IB · PSYCH HL

Psychology HL

HL extensions and data interpretation (2027)

Name: ____________________Date: November 19, 2026
  1. 1.

    Hypothetical source comparison, not published research: A reports more public agreement in community X than Y; B interviews people in X who describe disagreement expressed privately as respectful. Evaluate a cultural explanation using perspective, identify a measurement problem and explain why these sources do not justify a national stereotype.

    [3 marks] · no calculator

    Syllabus placement evidence: Psychology updates, HL Extensions and syllabus component table (culture, motivation, technology; data analysis and interpretation). Prescribed HL extensions are core for HL, not optional beyond-course extension; integrates selected culture, motivation, technology and data interpretation tasks.

    Worked explanation

    1. If respectful disagreement is private, public agreement can conceal rather than measure agreement of belief. Qualitative accounts help reinterpret the behaviour counted in A.
    2. Check common operational definitions and sampling before comparing groups. Culture can help explain practices without making every individual behave alike.

    Common mistake: Treating a group average as the essential nature of all members commits an ecological overgeneralisation.

    Marking points

    • Uses B to distinguish culturally shaped expression from absence of disagreement.
    • Recognises that public agreement may not equivalently measure private belief across settings.
    • Considers within-community variation and sampling; neither source warrants attributing a fixed trait to all members.

    Examiner tip: Treating a group average as the essential nature of all members commits an ecological overgeneralisation.

  2. 2.

    Hypothetical motivation example: an optional puzzle activity is offered with controlling cash rewards in one group and choice plus feedback in another. Explain intrinsic versus extrinsic motivation, apply autonomy/competence and identify evidence needed before claiming rewards always reduce motivation.

    [3 marks] · no calculator

    Syllabus placement evidence: Psychology updates, HL Extensions and syllabus component table (culture, motivation, technology; data analysis and interpretation). Prescribed HL extensions are core for HL, not optional beyond-course extension; integrates selected culture, motivation, technology and data interpretation tasks.

    Worked explanation

    1. Cash can be an extrinsic reason to complete a task; interest is intrinsic. A controlling reward may be experienced differently from informational feedback, while choice may support autonomy.
    2. Measure subsequent optional engagement after the reward ends under a controlled comparison. One invented scenario cannot show that all rewards harm intrinsic motivation.

    Common mistake: Equating any external feedback with controlling reward ignores how participants interpret it.

    Marking points

    • Distinguishes enjoyment/interest in the task from an external reward as a reason to act.
    • Connects choice to autonomy and useful feedback to competence, without equating all reward with control.
    • Requires measures such as later voluntary engagement and information about reward framing, baseline interest and allocation; rejects an absolute universal claim.

    Examiner tip: Equating any external feedback with controlling reward ignores how participants interpret it.

  3. 3.

    Hypothetical technology evaluation: an adaptive app group reports higher study confidence than a paper-notes group; app users also receive more feedback and volunteer for the app. Evaluate the causal claim that the technology itself improves memory and propose a better outcome measure/design.

    [3 marks] · no calculator

    Syllabus placement evidence: Psychology updates, HL Extensions and syllabus component table (culture, motivation, technology; data analysis and interpretation). Prescribed HL extensions are core for HL, not optional beyond-course extension; integrates selected culture, motivation, technology and data interpretation tasks.

    Worked explanation

    1. App preference may reflect prior motivation, and feedback differs alongside medium. Neither is isolated by the group comparison.
    2. Use delayed performance on comparable content rather than confidence alone, matching feedback and study time. A benefit in one app design would not establish a universal technology effect.

    Common mistake: Substituting self-confidence for memory accuracy overstates the construct actually measured.

    Marking points

    • Identifies self-selection and additional feedback as competing explanations.
    • Distinguishes confidence from objectively measured memory performance.
    • Proposes a controlled allocation with matched feedback/time and an equivalent delayed recall or transfer measure.

    Examiner tip: Substituting self-confidence for memory accuracy overstates the construct actually measured.

  4. 4.

    Hypothetical mixed sources: A, app group mean recall 8/12 words (n=20) and paper group 7/12 (n=20), with no spread or allocation details; B, a fictional interview note says 'The reminders helped, but I checked messages too.' Calculate the mean difference, integrate B without overgeneralising it, and state two missing details needed for stronger interpretation.

    [3 marks] · no calculator

    Syllabus placement evidence: Psychology updates, HL Extensions and syllabus component table (culture, motivation, technology; data analysis and interpretation). Prescribed HL extensions are core for HL, not optional beyond-course extension; integrates selected culture, motivation, technology and data interpretation tasks.

    Worked explanation

    1. 8 - 7 = 1 word on the task; n=20 per group does not reveal the variance or uncertainty. The fictional interview suggests reminders and distraction may coexist.
    2. Ask for score distributions and allocation/baseline comparability. A single quoted account cannot tell how common either experience was, and these sources are invented for analysis, not empirical evidence.

    Common mistake: Inferring statistical significance from a mean gap and sample sizes alone ignores variance and design.

    Marking points

    • Mean difference is 1 correctly recalled word out of 12.
    • Uses B to suggest both support and distraction mechanisms, not a prevalence estimate or proof of mechanism.
    • Requests dispersion/uncertainty and assignment/baseline information before significance or causality claims.

    Examiner tip: Inferring statistical significance from a mean gap and sample sizes alone ignores variance and design.

  5. 5.

    Original hypothetical 2027 HL technology/data-interpretation exercise. Successful completion of a learning task is invented as follows: experienced learners using an app 8/10, using paper 27/30; novice learners using the app 12/30, using paper 5/10. Calculate each format's overall success rate and its rate within each experience group. Explain why the aggregate comparison could mislead and why these non-randomised data do not prove either format causes better learning.

    [4 marks] · no calculator

    Syllabus placement evidence: Psychology examiner instructions 2027: HL paper 3 - quantitative/qualitative source interpretation focused on culture, motivation or technology. Combines technology with quantitative source interpretation for the prescribed 2027 HL extension, which is core within HL and not an SL assessment requirement. Invented contingency counts are practice evidence only, not research findings or complete Paper 3 preparation.

    Worked explanation

    1. Sum successes and denominators separately: app (8 + 12)/(10 + 30) = 20/40; paper (27 + 5)/(30 + 10) = 32/40. Do not average subgroup percentages equally when their sizes differ.
    2. Experienced success is 80% versus 90%; novice success is 40% versus 50%. Paper has more experienced learners (30/40 rather than 10/40), inflating the aggregate gap relative to either within-group gap; there is no reversal of direction in these numbers.
    3. Stratification exposes one compositional influence but does not randomise format or control every difference. A causal conclusion would need a stronger design and a learning measure appropriate to the claim, such as delayed retention rather than completion alone.

    Common mistake: Calling every aggregate/subgroup discrepancy a direction reversal, or averaging percentages without weighting their denominators.

    Marking points

    • Calculates overall app success 20/40 = 50% and paper success 32/40 = 80%.
    • Calculates experienced app/paper rates 80%/90% and novice app/paper rates 40%/50%.
    • Explains that paper exceeds app by 10 percentage points within each subgroup but by 30 points overall because experience composition differs.
    • Explains that non-random selection and other confounds prevent a causal format claim; task completion is not automatically durable learning.

    Examiner tip: Calling every aggregate/subgroup discrepancy a direction reversal, or averaging percentages without weighting their denominators.

  6. 6.

    Original hypothetical 2027 HL technology/data-interpretation task, not a real study. Two non-randomised groups of 20 adults complete the same 12-word delayed-recall task. Supplied box-plot summaries are minimum, lower quartile, median, upper quartile, maximum: app group 1, 4, 7, 9, 12; paper group 2, 5, 7, 8, 11. Use these quartiles as supplied, without reconstructing individual scores. Calculate each interquartile range, compare the medians and middle-half dispersion, and evaluate the claims 'the means must be equal' and 'the app causes better learning'. Do not calculate a p-value from these summaries.

    [4 marks] · no calculator

    Syllabus placement evidence: Psychology guide, first assessment 2027: Table 3.4, Data analysis and interpretation; HL extensions, technology. Adds box-summary interpretation absent from the existing five-task HL evidence topic. This is current HL formative data interpretation, not an SL examination requirement, a real study or a statistical-significance result.

    Worked explanation

    1. Subtract lower from upper quartile, not maximum from minimum: 9-4=5 and 8-5=3. Both central medians are 7, so the supplied evidence distinguishes spread rather than median advantage.
    2. A five-number summary leaves many individual values unknown. A mean needs the sum of all 20 scores; the shared median cannot supply it. Neither an IQR gap nor n=20 alone supplies a justified inferential test result.
    3. Existing differences in experience or motivation could affect group outcomes. Even a better observed score on this specific delayed task would require careful design before attribution to format, and would not establish that every app improves all learning. No experimental or IA completion is claimed.

    Common mistake: Calling the maximum-minus-minimum the IQR, or inferring equal means from equal medians.

    Marking points

    • Calculates app IQR=9-4=5 words and paper IQR=8-5=3 words.
    • Identifies both medians as 7 words and the wider middle-half spread for app scores; equal medians do not mean identical distributions.
    • Rejects necessary equality of means: the summaries do not specify every score or each group's sum, so the means cannot be calculated from them.
    • Rejects the causal superiority claim: non-random allocation permits confounding, and these summaries do not establish a higher median, statistical significance or a general learning benefit.

    Examiner tip: Calling the maximum-minus-minimum the IQR, or inferring equal means from equal medians.