Wednesday, October 3, 2018

A Shout-Out for The Replication Network

In May 2015 I posted about the newly-formed The Replication Network (TRN). Since then, their team has been extremely busy promoting and fostering their objectives to serve "...... as a channel of communication to (i) update scholars about the state of replications in economics, and (ii) establish a network for the sharing  of information and ideas." TRN's "..... goal is to encourage economists and their journals to publish replications."

And they're doing a great job!

As a member of TRN I receive email newsletters from them regularly. I thought I'd share the one that I received this morning, in the hope that it might encourage some of you to become TRN members.

Here it is:

"Dear Member of The Replication Network (TRN),
If you have been checking into TRN over the last several months, then you know that a lot has been happening in the replication world. If you haven't, this email is a good chance to get you caught up with some of the highlights.
Since the last update, a major, new replication project of social science experiments was published in the journal Nature. It received headline coverage in multiple leading news outlets. You can read about it HERE. In other news, ABEL BRODEUR of Star Wars is back again with another study providing evidence of QRPs (Questionable Research Practices) in economics journals. SHRAVAN VASISHTH gives a very clear illustration of the problem of Type M error. It provides a compelling answer to the question, are significant effects from small samples more believable than significant effects from large samples? 
Recently, the subject of power, or lack thereof, has received much attention. The problem is, calculating statistical power is difficult in the best of circumstances, and trying to work backwards from an estimated regression to determine its power is generally bad practice. However, in a major contribution, BROWN, LAMBERT, and WOJAN develop a simulation procedure for calculating ex-post power curves. This is especially useful when it comes to interpreting null effects. And for those of us who have placed much hope in the benefits of replication, NATHAN GOODMAN provides some discouraging words. Sorry!
To get a flavor of recent coverage of replication in the news, check out these stories in MOTHER JONESBLOOMBERGTHE CHRONICLE OF HIGHER EDUCATION, and THE GUARDIAN.
Oh, and one more thing: If you happen to be travelling on the South Island of New Zealand on Friday, October 26th, come to the University of Canterbury and check out their workshop on "Reproducibility and Scientific Integrity in Science". There will be excellent speakers and an excellent registration fee (= zero). A detailed program, with a list of speakers and abstracts, is given HERE
As always, we hope you will encourage others to join TRN. We currently have 496 members (COUNT THEM!).  We really, really, really want to push this number past 500. Is there anybody you know to whom you would be willing to forward this email and ask to consider joining?
The ultimate goal of this website/network is to encourage journals to publish more replication studies.  The more members we have, the more journals can see that there is a substantial research community who support a greater role for replications in economics.
The Organizing Committee of TRN"

© 2018, David E. Giles

No comments:

Post a Comment