Appraisal Theory Analysis (Martin and White)

Description

Codes evaluative language into attitude (affect, judgement, appreciation), engagement and graduation to analyse how users express feelings and stance far more precisely than sentiment scores.

Sentiment analysis reduces evaluation to positive or negative. Appraisal theory distinguishes Attitude: Affect (emotions: 'I was furious'), Judgement (evaluating people's behaviour: 'they were dishonest') and Appreciation (evaluating things: 'the app is clunky'); Engagement (how voices are positioned: acknowledging, denying, conceding); and Graduation (force and focus: 'really', 'kind of'). This reveals, for example, whether complaints target the product (appreciation) or the company's integrity (judgement), which have different implications. Systematic coding with defined categories and reliability checks provides rigorous evaluative analysis of interviews, reviews and support conversations.

Attributes

Loss type
Unmigrated
Effort
Weeks

Steps

  1. Select texts and define the evaluative targets of interest (product, company, self, others).
  2. Code instances of Affect, Judgement and Appreciation with polarity and target.
  3. Code Engagement (monoglossic vs heteroglossic resources) and Graduation (force, focus).
  4. Check inter-coder reliability on a subset and refine coding guidelines.
  5. Aggregate patterns by target, segment and context; read examples in context.
  6. Translate findings into design or communication actions (e.g. integrity concerns need trust repair, not UI polish).

Sources

  • Martin & White, The Language of Evaluation: Appraisal in English (2005)
  • Halliday & Matthiessen, Halliday's Introduction to Functional Grammar (4th ed., 2014)

Also known as

  • Appraisal framework
  • Evaluative language analysis
  • Systemic functional appraisal

Categories

  • Content
  • Linguistics, Semiotics & Discourse