3 Conclusions

        The goal for this project was to conduct a longitudinal analysis to expand on previous projects. However, much more was accomplished; Python coding, the basics of APIs, how to display data in R, how to write a paper and publish it online through R Bookdown, and much more was learned through this project.

3.1 Summary: Interpretation

        Through this analysis, the variables year, energy, mode, speechiness, acousticness, instrumentalness, liveness, valence, tempo, year * speechiness, and year * valence were all found to be significant due to their p-values being less than 0.05. It was interesting that, in addition to the variables determined to be significant through previous projects, mode, liveness, and instrumentalness were also significant predictors. Adding more datapoints and the variable of year caused these variables to be significant. One limitation that was not accounted for was the occurrence of one artist having more than one song in the Top 100 tracks.

3.2 Suggestions for Further Study

        In future analyses, it would be interesting to see if there were any distinct differences of top songs in 2020. It has been such a turbulent year, through COVID-19, the riots, and the election. There could easily be a noticeable impact on the types of songs Spotify users have been drawn to listen to the most. Another interesting topic to research would be generational differences in music choices. Through the Spotify Web API, public user profiles are accessible by the user’s URI. Future studies could take playlists from these public profiles and compare the types of music listened to by different age groups to see if there is any relationship between age and genre of music that is most listened to.

Impact

        The datasets created for this project have been uploaded to Kaggle so future analysts with an affinity for music have more options for data analysis.