Bar Songs was a small project I carried out for a graduate course in qualitative research methods. The basic idea behind it was to make short clips of music based on simple data sets well suited for traditional bar charts.

The Bar Songs prototype I built with Processing was deployed across a convenience sample of UMBC students, and I collected survey and interview responses for analysis.

I found that users of Bar Songs were capable of identifying gross patterns, but that interpretation of the generated songs was not reliable for detailed analysis of the data sets.

In all, this was a fun project, and gave me a chance to learn a new programming language and experience all the neat little foibles and follies of a first-time qualitative study deployment.