IB Psychology IA

The IB Psychology Internal Assessment is a replication study: you choose an existing published psychological study, adapt it for an ethical school setting, collect data from a minimum of ten participants per condition, and write a formal report in the style of a scientific paper. It is worth 25% of your final grade at SL and 20% at HL. Unlike other IB IAs, Psychology explicitly requires you to replicate an existing study rather than design a completely original experiment — and this constraint is both an advantage and a challenge.

The Four Sections of the IB Psychology IA Report

The Psychology IA has a specific required structure that differs from other Group 3 IAs. The four sections and their approximate word allocations are:

• Introduction (approx. 200–300 words): Background to the original study, explanation of the research question and aim, and a clear directional hypothesis (one-tailed or two-tailed).
• Exploration (approx. 200–350 words): Design of your study — independent variable, dependent variable, controls, sampling method, ethical considerations (including informed consent and right to withdraw).
• Analysis (approx. 150–250 words): Descriptive statistics (means, standard deviations) presented in a table, an appropriate graph, and the result of an inferential test.
• Evaluation (approx. 200–350 words): Comparison of your results to the original study, discussion of limitations, and suggested improvements.

Total word limit: 1500–2200 words (not including references, raw data in appendix, or participant consent forms). Staying within the word limit is itself assessed under Communication.

Choosing the Right Study to Replicate

The choice of original study determines how straightforward your IA will be. A good replication study for the IB Psychology IA has four properties:

1. Simple, operationalisable IV and DV: The independent variable must be something you can manipulate in a school setting (two word lists, two images, two conditions) and the dependent variable must be measurable numerically (number of words recalled, response time in seconds, rating on a scale).
2. Ethical with secondary school participants: Studies involving deception are acceptable if debriefing is provided, but anything involving stress induction, deception about a participant's health or performance, or sensitive topics (eating, depression, self-esteem) requires especially careful ethical planning.
3. Accessible to replicate with ≥10 participants per condition: A study requiring clinical populations, specific age groups you cannot access, or expensive equipment is unsuitable.
4. An expected result that is testable with a Mann-Whitney U test or Wilcoxon signed-rank test: These are the two inferential tests specified in the IB Psychology course. Your study design must produce data suited to one of them.

IB Psychology IA Study Examples That Work Well

1. Replication of Loftus and Palmer (1974) — leading questions and eyewitness memory

Show participants a short video clip of an everyday event (not a car crash — less distressing). Vary the verb in a critical question ("How fast was the car going when it contacted/smashed into the other vehicle?"). Compare estimated speeds between the two conditions. This is a between-subjects design; use the Mann-Whitney U test. Well understood, easy to run, clear results.

2. Replication of Stroop (1935) — cognitive interference

Create two word lists: colour words printed in a congruent colour (RED in red) and the same words printed in an incongruent colour (RED in blue). Measure time to name the ink colour for each list. This is a within-subjects design; use the Wilcoxon signed-rank test. Response time is the DV — measurable with a stopwatch, no specialist equipment needed.

3. Replication of Miller (1956) — the magical number 7 and working memory capacity

Give participants a sequence of random digits (lists of varying length: 5, 7, 9, or 11 digits) to recall immediately. Measure accuracy by list length. A simpler version: compare recall accuracy for a 7-item list vs a 11-item list (between subjects). This directly tests Miller's claim about working memory capacity limits. Easy to standardise, clear ethical picture.

4. Replication of Craik and Lockhart (1972) — levels of processing

Present participants with a word list under two encoding conditions: one group answers shallow questions about each word (is it in capital letters?), the other answers deep questions (does it fit in the sentence "I saw a ___ in the park"?). Measure free recall after a distractor task. Between-subjects design; Mann-Whitney U test. This directly tests the levels of processing effect.

The Statistical Test: Mann-Whitney U vs Wilcoxon

IB Psychology requires one of two non-parametric tests. Choosing correctly is assessed under Analysis:

• Mann-Whitney U test: Use when you have two separate groups of participants (between-subjects / independent measures design). Your two groups are independent — no participant appears in both conditions.
• Wilcoxon signed-rank test: Use when the same participants take part in both conditions (within-subjects / repeated measures design). Each participant's two scores are paired.

Calculate the test statistic by hand (the IB requires this) and compare to the critical value table at p ≤ 0.05. State whether you reject or retain the null hypothesis. Do not interpret a non-significant result as "the study failed" — a non-significant result still has things to say about the original study, your methodology, and the population you sampled.

Ethical Considerations: More Than a Checkbox

The Psychology IA requires explicit ethical planning — informed consent, right to withdraw, confidentiality, and debriefing. These are not formalities: examiners check whether your ethical procedures were actually appropriate for the study you ran. Common problems:

• Informed consent forms that describe the study so specifically that they prime participants — defeating the purpose of any deception. Solution: describe the general topic (a study about memory) without revealing the specific hypothesis.
• Studies run on friends or classmates without a genuine right to withdraw — social pressure makes "you can leave at any time" meaningless. Note this in Evaluation as a limitation.
• No debriefing documented. Debriefing must explain what the study was actually testing and why any deception (if used) was necessary.

Evaluation: The Most Commonly Under-Scored Section

Psychology IA Evaluation requires you to do three things, and most students only do one:

1. Compare your results to the original study: Did you replicate the finding? If not, why might the results differ? Consider participant demographics, cultural context, the specific stimuli used.
2. Identify and explain specific limitations of your design: Convenience sampling (you used your classmates — how might this affect generalisability?), demand characteristics, social desirability, order effects if within-subjects.
3. Propose concrete improvements: Not "use more participants" without justification — specify how many, why that number, and what statistical power that would give you. A specific suggested modification to reduce a named confound is much stronger than a general statement.

Before Your Supervisor Reads Your Final Draft

Psychology IA marks are largely determined before the moderator sees the work — your supervisor's assessment is moderated but is usually close to the final mark. Getting criterion-by-criterion feedback before your supervisor finalises their mark is the highest-leverage intervention available to you.

IBLens analyses your Psychology IA report against the official IB marking criteria and tells you exactly what each criterion needs to reach the next mark band — before it is too late to change anything.

Upload your Psychology IA draft to IBLens for rubric-based feedback →