Hi Daniel,
I think you would like to automatically change the masters so that whatever has not been coded will never be a master item: after this is done, running "mark automatically" will catch more items and spare you some manual work, is that what you have in mind? In any case, at the moment it is not possible, I'm afraid! However, it is something I have in the back of my mind: it's tricky to implement, so I've left it in my wish-list for the time being.
Still, I have another suggestion for you (I gave a look at your review stats, I hope you don't mind): since you have already done many exclusions, it might be unnecessary to de-duplicate all your items. EPPI-Reviewer allows to start the screening with all items marked as Included and then gradually mark them as Excluded, but also permits the other way round (starting with everything Excluded); for this reason, the duplicate search algorithm evaluates both included and excluded items: what is does not evaluate (with one exception*) is deleted items.
What you could do is:
1) Walk through your list of excludes and mark them as deleted.
2) In duplicate checking, choose the first “reset” option.
3) Get new duplicates once more.
At this point you should have much shorter list of duplicate groups to deal with. Also, if you know that all remaining items will have at least one code applied (I’m thinking of an “include” code of some kind, for example), you could use the “Advanced Mark Automatically” and change the “Codes Threshold”.
4) You will now be able to walk through all your deleted items and mark them as Excluded once more (if needed) – it can be done before, after or during your duplicate sifting: it doesn’t matter when. This will work for you because you have only one item that was explicitly marked as deleted (the only “manually entered” item); all other items are either Included, Excluded, Duplicates, or part of a deleted source. When you’ll list the deleted items, you’ll get all items that were previously excluded (plus one), and hence you won’t lose any information – as a double safety: duplicates and items contained in a deleted source do not show up when listing “Deleted Items”, and even if you’ll explicitly try to mark them as included/excluded, they will remain untouched.
Of course, proceeding in this way is possible only if you don’t need to know how many duplicates you had in the excluded items list.
Note for the other readers: what I’m suggesting here applies to Daniel’s review because he didn’t finish his duplicate checking after importing all items; instead, he proceeded directly to screening and coding. The best practice remains different: whenever possible, the quickest way to get duplicates out of the way is:
- Import all items (without doing any coding)
- Run “get new duplicates” (just once) and use a mixture of manual and automatic marking to evaluate all groups. At the end of this phase, all duplicate groups should be marked as completed.
- Proceed with the rest of the review: should you stumble on some more duplicates, you’ll be able to manually mark them at any time.
I hope this helps,
Sergio
* The one exception is: whenever an item is already present in some duplicates-group, it will always be re-evaluated without exceptions. This is useful to minimise the amount of overlapping groups that would be created when running “get new duplicates” incrementally.