A key process in qualitative research data collection is the transcription of interviews, focus groups or fieldwork notes.

Transcription can be done manually by the study team or third-party service provider. It can also be done by artificial intelligence (AI) software, which is becoming the preferred way researchers are choosing to transcribe their research data.

It is important when transcribing research data that you use transcription tools that have been vetted by the university.

Campus has contracted with three media vendors that provide manual transcription of data, for a fee. More information about these vendors can be found here.

Campus also has several approved AI tools. However, before using any of these AI tools for transcription, study teams should first understand the classification of the data they plan to transcribe. Information about UW–Madison’s current data classification framework is available here. Data classification helps determine the level of protection required for data and which tools may be used appropriately. Human research data are often classified as Sensitive or Restricted under the current classification framework, and PHI is considered Restricted. Investigators are responsible for classifying their data before determining whether a particular AI tool is appropriate for use.

UW–Madison’s approved generative AI tools identify the data classifications for which each tool may be used. Currently, Zoom and Webex are the two campus-approved tools that offer AI transcription features. If a study team wishes to use a third-party AI tool for transcription not approved by the university, determine whether a cybersecurity risk review is required or consult the Risk Management & Compliance cybersecurity team for help.

As a reminder, any use of AI in a research study must be described in the IRB application and/or study protocol and receive IRB approval before implementation.