The following is a recap of the group discussion, “Experimental Data Science - Lessons from the Lab,” led by Jason Cherry, data scientist, advancement strategy at the University of Colorado. This group discussion was part of the Data Analytics Symposium (DAS) at Prospect Development 2019. Check out more PD 2019 group discussions here. Check out their notes below.
Many participants shared challenges when it comes to data preparation, including:
- Data is not always in the correct format for analysis
- Often much information needs to be collected about the meaning of the data when its received
Communicating or convincing others of the need for data can sometimes be a challenge, too. It’s generally best to start with small, easy projects and build a persuasive portfolio from there.
Challenges when collecting data from donors include:
- Asking for too much info on first contact leads to people not filling out the form
- Asking for too little leads to not enough data
- A two-stage data request may be appropriate — get enough data initially to follow up, then follow-up comprehensively later