Analysis of primary care staff language about aggression at work using concordancing techniques to identify themes

Yvonne H Carter, J R Skelton, J E Kenkre, F D Richard Hobbs

Research output: Contribution to journalArticlepeer-review


OBJECTIVE: We aimed to describe features of language used during interviews about the extent of aggression and violence at work and their effect on primary care staff.

METHOD: Forty-four primary health care team members in the West Midlands were interviewed, and interviews were recorded on videotape. The language content of these interviews was analysed using Cobuild concordancing software. Outcome measures used were word frequency, collocation and mutual information (MI) scores for language use.

RESULTS: A total of 17517 words spoken by interviewees were analysed. Violence in this sample was perceived as occurring principally in connection with unmet demands for such things as prescriptions and referrals. Only patients were perceived as violent; health care workers used other terms to describe their own feelings and responses. Sixty-eight specific incidents of violence were recounted, features perceived as salient being drink, youth and to a lesser extent mental illness.

CONCLUSIONS: Concordancing software can be successfully used in the qualitative examination of videotaped interviews. In this study, the technique rapidly identified a number of perceived training needs among a variety of primary care staff.

Original languageEnglish
Pages (from-to)136-41
Number of pages6
JournalFamily Practice
Issue number2
Publication statusPublished - Apr 1997


  • Aggression
  • Attitude of Health Personnel
  • England
  • Humans
  • Interviews as Topic
  • Language
  • Patient Care Team
  • Physician-Patient Relations
  • Primary Health Care
  • Software
  • Violence
  • Journal Article
  • Research Support, Non-U.S. Gov't


Dive into the research topics of 'Analysis of primary care staff language about aggression at work using concordancing techniques to identify themes'. Together they form a unique fingerprint.

Cite this