The file "data_analysis.R" shows some frequently executed statistical analyses on data collected in speech quality assessment studies. The files "itut.results15.csv" and "Batch_*.csv" represents the speech quality ratings collected from the crowdsourcing listeners, whereas "Laboratory_data.xlsx" are the quality ratings from the listeners in the Laboratory that executed the listening test in an optimal environment.
"itut.results15.csv" corresponds to the data gathered in the study described at [1] and [2] (CS 501).
Most of the analyses performed in "data_analysis.R" were for the article [3].
[1] B. Naderi, S. Möller, and R. Zequeira Jiménez, "Evaluation of the Draft of P.CROWD Recommendation," ITU, CH-Geneva, May 2018. [http://www.qu.tu-berlin.de/fileadmin/fg41/publications/naderi_2018_evaluation-of-the-draft-of-p.crowd-recommendation.pdf]
[2] B. Naderi, R. Zequeira Jiménez, M. Hirth, S. Möller, F. Metzger, and T. Hoßfeld, "Towards speech quality assessment using a crowdsourcing approach: Evaluation of standardized methods," Qual. User Exp., vol. 5, no. 1, pp. 1–21, 2020. [DOI: https://doi.org/10.1007/s41233-020-00042-1]
[3] R. Zequeira Jiménez, B. Naderi, and S. Möller, "Influence of Language in Speech Quality Studies in Crowdsourcing," in submitted to: 13th International Conference on Quality of Multimedia Experience (QoMEX), 2021.