being replicated).
Papers over the limit will be returned to the authors without being reviewed.
A 100-word abstract and 10 references can be included. The
replication should include a short introduction on the chosen effect and the
difference between the original paper and the replication. This should be
followed by a more detailed report of the procedure, summary of the results,
and limitations of the replications. In general, replications should be written
in a concise and clear writing style.
replication should include a clear comparison of the results, as well as any
difference in the data collection and the methods that were used between
the original paper and the replication. The authors are encouraged to
provide this comparison in a table.
tables, the data, and the code used should be submitted and will be
posted on the website of the journal, for public use.
replication corner has four editors: Eric Bradlow, Joel C. Huber, Don
Lehmann, and John Lynch, one of them will process the submission.
will consider reports of failed replications as well, although we might ask one
of the original authors to review the replication attempt. In such cases, a
careful examination of quality checks (e.g., attention) becomes an important
part of the report.
If the original
author reports a significant effect and a second author finds no
significant effect, it is always unclear whether the difference in results
is a “failure to replicate” or just what one would expect from random
draws from a common effect size. We would like to ask authors to include
in their papers or at least an online appendix a meta-analysis
including the original Study and attempted replication. An example
of such a meta-analysis in the online appendix comes from Chark and Muthukrishnan
(IJRM Dec 2013)A good short summary of how to do this is in Rosenthal and de
Matteo 2001 Annual Review of Psychology. It is easy to have a
situation where one effect size is significant and another is not,
but no significant heterogeneity across the studies. If the
heterogeneity is not significant, then one can calculate
the weighted average effect size and test whether the effect
is significant pooling across all the available studies. If there is
no significant heterogeneity but the weighted average effect size remains
significant, the original conclusion would stand. If there is no significant
heterogeneity and the weighted average effect size is NOT significant, then
this calls into question the original finding. If there is significant
heterogeneity, then this raises the question of what is the moderator or
boundary condition that explains the difference in results.
the cover letter, authors should explain why they think their replication is
important and provide evidence for it (i.e., citation estimate in the important
authors have to send along with their submitted replication the paper where the
original phenomenon is reported.
of the important considerations in reviewing the replication is how appropriate
is the use of subjects in experimental or empirical replication study.
10. Before embarking on a replication study
on an effect, authors may consult with the editors by submitting a
one-paragraph proposal for a pre-study evaluation.