Department of Biostatistics Satisfaction Survey

This CQI project lead by William Wu is to development and implement a satisfaction survey for the department. It will be used periodically to gather anonymous feedback from our collaborators, providing information to guide our improvement efforts. This project's team members includes: Frank Harrell, Daniel Byrne, Robert Greevy, and Cole Beck.

The current version of the survey is viewable here.

Data Handling Protocols

Filtering Names

Some survey responders may break anonymity by including names, theirs or others. We have the following protocol to re-anonymize the data.

1. We will finalize the survey by taking it offline 5 weeks from the date we emailed out the survey requests. The link will redirect to a page thanking the responder for trying to answer the survey and informing them the survey has been closed.
2. The final survey dataset will be stored as a text file, which no one will be allowed to access directly.
3. A names filter will be used on the final survey dataset to filter out any names.
4. The filtered dataset will be opened and checked for integrity by Rob, Cole, and William. It just needs to be readable by a human.
5. If the filtered dataset is readable, the unfiltered survey data will be deleted. Any names remaining in the dataset will be deleted by hand and the filter will be updated for the next time the survey is done. If the filtered data is not readable, we will rewrite the names filter as needed to clean the data. (Loop steps 3-5 as needed.)

A note on the data filter
If Jane Doe is providing email addresses for her top two collaborators, Bill Gates and Steve Jobs, then she will also provide names for the names filter, e.g.
Jane
Doe
Bill
William
Gates
Steve
Stephen
Steven
Jobs.

The filter will only treat special names like Jobs as case sensitive. For example, the filter will take a sentence like

• I've been very pleased with jane. Jane's come through in the clutch on many jobs! Thanks, Jane. - Steve Jobs
  

and return a sentence like

• I've been very pleased with name. name_'s come through in the clutch on many jobs! Thanks, _name. - name _name_