I am trying to write a script for reliability check. I have the template script for checking reliability that compares two cells that have the same ordinal number. However, my coding doesn't work in this way. Each coder code the behavior whenever they observe (event-coding), so each may have different numbers of cells. I have tried putting "onset" instead of "ordinal" in the template that I have, but the result didn't seem right (agreements were 100% although those weren't). I found a similar question titled "Reliability scripting question", but I couldn't quite understand the script given in the answer. Since I am a beginner of writing a script, I couldn't fully understand the script, and therefore, I'm not sure which part should be changed when I edit it to fit to my data.
asked 12 Aug '16, 00:36
I just reviewed the post you mentioned and while the sample code in the answer was a bit verbose, it can serve as a good starting point. The basic idea is to construct a column that merges together two columns such that a cell is created in the new column whenever a cell starts or stops in either of the original columns.
You would then check each of the cells in the merged column and see if it denotes a disagreement or agreement (when the value of a code is blank, denoted in the image above as the name of the code inside less-than and greater-than symbols, it means it was not coded in the respective column).
Documentation on how to use the create_mutually_exclusive function can be found on the API reference page: core documentation
Additional documentation is also available at: full documentation
Since you are new to scripting, here is a quick skeleton script to get you started:
answered 15 Aug '16, 16:46
Shohan Hasan ♦♦