Six marks on the quality of a piece of research. Question 3 comes in three flavours — credibility, bias or transferability — and the source is usually qualitative. It is the most technical question on the paper and, once you know the vocabulary, one of the most predictable.
📘 What you need to know
Q3 is worth 6 marks and evaluates one source, usually a piece of qualitative research.
Three possible versions: improve the credibility, avoid the bias, or judge the transferability.
The structure is the same each time: identify the issue, explain why it is a limitation, suggest an improvement, and link back to validity or trustworthiness.
The command term is usually discuss, which means a considered review: limitations, suggestions and some balance.
Credibility tools: triangulation, reflexivity, member checking, audit trail, peer or expert review.
Transferability answers must argue both ways and then judge.
Which version have you been given?
Read the command carefully. “Improve the credibility” and “avoid the bias” want different answers even about the same study.
The credibility toolkit
Three of these, each explained and tied to the source, is a comfortable six-mark answer.
Reflexivity is the one students define wrongly. It does not mean being neutral. It means writing down what you already believe, so that other people can see where your interpretation might be coming from.
If the question is about bias
Name the specific bias, say where it enters this study, and give a fix. The usual suspects:
Culture bias — broaden the sample beyond one country or ethnic group.
Gender bias — include the missing gender, and say what new information that would give.
Confirmation bias — reflexivity, plus a second coder who does not know the hypothesis.
Social desirability and acquiescence bias — check the wording of questions, mix positively and negatively worded items, and keep responses anonymous.
Sampling bias — move away from opportunity or self-selected samples where possible.
If the question is about transferability
Transferability is not the same as generalisability from a big sample. It asks how far a reader could reasonably apply these findings to another setting. A strong answer does three things: limits the claim, then argues for what does transfer, then judges.
The transferability move
The specific findings should not be generalised because [sample limits]. The underlying idea does transfer because [it is about a situation, not these people]. So: limited for the numbers, reasonable for the principle.
Worked examples
WORKED EXAMPLE
Discuss how the researcher could improve the credibility of the findings in Source 3. [6 marks]
The source is a qualitative study in which one researcher collected and interpreted participants’ retellings of a story.
Identify the issue
Credibility is limited because a single researcher collected and interpreted all the data, so the findings rest entirely on one person’s reading of them.
Improvement 1: triangulation
Method triangulation would add a second route to the same data, for example a short questionnaire alongside the retellings. Researcher triangulation would bring in other experts to code the material, so agreement between them can be checked.
Improvement 2: reflexivity
Recording his own expectations before analysis would keep subjectivity visible, since he would otherwise be well placed to notice only the changes that fit what he predicted.
Improvement 3: audit trail
Keeping precise records of every decision about collection and analysis lets others follow how conclusions were reached.
Balance, as “discuss” requires
None of this makes qualitative interpretation objective; it makes it transparent and checkable, which is what credibility means here.
Three techniques, each explained and appliedMember checking would also work here. Three well-explained points is plenty for six marks.
WORKED EXAMPLE
Discuss how the researcher could avoid bias in a study of 80 mothers and their children in one country. [6 marks]
Culture bias
The sample comes from one country, so the findings reflect one cultural setting. Broadening it to include families from other countries, and a wider ethnic and socio-economic range within the same country, would make it more representative.
Gender bias
Only mothers took part. Including fathers would show whether parents differ in how they see the issue, rather than assuming one parent’s view stands for both.
Confirmation bias
The researcher should practise reflexivity before, during and after the study, so that she is aware of what she expects to find and does not report only the data that confirms it.
Round it off
Together these steps do not remove bias, but they make it visible and limit its influence on the conclusions.
Three named biases, three concrete fixesName the bias precisely. “The sample was limited” is much weaker than “culture bias”.
💡 Exam tip
Learn the five credibility techniques by name. They apply to almost any qualitative source.
Always attach the fix to the study: “in this study, member checking would mean…”.
“Discuss” wants balance, so end by admitting what the improvements cannot solve.
Three developed points is the right size for six marks.
Give this question about 15 minutes.
⚠ Common mix-up
Confusing credibility with reliability. Credibility is about trustworthy interpretation in qualitative work; reliability is about consistency of measurement.
Listing techniques with no explanation. Each one needs to be tied to the source.
Saying “get a bigger sample” as the whole answer. Qualitative research is not trying to be representative in that way.
Only criticising in a transferability question. You must argue what does transfer as well.
Thinking reflexivity removes bias. It exposes it.
Up next: Paper 3, Question 4: The Extended Response — fifteen marks, three sources and a claim to weigh up.
Want this explained one-to-one?
Book a free session with an experienced IB Psychology tutor and get your trickiest topics made simple.
