Skip to content
PaperFren

Research method

Thematic Analysis

8 studies

Ask PaperFrenWatch this

Thematic analysis codes qualitative material — interview transcripts, focus-group talk, open survey comments, reflections — into patterned meanings. Reflexive versions interpret through a theoretical lens (ableism, culture, belonging); more descriptive versions catalogue what participants said about a tool. Themes are researcher-made patterns, not prevalence estimates, not social-network coefficients, and not randomised outcome tests. Some papers tagged here mix themes with surveys or learning-analytics traces; the theme layer is still not the trace.

Online-learning papers use themes when they need to say what exclusion, loneliness, competence training or blended (dis)engagement meant to participants. They answer 'what meanings recur in these texts?' Their main limitation is over-reading: 56 students saying accommodations helped, or overwhelmingly satisfied live-chat ratings, can sit beside themes without becoming a trial of assessment redesign or chat software.

Studies

8

Findings

6

6 supporting · 0 challenging · 0 qualifying citations

Open tensions

2

Currently

What we know

  1. At a Finnish research university, a pilot of 25 and a main survey of 139 disabled students were interpreted with reflexive thematic analysis through ableism and disablism. Although the prompt asked about assessment broadly, experiences centred on exams. Fifty-six respondents explicitly said accommodations helped them show their abilities, yet students still wanted more authentic, student-centred assessment. Time, visuality and productivity norms were read as excluding. Self-selected responses at one institution are not a population barrier rate.
  2. Two hundred forty-six Australian psychology students with live chat available (155 blended, 91 fully online) were overwhelmingly satisfied and felt chat showed the teaching team cared; fully online learners were often more positive, using chat for timely academic help rather than only email or boards. That is perception and self-reported use, not experimental learning gains.
  3. Across three summers, 43 EMBARC students joined five semi-structured focus groups (n = 8, 9 and 26 by year). Themes emphasised mentorship and navigation, hard skills and hidden-curriculum knowledge, peer bonding, and mindset shifts; many described a change in 'embedded inner culture' that validated cultural strengths rather than demanding assimilation. Themes do not estimate degree-completion effects versus a control.
  4. Nine online disabled students’ interviews linked asynchronous pacing to reduced stigma and to loneliness, with misunderstanding of disability as a social-difficulty theme. Informal activities and staff dialogue were proposed supports. Theme presence in nine transcripts is not an intervention effect.
  5. A five-month upper-secondary blended case used observation to analyse (dis)engagement. Digital tools were sometimes used to enhance learning and sometimes deliberately avoided. Teachers’ observed work pace exceeded students’ on average, producing tension. Engagement was negotiated across context, teacher leadership, activity and the student-as-learner — an interpretive case, not a national blended-learning effect.

Largest unresolved question

Help-seeking and belonging are not one theme. Live-chat students were overwhelmingly satisfied and treated chat as just-in-time academic access, especially fully online learners (91 vs 155 blended). Loneliness interviews and ableism themes describe exclusion, stigma and exam visuality. K–12 blended observation describes negotiated disengagement when teacher pace outstripped students. Positive tool themes and exclusion themes can both be accurate in different rooms.

Procedure

How it works

The main stages of the method. Specific protocols and analysis choices vary between studies.

  1. Assemble qualitative material

    Interview transcripts, focus-group talk, open comments, or reflections are gathered. Who wrote or spoke, and under what prompt, already limits which meanings can later be found.

  2. Code segments of text

    Passages are labelled. Reflexive versions interpret through a theoretical lens such as ableism or belonging; more descriptive versions catalogue what participants said about a tool.

  3. Group codes into themes

    Themes are researcher-made patterns across the corpus, not social-network coefficients and not randomised outcome tests. Recurrence in these texts is not a population rate.

  4. Write what the patterns mean

    The report says what exclusion, loneliness, competence training, or blended (dis)engagement meant to these participants. A survey rating sitting beside the themes is still a rating, not a trial of the tool.

  5. Do not treat a recurring theme as prevalence

    Thematic analysis answers what meanings recur in these texts. Self-selected disabled students at one university, volunteers in a small yearly sample, or satisfied live-chat comments do not become a trial of assessment redesign or chat software.

Common misconceptions

  • If a theme recurs, it is common in the population.

    Exam-centred ableism themes come from 139 self-selected disabled students at one university (plus a 25-person pilot). EMBARC themes come from 43 volunteers whose yearly n ran 8–26. Recurrence in those texts is not a base rate.

  • Thematic analysis of end-of-course reflections shows that an AI dashboard improved learning.

    The graduate-course paper reports SNA differences (density F = 4.90) and frames prediction as a decision aid. It is not a multi-institution randomised trial and does not guarantee later achievement for each learner.

  • Overwhelmingly positive live-chat comments prove the tool closed the support gap for all online students.

    Ratings come from 246 psychology students at one university who had chat available. The study does not measure learning gains, and loneliness/ableism papers show other online students describing isolation and exam exclusion.

Related