Frequently Asked Questions
What is Failure to Replicate?
Failure to Replicate is, as the name suggests, a database of failed attempts to replicate previously-reported psychological findings.
What's the point?
The point is to maintain as comprehensive a record as possible of the psychological research literature. That includes not just studies that successfully replicate an effect, but also those that fail to. Think of it this way: if you're thinking about starting a program of research that builds in part on someone else's published work (which is to say, almost always), you'd like to know what the odds of replicating that work are. If there are 10 published replications of an effect but, say, 60 unpublished failures to replicate, the odds of your project succeeding are going to be much lower than they'd be if all 10 attempts to replicate the effect had succeeded and been published. But you wouldn't know which of the two scenarios (or any number of others) was actually true unless you had some way to track both successes and failures. That's the point of this database.
Who can use this website?
Anyone with an internet connection! That said, the database is unlikely to be of much use to non-psychologists.
Why do I have to give an email address when adding an entry? I don't want to be spammed!
An email address is necessary so that other researchers who want more information about your study can contact you. Email addresses are only displayed to registered members, so your address won't be displayed to nasty, malicious email harvesting bots. Of course, if you're really worried about it, you're free to use a fake email address. We promise we won't tell.
How much information should I provide about my study?
As little or as much as you like! It's entirely up to you. There are no guidelines to speak of; the point here isn't to formalize some sort of standard for reporting, it's just to ensure that a minimum of information about as many studies as possible is available to others. Obviously, you should provide enough information that other users can determine what effect you tried to replicate (especially if it's a well-known one); but beyond that, it's your call.
I'm an [anthropologist / economist / biologist]. Can I add my own replication failures to the database?
At the moment, the database is designed specifically for psychology studies (e.g., witness the spiffy "sample size" and "summary of experiments" fields). It's possible that at some point in the future, we'll expand the database to include studies conducted in other scientific disciplines, but there are no current plans to that effect. Sorry!
Should I report the results of studies that I expect to publish or that have already been published in peer-reviewed journals?
The objective of this website is to maintain a repository of only those studies and experiments that cannot be found in the primary peer-reviewed literature. You should only report a study here if you're quite sure it's never going to see the light of day in a peer-reviewed journal. (If you report a study here and later publish it after all, you can always delete it from the website.)
Why are "full replications" listed on this site? The name says failure to replicate!
Emphasis is placed on replication failures rather than successes because successful replications and extensions of previous findings are published at a disproportionately high rate relative to unsuccessful replications. Most replication failures would never be reported anywhere if not for informal repositories such as this one (the "file drawer" problem). Many, if not most, successful replication attempts do end up being published in peer-reviewed journals. Nonetheless, since the overarching goal of this website is to provide as comprehensive an accounting as possible of the research literature, we encourage researchers to report any replication attempt that is not expected to be reported in a more formal venue, successful or not.
Can I report replication attempts that I was only peripherally involved with or have only second-hand knowledge of?
Yes, but please make sure the lead researcher (and ideally, any other people involved) is aware that you are adding the information to the database. Also, please conduct a search first to make sure the study has not already been added by someone else.
Should I report the results of studies that test novel hypothesis rather than attempting to replicate previous effects?
No. This may change in the future, but for now, you should only report studies/experiments that involve a replication of previous findings. It's fine to report studies that replicate previous findings as part of a broader objective (e.g., to test whether an effect generalizes to a new situation or population), but you should focus primarily on the replication component rather than the novel extension. Similarly, conceptual replications are ok, but make a point to note the salient difference between your experiments and the canonical result.
Who made this?
The site is built and (occasionally) maintained by Tal Yarkoni.
I found a bug! I have a suggestion! I don't like this place at all!
Send all bug reports, comments, and suggestions here! Preferably with lots of exclamation marks!