We encountered a very strange issue with Editorial Activity Statistics (Statistics > Editorial) in one of our journals. I wonder if anyone has experienced the same or has any insight into the issue. Any suggestions will be greatly appreciated.
We looked at the editorial activity data for 2020 (period from 2020-01-01 to 2021-01-01) on the following three dates (Jan 29, Feb 9 and Feb 16, 2021). The stats looked different on each date (please see attached image comparing the collected data see by side). Where the data mostly varied was percentage calculations.
For example, Post-Review Rejection Rate changed in each case while the underlying totals used to calculate it remained unchanged – Received submissions (650) and Submission Declined (Pos-Review) (201). Additionally, there seems to be an error in calculating the percentage, which should be 30.92%.
Unfortunately PKP Documentation on Editorial Activity Statistics (Statistics) doesn’t provide any in-depth description about calculations logic. Hoping the community wisdom will shed some light on this. Thanks.
Thanks for posting your question. Newer OJS versions actually provide an explanation of how the Acceptance and Rejection rates are calculated. I’ve provided the explanation for you below.
The percentage for the selected date range is calculated for submissions that were submitted during this date range and have received a final decision.
For example, consider the case where ten submissions were made during this date range. Four were accepted, four were rejected and two are still awaiting a final decision.
The acceptance rate will be 50% (4 of 8 submissions) because the two submissions that have not reached a final decision are not counted.
I doubt that the above explanation of how acceptance rates are calculated in OJS is correct. Looking at the (rather complex) SQL query used to calculate the acceptence rate the denominator is the count of all submissions that were submitted during the date range (not only submissions with a final decions). So for the above example given by Patricia I expect that OJS (3.1.2) would calculate an acceptence rate of 40% (4 of 10 submissions).
Basically the numerator of the acceptance rate includes all submissions that were submitted during the date range AND where the last decision was either “accept” oder “send to production”. The denominator is the number of submissions submitted during the date range.
Below is a simplified version of a SQL statement that can be used to get the OJS acceptance rate for a journal with id = 5 (s.context_id = 5) and year 2022 (s.date_submitted BETWEEN ‘2022-01-01 00:00:00’ AND ‘2023-01-01 00:00:00’).
I’m not sure if the OJS definition of acceptance rate is common in publishing. Most definitions I’ve seen simply divide the number of accepted submissions in a given period by the number of new submissions in the same period. (My understanding is that neither the accepted submissions [for the numerator] need to have been submitted in that time period, nor do the new submissions [for the denominator] need to have a final decision in that time period.)
If anybody knows an agreed upon standard for defining the acceptance rate please let me know!
Best wishes
Armin
SELECT COUNT(CASE WHEN ed.decision IN (1, 7) THEN 0 END) AS numerator, COUNT(s.submission_id) AS denominator, COUNT(CASE WHEN ed.decision IN (1, 7) THEN 0 END) / COUNT(s.submission_id) * 100 AS submission_acceptance_rate FROM submissions as s left join edit_decisions as ed on ed.edit_decision_id = (select ed.edit_decision_id FROM edit_decisions as ed where ed.submission_id = s.submission_id order by ed.stage_id desc, ed.round desc, ed.date_decided desc limit 1) WHERE s.submission_progress = 0 and s.context_id = 5 and s.date_submitted BETWEEN '2022-01-01 00:00:00' AND '2023-01-01 00:00:00'