- Which Psychology IA Applies to You?
- What the New Psychology IA Asks You to Do
- The Four IA Research Methods and When Each Fits
- Choosing a Real-Life Problem and Population of Interest
- Psychology IA Criteria: What Each 6-Mark Criterion Needs
- Procedure, Sampling and Design Decisions
- Building Your Data Collection Tool
- Ethics Must Be Linked to Your Participants
- Writing the Discussion Without Results
- Do You Still Need Statistics?
- Common Psychology IA Mistakes
- The Previous Experimental IA (Exams in 2026)
Quick Answer
From May 2027, the IB Psychology IA is a research proposal, not an experiment. You design, but do not carry out, a study of a specific population of interest using one of four methods: experiment, interview, observation or survey/questionnaire. The limit is 2,200 words, there are four criteria worth 6 marks each (24 in total), and it counts for 30% at SL and 20% at HL.
What You'll Learn
- Exams from May 2027 (students who started in August 2025 or later): a research proposal marked out of 24
- You choose one of the four methods you practised in class practicals: experiment, interview, observation or survey/questionnaire
- You collect no data, so there is no results or statistics section; you discuss what the study could find instead
- Every choice must be justified for your specific population of interest and real-life problem
- Exams in 2026 used the previous experimental IA (a replication marked out of 22)
Which Psychology IA Applies to You?
The IB introduced a new Psychology course for first teaching in August 2025, with first exams in May 2027. The IA changed completely, so check which version you are on before you follow any guide, video or exemplar.
- Exams in May 2027 or later: the new research proposal described on this page. You design a study but do not run it
- Exams in 2026 (May or November): the previous guide's experimental IA. You replicated a simple published experiment, collected data and analysed it with statistics
- Most exemplars, ideas lists and "how to get a 7" videos made before 2025 describe the old experiment. Their advice on data analysis and inferential tests does not apply to the new task
Watch Out
If your teacher talks about a research proposal, class practicals or a population of interest, you are on the new guide.
What the New Psychology IA Asks You to Do
The IA asks you to write a research proposal: a detailed, justified plan for a study that would investigate a real-life problem affecting a specific population of interest. You use one of the four research methods you practised in your class practicals, and you explain why it suits the problem and the population. Because you never collect data, the marks reward the quality of your reasoning: why this aim, why this method, why this sample, why these questions or measures, and what the findings could mean. SL and HL students complete exactly the same task.
- Word limit: 2,200 words, written individually
- Marks: 24, from four criteria worth 6 marks each
- Weighting: 30% of your final grade at SL, 20% at HL
- Time: the IB recommends 20 hours of class time
- Marked by your teacher and moderated by the IB
- Supported by class practicals: the course includes at least four, one in each context of study, and these give you hands-on experience of the methods
Pro Tip
Think of the proposal as something you could hand to a researcher who would then run the study. If they would need to ask you how, who, where or why, your proposal is not detailed enough.
The Four IA Research Methods and When Each Fits
Your proposal must use one of the four research methods from the class practicals. The best choice is the one that fits your aim and population, not the one that sounds most scientific. Moderators reward a method you can justify convincingly over a method that is simply familiar.
- Experiment: you manipulate an independent variable and measure its effect on a dependent variable. Best when your aim is about cause and effect, for example whether a short mindfulness exercise changes test anxiety scores in a specific group. You must operationalise both variables and choose an experimental design such as independent measures or repeated measures
- Interview (including focus groups): best for exploring experiences, meanings and perceptions in depth, for example how first-generation university students describe their sense of belonging. You design an interview schedule and decide between structured, semi-structured or unstructured formats
- Observation: best when you want to record what people actually do rather than what they say, for example how young children share toys in a nursery setting. You decide on covert or overt, participant or non-participant, and naturalistic or controlled, and you build behavioural categories
- Survey or questionnaire: best for collecting views or self-reported behaviour from a larger sample, for example screen-time habits and sleep quality among Year 12 students. You design the items, choose closed or open questions and rating scales, and plan how to avoid leading wording
Pro Tip
Ask what kind of answer your aim needs. A "does X cause Y" aim points to an experiment, a "how do people experience X" aim points to an interview, and a "how common is X" aim points to a survey.
Choosing a Real-Life Problem and Population of Interest
Your starting point is a real-life problem, and your population of interest is the specific group affected by it. Vague populations such as "teenagers" or "students" make it hard to justify anything that follows, because sampling, ethics and the data collection tool all depend on who you are studying. A focused population lets you explain why this group matters, what is already known about them and why your method is appropriate for them.
- Weak: "stress in teenagers". Stronger: "exam stress among 16 to 18-year-old boarding students in their final year"
- Weak: "social media and mood". Stronger: "the effect of passive social media scrolling on mood in university students who live away from home"
- Pick a problem with published research you can use: the introduction needs the findings and conclusions of two relevant pieces of research
- Make sure the population is one you could realistically reach in principle, even though you will not actually run the study
- Choose a topic that links clearly to course content, such as health and well-being, human development, human relationships, or learning and cognition
Psychology IA Criteria: What Each 6-Mark Criterion Needs
The four criteria carry equal marks, so a strong proposal gives each one proper attention rather than spending half the word count on the introduction. The summary below paraphrases what each criterion rewards. Your teacher can share the full descriptors from the guide.
- 1Criterion A, Introduction (6 marks): a clear, focused aim; a real-life problem, with its impact on your population of interest explained; and the findings and conclusions of two pieces of relevant research that justify why the study is worth doing
- 2Criterion B, Research methodology (6 marks): why your chosen method suits the aim and population; the procedure, including sampling, design and setting; and ethical considerations that are explicitly linked to your study and participants
- 3Criterion C, Data collection (6 marks): the tool you created to measure behaviour, such as an interview schedule, observation checklist, questionnaire or experimental materials; the decisions behind it; and the challenges you would expect in collecting data
- 4Criterion D, Discussion (6 marks): what the study could find and what those findings could mean for policy or practice; how researcher bias could affect the study; and one other research method, with what it would add
Procedure, Sampling and Design Decisions
Criterion B rewards justified choices. Every decision in your procedure should be explained in terms of your aim and population, not just stated. A moderator should finish this section knowing exactly who would take part, how they would be recruited, where the study would happen and what each participant would experience.
- Sampling technique: random, stratified, opportunity, purposive, snowball or self-selected, with a reason it fits your population and its trade-offs
- Sample size: a realistic number with a reason, for example a small purposive sample for in-depth interviews or a larger sample for a questionnaire
- Design: for experiments, independent measures, repeated measures or matched pairs, with how you would handle order effects or participant variables
- Setting: where the study would take place and why, including how the setting could affect behaviour
- Standardisation: scripted instructions and consistent conditions, so another researcher could repeat the procedure
- Controls: what could distort the results (demand characteristics, researcher effects, environmental distractions) and how your design reduces it
Watch Out
A procedure that sounds clear on paper but leaves out who reads the instructions, how long each stage takes or what happens if a participant withdraws will look underdeveloped to a moderator.
Building Your Data Collection Tool
Criterion C asks you to create a tool, so this is where your proposal becomes concrete. The tool must match your method and actually measure the behaviour or experience in your aim. Explain your design decisions: why each question, category or measure is there, how it operationalises the variables or themes in your aim, and what problems you expect in using it.
- Interview schedule: open questions in a logical order, with planned follow-up prompts, and no leading wording
- Observation checklist: clear behavioural categories that two observers would code the same way, with a sampling method such as event or time sampling
- Questionnaire: items linked to your aim, consistent rating scales, a mix of question types where useful, and checks for social desirability
- Experimental materials: the stimulus, task and measure of the dependent variable, with exactly how it would be scored
- Challenges: for example participants giving socially desirable answers, observer drift, unclear wording or low response rates, and how you would reduce each
Ethics Must Be Linked to Your Participants
Ethics is assessed as part of the methodology, and generic lists score poorly. Explain how each principle applies to your population and procedure. The more vulnerable your population, the more detail moderators will expect.
- Informed consent: who gives it, how it is explained, and whether parental or guardian consent is needed for younger participants
- Right to withdraw: how participants can leave at any point, and what happens to their data
- Protection from harm: how you would avoid distress, especially on sensitive topics such as anxiety or body image, and where participants could get support
- Confidentiality and anonymity: how data would be stored and reported
- Debriefing: what participants would be told afterwards, especially if any detail was withheld
- Trade-offs: where an ethical safeguard could affect validity, and why you accept that
Writing the Discussion Without Results
Because you collect no data, criterion D asks you to think ahead. Describe the findings your study could produce, ideally more than one possible outcome, and explain what each would mean for the real-life problem, for example for a school policy, a health service or a parent support programme. Then discuss how researcher bias, such as confirmation bias in interpreting interviews or expectations during observation, could affect the study and how you would limit it. Finally, propose one other research method and explain what it would add, for example following a questionnaire with interviews to explain why patterns occur.
Do You Still Need Statistics?
Not in the new IA. There are no results to analyse, so there is no descriptive or inferential statistics section in the research proposal. Data analysis and interpretation are still part of the course: HL students are assessed on them directly in Paper 3, which uses quantitative and qualitative research findings. SL and HL students are also expected to think critically about data in research and media claims. If you were on the previous guide, statistics were central: the experimental IA required descriptive statistics, an inferential test and a justified interpretation of the result.
Common Psychology IA Mistakes
These problems weaken research proposals most often, and several come from following advice written for the old experiment. An IB Psychology tutor can explain, using published studies rather than your proposal, how a method should fit an aim and how to justify each design choice, while the proposal stays your own work.
- Treating the task as a replication of a classic experiment, or reporting invented results
- A population of interest so broad that no choice can be justified
- Choosing a method because it is familiar rather than because it fits the aim
- Background research that summarises studies without linking their findings to your aim
- Listing ethical principles without applying them to your participants
- A data collection tool that does not measure what the aim describes
- A discussion that predicts one outcome and ignores what other findings would mean
- Suggesting another method without explaining what it would add
The Previous Experimental IA (Exams in 2026)
For reference, students sitting exams in May or November 2026 followed the previous guide. The IA was a report on the replication of a simple published experiment in which you manipulated one independent variable, collected data from participants and analysed it. Reports were 1,800 to 2,200 words, marked out of 22 across four criteria: introduction (6 marks), exploration (4 marks), analysis (6 marks) and evaluation (6 marks). The IA was worth 25% at SL and 20% at HL. None of this applies to students sitting exams from May 2027.