Hi Rachel,
It seems that you have some overlaps in your second set of allocation codes. (We are not sure how this has occurred, but you could have inadvertently selected both included and excluded items for one group, or some such. We find people tend to use the I and E flags differently, according to their own personal preferences and working methods.)
If you look in the Search tab of your review 11970, you will see we have carried out some searches (numbers 141 to 153), and found there are definitely some items that have been allocated to more than one person. (It appears 170 items have been coded with Ann's Group and RAK-Group, and 165 items coded with Ann's Group and Mairead's Group.)
The total number of items coded in all three groups is 1470 (adding 894 + 288 + 288), whereas the total number of unique items coded in any one of the three groups is 1135. (1135 plus the duplicates 170 + 165 = 1470.)
The twenty items that have gone astray in your numbers may just have arisen as you are only looking at certain included items or some such.
(One way to reallocate and check again would be to create a new code, select the items which have been already coded by each group (i.e. are started) - by clicking the Collaborate tab, clicking on the relevant number, selecting all the items that are listed, and right-clicking on the new code assigning it to all items. Then create another second code and randomly assign items without the first new code to one of 5 child codes below the second new code, using the inbuilt function to assign codes randomly... You could then assign items with one of the five child codes to Ann's group, items with another of the five child codes to Mairead's group, and finally items with any of the remaining three child codes to RAK's group, giving a 20%, 20%, 60% split.)
Let us know if you require any further assistance addressing this issue. (We did not want to change coding or anything else on your review whilst it was being accessed.)
Cheers,
Zak