Save valuable time and reduce errors by automating processes

The challenge

Science on Stage e.V. tried to work in a more data-oriented way, but data from many different sources proved to be a significant hurdle. In particular, the manual nature of the export, cleaning and analysis of the data made these processes tedious, time-consuming and error-prone.

The data

The project focused on the data sources Surveymonkey (survey data from participants) and Matomo (anonymized download statistics for the organization's educational material).

The solution

The CorrelAid team worked closely with Daniela from Science on Stage e.V. and supported her by developing R scripts to (partially) automate and better structure the previously manual processes for those data sources. As part of the project, the CorrelAid volunteers split into small sub-teams and tackled data processing steps, e.g. the extraction and preparation of survey data from Surveymonkey including the immediate and long-term feedback for the NPO's webinars. The team created analysis scripts for relevant metrics such as length, topics and number of participants. Another part of the team focused on extracting and analyzing anonymized web tracking data from Matomo on download statistics of Science on Stage's educational material. The volunteers analyzed the data and provided CSV files and HTML reports with visualizations that broke down the relevant metrics by various categories relevant to Science on Stage.

The scripts and the associated folder structure were handed over to Daniela Neumann at the end of the project.

The impact

Thanks to the early and close involvement of Daniela from Science on Stage in the project process, results could be used and consolidated in the long term: The scripts are still being used two years later and Daniela independently takes care of necessary adjustments to the code, e.g. when switching infrastructure provider from Surveymonkey to Lamapoll. The outputs and experiences of the project also served Science on Stage as an inspiration and basis for new use cases: Daniela was able to transfer the code to extract anonmyized statistics on access numbers to web offers in addition to the download figures for PDF documents. And after master data had been manually copied from over 300 Word application forms every year, this is now done by an R script developed by Daniela.

By consolidating the project results and developing new use cases based on the code, Science on Stage e.V. was able to save a considerable amount of time and avoid serious errors in data processing. An important step in the evolution of Sciene on Stage towards more data orientation in their work!