Today I spent some time changing the Python program that correctly formats the email that we send of our data. Earlier, it worked such that data from one text reading had to be in a separate file in order to be processed, but what we’re doing now is having all of the data from one person — i.e. 7 texts, in the case of this first study — in one file. So rather than having to open up that file and manually split it up, I automated the process. Now, the program takes in all this raw data, splits it into as many files as necessary (in a folder it creates called RawFiles) and then carries out all of the tidying up that puts the data back into dictionary form. Then the output of my program is all in one text file that can be put in Adam’s program for fairly straightforward plotting.
After doing this, I suggested we try to do a run-through of our study to smooth out issues in the app, the texts and the survey. We definitely discovered a lot of these. We decided that once the user tells us what kind of font size or controls they prefer, the rest of the study should be performed with those settings, rather than forcing something that may interfere with their judgement of everything else that we want them to pay attention to. Then we discovered some issues with how we would reset the app after reading a text. We don’t really want to have to email the data after every text, and wanted to be able to have all the data from one person in one email, which wasn’t what was happening. That was changed, and we decided it would be useful to have the settings be remembered so we don’t have to input the next text or the ID number every time. Overall, I think the attempt to do it just between us what helpful, particularly for me — I haven’t spent a lot of time using the app myself, so this showed me a lot of things I hadn’t realized before, such as how the high sensitivity is far too high for my taste, and even the smallest font size is large by my standards. I think there is a lot to learn about what people prefer on average so I think that the study will yield interesting results.