Political Methods & Tools (DP IB Theory of Knowledge): Revision Note

Surveys and opinion polling

Collecting responses from a sample of individuals to estimate attitudes, beliefs or voting intentions in a wider population

Sampling bias: the sample does not match the population, so results reflect who was included rather than the wider opinion.

Non-response bias: certain groups refuse to respond or cannot be reached, so their views are missing.

How the questions are worded has an impact, e.g. leading or emotionally loaded terms push respondents toward particular answers.

Sometimes respondents give answers that sound acceptable rather than reflect what they truly believe or would do.

Interviews and focus groups

Asking one person questions to explore their reasons and experiences in depth, or asking questions to a small group at the same time and using the group discussion to reveal shared views.

Small or unrepresentative participation: a few voices are studied, so conclusions may not be applicable to the wider population.

Interviewer influence: tone and follow-up questions steer what participants choose to say.

Group dynamics: dominant speakers and conformity pressure can suppress disagreement in focus groups.

Case study research

Investigating one case in depth to explain causes, decisions, and outcomes.

Limited generalisability: What explains one case may not apply to other countries, times or policies.

Selection bias: choosing an “interesting” case can produce conclusions that do not hold in typical cases.

Interpretation dependence: conclusions rely on how the researcher interprets documents and events.

Content analysis

Analysis of texts, e.g. speeches or media articles, to identify patterns in claims and language.

Source selection bias: the texts chosen may not be representative.

Context loss: meaning depends on the audience and situation, which can be missed in text-only analysis.

Platform effects: different media formats shape content, so patterns may reflect the platform more than the message.

Question banks

A prepared set of standardised questions used to collect comparable responses across many people.

Consistent wording makes responses easier to compare across respondents and reduces variation between researchers.

Recording and transcription tools

Audio/video recordings, fieldnotes and written transcripts that capture what participants said or what was observed.

A full record reduces memory bias and lets others check interpretations against what was actually said.

What gets captured can still be selective, transcripts can contain errors, and privacy/consent rules may restrict what can be recorded or shared.

Official datasets and records

Data collected by institutions, e.g. election results, budgets, census data and voting records.

Large-scale, repeated data collection supports trend analysis and comparisons across time and place.

Some data may be missing or underreported, and political incentives may affect what is measured or published.

Statistical tools

Techniques for summarising and analysing data, e.g. percentages, averages and correlations.

Quantifies patterns across large datasets and can express uncertainty as a quantity.

Averages can hide variation, correlations can be misread as causes, and outputs depend on the quality of the underlying data.

Models and simulations

Simplified representations that use assumptions and data to estimate outcomes or test “what if” scenarios.

Allows systematic testing of different scenarios.

Outputs can give false certainty, and results may fail when real-world behaviour changes or the model’s assumptions do not hold.