Key facts

• The categories (coding units) are decided before the analysis and must be operationalised, so anyone would code the same way.
• Tallies become frequencies, which are quantitative data.
• Thematic analysis looks for themes instead and keeps the data qualitative.
• Inter-rater reliability: two coders work independently and their tallies are correlated. +0.80 or above is good.

How to do it

1. Read each example and match it to the category whose definition fits.
2. Tally each category and find the total.
3. Percentage = frequency ÷ total × 100.
4. Percentage change = (new − original) ÷ original × 100.