Most of today was also spent conducting studies with 5 sets of participants. Today the data was more interesting, as I personally noticed a wider range in what people seem to like and dislike, such as two people preferring the large font size, one person not being able to tell the difference in the sensitivity levels, and one person saying they experienced high and low sensitivity exactly opposite from how they should be. We have ended the study with a grand total of 27 participants, and the summary of the results can be found here. I think this gives us a good base line to decide the presets of the app for the next study. Hopefully we can try some of these out tomorrow at the CS Open House.
After the study, I have simply been trying to get our data to be more user-friendly. Using this module for matplotlib called mpldatacursor, I have so far gotten our plots to display either their coordinates or a totally custom text, so it seems pretty doable. The only problem is that what we want the labels to say is probably likely to be whatever words were at that point in the text, and so far I haven’t found a way to do this. While we already have a dictionary which maps indices to words in the text, we still need to have some way of looping through every index, getting the corresponding word, turning it into a label, and then applying that label to the individual point that index represents. Unfortunately, the plotting isn’t so convenient — all we do at the moment is call the plot function on a list of numbers. So creating a label that is customized to each point is a problem I need to tackle tomorrow.
Aside from this, hopefully the Open House tomorrow will give us an opportunity to determine whether the results of the study are helpful in indicating general opinions tomorrow. We will try to get people to sign up for our second study, but without actual dates, I am not sure how successful we will be. Hopefully an Open House will give us more traffic than the 27 people we did see.