Publishing Research and Getting Feedback

Posting a new project is straightforward. Mandatory fields are only the title, research field, authors, and abstract. You are welcome to use links to Arxiv or SSRN to connect to your existing projects, which should save time for you and a bit of storage space for us.

We use research groups and projects' timestamps to protect your ideas. You can only see the projects posted in the research groups to which you belong. Other users, including non-logged or anonymous users, cannot view, rate these projects or even know about their existence. The only exception is the Playground group which we use to demonstrate the platform. Every user has Playground among their research groups. You can post there if you want wider exposure.

The project's timestamp shows the exact date and time the project was first published on the platform. You can use this information to argue about scientific priority of published ideas, but the platform does not provide any additional tools or support in these disputes (similar to most other academic repositories). It might be also hard to argue about the priority of early stage research, because many people can independently come with similar questions.

Other users can read and review the research as soon as you publish in their groups. You will see the distribution of ratings on the project’s page after it has received enough reviews. You cannot edit your published projects because we need the project to match its comments and reviews. However, you can always delete it and make a new one. Please avoid doing to projects with attached referee reports as it also deletes all the reports, comments and ratings given to the project.

You can also publish the project anonymously. You will still see it in your projects and observe its reviews, but others will not be able to see its authors or see it among your projects on the author’s page. Doing so might be a good option for projects you are uncertain about or willing to donate.

Have an open research question or proposal you want others to explore? We would be really glad to see it. Please post this project and choose the "Open question" adoption status. But beware that by doing so you might lose the rights to authorship and intellectual property derived from the project.

Our main mission is to make feedback fast and accurate. There are three forms of feedback which can be used simultaneously or on each own:

  1. Referee reports: extended reports describing the research’s contribution, deficiencies and ways to improve it. Authors of referee reports can choose to hide or disclose their identity from project’s author, but always remain anonymous to third parties.
  2. Comments: short public communication about the project which can be anonymous or not.
  3. Projects' Ratings: categorical ratings of project’s design and contribution. All the ratings are anonymous by design.

Projects' Ratings

You can rate each research project on two dimensions: research design and potential contribution. Potential contribution is a measure of changing our beliefs on important questions within the discipline/field. Research makes a large contribution if it makes large shifts in beliefs on questions of high importance. If the project is in early stages, think of its impact on the beliefs when it is completed and published. However, if the project’s conclusions do not follow logically from its existing or potential results (interpretation issue), we suggest to adjust the evaluation of its contribution accordingly.

Research design reflects our confidence in validity of current and potential results. Theoretical research can have design issues if either assumptions are contradictory or derivations are incorrect. Empirical research can have design issues due to faulty data, unresolved significant endogeneity concerns and many others. Experimental research can have design issue if it leaves open alternative channels or if its results are unlikely to extend to real-life phenomena it studies (external validity).

It is important to calibrate the grading system so that any disagreement in rating comes from diverse evaluations and not from diverse interpretations of ratings. We suggest the following interpretation for the potential contribution ratings:

The contribution is impossible to discern from the description.
Minor or No Contribution
The contribution is not significant enough to be published in any top or field journal.
The contribution is significant enough to be published in a field journal if done correctly.
The contribution is large enough for a leading journal (discipline/general interest) if done correctly.

We suggest the following interpretation for the research design ratings:

It is impossible to discern the research design from the description.
There are serious weaknesses in this design. As it is, the project is unlikely to be published.
The design is acceptable within the discipline, but might have minor weaknesses.
The design meets or exceeds highest standards in the discipline.

Your karma reflects your positive contribution to the platform. Every new user gets 10 karma points after posting the first project under their real name. Reputation improves from providing accurate ratings of other projects and decreases from providing inaccurate ratings. You also earn karma by donating your research ideas: the idea adopter pays you in karma points. In a planned update of the platfom, giving good referee reports would also improve your karma. You can spend karma points on using AI assistance services and (in a coming update) on requesting feedback to your projects. To sum up, your Karma=Initial Endowment + Scoring Reputation + Net Change from Trading Ideas - Costs of AI services

How Project Ratings Affect Your Scoring Reputation

Your reputation gain from the project’s rating (accuracy score) depends on rating given by others both before and after you. You get the highest reputation rewards if your rating correlates more with ratings given to the rated project than to ratings given to other projects. More specifically the following formula determines your accuracy score reward for one project if you give at a score $x$:

$$S=\alpha\left({1\over F(x)}{n_m\over n}-1\right)$$

Here $F(x)$ is the proportion of ratings equal to x out of the whole universe of ratings (across all the items), $n_m$ is the number of matching ratings for the same item and $n$ is the total number of ratings for this item. The constant $\alpha$ is a scaling constant which is currently equal to 0.5 but we expect to do more experimentation with its value.

Given the formula, your reputation gain $S$ increases when you give uncommon ratings (low $F(x)$) which nonetheless match ratings given by other users. Note that grading all the projects uniformly regardless of their quality is not going to deliver any positive reputation gains. This holds true even if others are doing the same. As a good community member, you need to read the project and think it through before rating. This is also necessary in order to obtain positive reputation from scoring.

Tips to improve your scoring reputation:

  1. Rate only the projects which you are qualified to evaluate.
  2. Write a comment, if you see issues or potential contributions others are unlikely to see.
  3. Rate more projects.