Critical Incident Semantic Analysis Suite
CISA is a narrative-discursive methodology for analysing identity construction in interview and narrative data. It combines template analysis with computational pattern recognition, offering systematic detection while keeping the researcher in interpretive control.
CISA scans transcripts for discourse markers across eight categories — identity markers (I am, who I am), temporal shifts (now, I used to), emotional language, conflict indicators (but, torn between), modal verbs, and more. When multiple marker types converge, this signals a critical incident — a moment where identity work is actively occurring.
"I feel like I am being pulled in different directions constantly. I belong in the classroom with my students but I am isolated in the business world. I know who I am as a teacher but I am lost as an entrepreneur."
Here, repeated I-statements, conflict markers (but appears three times), and emotional language (feel, isolated, lost) converge — signalling active identity negotiation between two professional selves.
Once critical incidents are identified, CISA offers different approaches to analysing tensions and contradictions:
Both tools generate reports at transcript and dataset level.
All tools run locally — your data never leaves your device.
Developed by Dr Pauline Prevett
University of Manchester
School of Environment, Education and Development
For enquiries: pauline.prevett@manchester.ac.uk