Hello Sam,

I just had a quick look at your review. The issue is in how you are setting up your comparison.

When you click on the Create comparison button (in the Collaborate tab), after selecting Reviewer 1 and Reviewer 2, for the Code set you want to select your 'Screening criteria' codeset. For the 'Set attribute' option, you want to select your Group one code. This means you are running a comparison how Reviewer 1 and 2  coded the items in 'group 1' using the 'Screening criteria' codeset. I think you may had this turned around.

Then, when you 'run'  a quick report, select the 'Screening criteria' codeset for the 'Select this code(set)' option. You will then get a report showing how Reviewer 1 and 2 applied the codes from 'Screening critieria' against the items of in group 1.

I noticed that you have all of your codesets set for multiple data entry. You only need to set a codeset for multiple data entry if the intention is for 2 or more people to apply the codes from that codeset against the same item. If you are double coding/screening and you want to compare how each person coded the same item, then you will want your screening codeset to be set for multiple data entry. If you have an allocation codeset that is used to identify items for allocation, you would want that set for single data entry. That way you don't end up with items marked as 'not complete' that you actually want, and require. to be marked as complete.

Normally I keep all codesets set for single data entry unless 2 or more people will be assigning codes from that codeset to the same item and then comparing how each person coded that item .

Best regards,

Jeff

Hello Sam,

What you are suggesting would be fine. Many groups will double screen a number of items at the start of the project. Once they are sure that each coder is applying the criteria in the same manner they resort back to single coding to do the rest of the items.

What I would make sure of is that before you change back to single coding that you 'complete' all of your agreements and reconcile and complete all of your disagreements. Reviewers can get into difficulties if they switch back to single coding but still have a number of double coded items that have not been completed and reconciled. The 'un-completed' items do not show up up reports and can result in misleading reports. The functions to recover from this are available in the software but it is better to sort it out before you change your codeset from multiple to single data entry.

Best regards

Jeff