TeamFrames data analysis

Today I discussed with Pierre about the methodology of making sense of the data coming out of the TeamFrames questionnaires. Just looking at the categorisation I made of the respondents with the relative profile, he expressed the opinion that what we called “collaborators” are more cooperators and that our “cooperators” are more individualistic person. His approach is more to understand if the tool’s log may predict the study style and vice versa. He dislike the idea of validating the data through the results of the shs project because he feel that the students were too detached from a constant usage of the tool, consequently the outcomes might not connected with the usage patterns. Nevertheless a validation of the found clusters is possible and desirable. To achieve that we can try to follow the following paths:

a) clustering the users using the usage patterns and comparing the obtained clusters with the one already got;

b) asking a different bunch of question in the second questionnaire and see if the answer patter are kept the same;

c) find external validation in the literature.

Other interesting things to look at are:
1 – comparing the usage patterns with the clustering we have got: “Do cooperative people respected more deadlines?”

2 – compiling the matrix of matrix of group number X number of people for each cluster.

Leave a Reply