Hi @sarontt
I think you would also need to investigate those 204 submissions with the decision = 8, where date_decide < upgrade to 3.4 date. If they were all declined before the review, or if the decision should rather be 3 (external review). Those that also have decision = 6 should probably have 3 instead of 8, but this you should double check.
Best,
Bozana
here is a small sample of the edit_decisions table from the 204 submissions with decision = 8 and prior to our update to 3.4.0.* which shows (declined = 18536) (round 2 decline = 18537) (round 1 decline = 21449):
MariaDB [ojs]> select * from edit_decisions where submission_id in (18536,18537,21449);
±-----------------±--------------±------±----------±---------±--------------------±----------------±---------+
| edit_decision_id | submission_id | round | editor_id | decision | date_decided | review_round_id | stage_id |
±-----------------±--------------±------±----------±---------±--------------------±----------------±---------+
| 11072 | 18536 | NULL | 4912 | 8 | 2022-09-23 10:46:23 | NULL | 1 |
| 11074 | 18537 | NULL | 4912 | 8 | 2022-09-23 11:51:58 | NULL | 1 |
| 11523 | 18537 | 1 | 4912 | 4 | 2022-12-09 11:32:26 | 3446 | 3 |
| 11840 | 18537 | 1 | 4912 | 14 | 2023-02-09 16:58:39 | 3446 | 3 |
| 12434 | 18537 | 2 | 4912 | 6 | 2023-05-05 05:12:41 | 3686 | 3 |
| 13916 | 21449 | NULL | 5210 | 8 | 2023-11-12 20:51:09 | NULL | 1 |
| 14334 | 21449 | 1 | 5210 | 6 | 2023-12-20 19:02:35 | 4153 | 3 |
±-----------------±--------------±------±----------±---------±--------------------±----------------±---------+
7 rows in set (0.006 sec)
there were 108 instances of where there were decision = 6 that mistakenly had an 8 instead of having the 3:
select ed.edit_decision_id, ed.submission_id, ed.decision, ed.date_decided from edit_decisions ed join submissions s on s.submission_id = ed.submission_id where ed.decision = 8 and s.context_id = 34 and and ed.date_decided < ‘2024-03-07’ and ed.submission_id in ( select submission_id from edit_decisions where decision = 6 );
i ran the update on my test instance to fix those which changed the ‘submission declined’ in editorial workflow statistics accordingly– however it still doesn’t add up correctly– but it is a lot closer now– submission declined before the update 332 - the 108 that got corrected from 3 to 8 = 224 = what ‘submission declined’ is after the update–
224 submission declined + 115 submission accepted doesn’t equal 356 submission received– looks like it’s missing 17 somewhere
any suggestions to where to look for other problems with our data?
thanks for your help on this!