In one line: A finding is only as trustworthy as the controls for bias behind it.
A music-streaming company wants to show that background music helps people concentrate. Its own researcher recruits subscribers, tells them what the study is about, and scores the concentration task herself.
The finding, that music helps, may even be true. As the study stands, nobody can tell.
Two aims: Objectivity: in research that counts and measures, controls keep the researcher's influence out.
Credibility: in research that interviews and observes, the aim is a true picture of what participants think and do.
Controls for bias belong to every stage of research, from the plan to the publication. None removes bias completely, so a sound report also states what bias may remain.
Remember: For each bias, a control. For what is left, honesty.
Free preview
This is the free notes preview
You're reading the free notes. Aimnova Pro unlocks the full study experience — and you can try it with your first topic free to keep:
- FlashcardsLock in vocabulary and key terms with spaced repetition.
- Practice questionsAnswer exam-style questions and get instant AI marking.
- Mock exams & past-paper vaultSit full mocks and see exactly how examiners award marks.
- Personalised study planA daily plan built around your exam date and weak areas.
Key idea: The design decides how much bias a study starts with.
Three decisions made in advance
A research protocol
A written plan sets out exactly how the data will be collected and analysed. Every participant hears the same instructions and has the same time.
Here: the same task, the same volume and the same ten minutes for everyone.
A pilot study
A trial run on a few people tests the task, the instructions and the measures. It can show that something uncontrolled (a confounding variable) has been missed.
It costs time and money, so it is not always possible.
Sampling and allocation
A random sample is rarely possible, and volunteers are common. Volunteers can still be put into the conditions by chance (random allocation), which spreads their differences evenly between the groups.
Protocol · Pilot · Sample
In the music study, the subscribers who volunteer probably enjoy working with music. Random allocation cannot change who volunteered, so the result applies only to people like them. It does make the music group and the silence group alike.
Exam tip: Random sampling protects how far a finding can be generalised.
Random allocation protects the comparison between groups.
See how examiners mark answers
Answer exam-style questions with model answers. Learn exactly what earns marks and what doesn't.
Key idea: An expectation cannot act on information that has been withheld.
Single-blind
- Participants do not know the aim or their condition
- Reduces demand characteristics
- The researcher still knows
- Here: the task is presented as a proofreading study
Double-blind
- The person collecting the data does not know either
- Also reduces experimenter bias
- Scoring is done without knowing the condition
- Here: an assistant scores unlabelled scripts
A researcher who expects music to help may give warmer instructions to the music group without noticing. This influence of the researcher's expectations on the results is experimenter bias, also called the Rosenthal effect.
Hiding the aim is a mild deception, so it has to be justified and followed by a full debriefing.
When blinding is impossible: In an interview, neither side can be blind.
The controls are neutral, open questions, a trained interviewer, and answers given in private. In an observation, people can be given time to get used to the observer.
Key idea: The researcher is part of the study, so the researcher is checked too.
After the task, the company's researcher interviews twelve participants about how the music felt. She loves music and plays in a band.
Four checks on interpretation
Reflexivity
The researcher's awareness of her own views and values, and of how they may influence the research.
Personal reflexivity: she notes in a journal that she expects music to help. Epistemological reflexivity: she asks whether interviews run by the company can answer the question at all.
Triangulation
More than one researcher, method, kind of data or theory is used on the same question.
Here: a colleague codes the transcripts independently (researcher triangulation), which guards against confirmation bias.
Member checking
Participants read the findings and say whether they are a true picture of what they meant.
Openness
The company's interest is declared, the plan is registered in advance, and the results are published even if music made no difference.
This reduces publication bias.
Reflexivity · Triangulation · Member checking · Openness
Remember: These checks build credibility: confidence that the findings show what the participants think, not what the researcher hoped.
Study smarter, not longer
Most students waste 40% of study time on topics they already know. Our AI tracks your progress and optimizes every minute.
Key idea: A control is only useful against the bias it was built for.
The bias
- Sampling bias
- Non-response bias
- Demand characteristics
- Experimenter bias
- Confirmation bias
- Leading questions
- Social desirability
- Publication bias
A control that fits
- Random sampling; varied times and places
- A pilot; easy ways to reply
- Single-blind design
- Double-blind design; a fixed protocol
- Researcher triangulation; reflexivity
- A pilot; neutral, open wording
- Anonymous answers given in private
- Pre-registration; publishing null results
Every control has a cost. A pilot takes time. Blinding may need deception. A random sample may be out of reach. Discussing how to avoid bias means weighing each control against what it costs.
Exam tip: "Discuss" asks for a considered review.
Say what the researcher did well, what could be added, and what limits each suggestion.
How this is asked: On Paper 3, one 6-mark question is about a study described in a source. It asks how the researcher could avoid bias, how to improve credibility, or how far the findings transfer.
This page covers the first.
The pattern for 6 marks
- Pick three or four biases that this study invites.
- For each one, give the detail in the source that lets it in.
- Give the control that fits, and say why it works.
- Credit what the researcher already did well.
- Note a limit of your suggestions.
The trap: a general list: Controls that could be written about any study stay in the lower bands.
The highest band asks for explicit evidence from the source.
Discuss how the researcher could avoid bias in the source below.
Source: A researcher who plays in a band carried out semi-structured interviews with twelve students from the school where he teaches music. He wanted to understand how listening to music affects their studying. He asked questions such as "How does music help you to concentrate?" and took notes by hand. He analysed the notes himself and reported three themes, all of them positive.
Model answer plan
See the answer broken into clear points, with guidance on how each point earns marks.