2026 Cambridge IAL Psychology (9990) 模擬試題連答案詳解

Demand characteristics refer to environmental, procedural, or experimenter cues in a study that give away the true purpose, aim, or hypothesis to participants (1 mark). Consequently, participants modify their spontaneous behaviour, whether by trying to please the researcher, acting unnaturally, or intentionally disrupting the results (1 mark).

1 mark for stating that it involves cues/features in an investigation that reveal the aim/expected outcome to participants.
1 mark for explaining the outcome/effect: participants altering their natural behaviour / acting unnaturally / attempting to conform to or subvert expectations.

Note: Do not credit simply 'participants know the aim' without reference to how this arises from cues or affects their behaviour.

Inter-rater reliability is the degree of agreement or consistency between two or more independent researchers/observers (1 mark) who are recording, rating, or categorising the exact same behaviour, event, or response independently (1 mark).

1 mark for stating the consistency/agreement between two or more independent observers/raters.
1 mark for stating that they observe/code/rate the same behaviour/data independently (e.g. comparing their recorded scores to check for a high positive correlation).

Accept 'inter-observer reliability'.

A directional hypothesis is a precise statement or prediction about the outcome of a study (1 mark) that clearly specifies the direction of the expected effect, difference, or relationship (e.g., stating which group will score higher, or that a positive correlation will occur) (1 mark).

1 mark for stating it is a testable prediction of a difference / relationship between variables.
1 mark for stating that it specifies the direction of the outcome (e.g. stating which group will perform better / higher / faster OR that variable X will increase as variable Y increases).

Do not accept non-directional definitions (e.g. 'predicts there will be a difference' without stating the direction = 0).

Standard deviation is a measure of dispersion/spread (1 mark) that shows the average amount of variation or distance of individual scores around the calculated mean (1 mark).

1 mark for identifying it as a measure of dispersion / spread / variation in a dataset.
1 mark for identifying that it quantifies how far scores/data points deviate or spread out around the mean / average score.

Reject descriptions of the range (e.g. 'difference between highest and lowest value' = 0).

(a) A fully operationalised directional hypothesis predicts the specific direction of the effect and clearly states both levels of the independent variable and the operationalised dependent variable.
Example: 'Participants completing the reading test under bright lighting (500 lux) will achieve higher scores on the 20-question comprehension test than participants under dim lighting (100 lux).'

(b) The mean is the measure of central tendency most affected by extreme outliers. This is because calculating the mean involves summing all scores and dividing by the total number of participants; therefore, an unusually high or low score pulls the value towards the extreme, making it unrepresentative of the typical dataset.

(a) [2 marks total]:
- 2 marks: Directional hypothesis with both levels of the IV operationalised (bright/500 lux vs dim/100 lux) and DV operationalised (score/correct answers on the 20-question test).
- 1 mark: Directional hypothesis that is incomplete or lacks full operationalisation of IV or DV.
- 0 marks: Non-directional/null hypothesis, or no prediction of difference.

(b) [2 marks total]:
- 1 mark: Identifies the mean.
- 1 mark: Explains why in context (e.g. it takes into account every individual value in the calculation / an extreme value skews or pulls the total sum disproportionately).

(a) A scatter graph (or scattergram / scatter plot) is the appropriate graphical representation for correlational data involving two continuous co-variables.

(b) As screen time increases (from 1 to 6.5 hours), sleep quality ratings systematically decrease (from 9 to 2). This demonstrates a negative correlation.

(c) Correlations establish a relationship or co-occurrence between variables but do not demonstrate cause-and-effect. Anika cannot manipulate the independent variable or eliminate confounding variables (e.g., academic anxiety, caffeine consumption), meaning a third factor could explain the link, or poor sleep quality could cause increased screen time (reverse causality).

(a) [1 mark]:
- 1 mark: Scatter graph / scattergram / scatter plot.
- 0 marks: Bar chart, histogram, pie chart.

(b) [1 mark]:
- 1 mark: Negative (correlation) / inverse.

(c) [2 marks total]:
- 1 mark: Identifies that correlation does not establish causality / lack of manipulation / possibility of a third confounding variable.
- 1 mark: Contextualises the explanation to the study (e.g. stress might lead to both using the phone more and sleeping worse; reverse causality where inability to sleep causes phone use).

(a) Inter-observer reliability is assessed by having both observers view and code the cafeteria behaviour at the same time using standardised behavioural categories. Their sets of recorded data are then compared (e.g., using a correlation coefficient); a high positive correlation (typically r >= 0.80) confirms high inter-observer reliability.

(b) A key strength of a covert observation is the elimination of participant reactivity. Because cafeteria patrons do not realise they are under observation, they do not alter their actions to appear more polite or helpful (avoiding social desirability bias), resulting in high ecological validity.

(a) [2 marks total]:
- 1 mark: Both observers independently observe the same session/behaviour using identical operationalised categories / tally sheets.
- 1 mark: Compare / correlate their records (e.g. check for a high degree of agreement / consistency / statistical correlation).

(b) [2 marks total]:
- 1 mark: Identifies reduced demand characteristics / social desirability / natural behaviour.
- 1 mark: Explicitly links this to the cafeteria context (e.g. students will not clear trays just because they feel watched, reflecting genuine everyday behaviour).

(a) A Likert rating scale produces quantitative data. This allows for straightforward numerical scoring, statistical analysis (such as calculating the mean or median stress rating), and direct comparisons between nurses working in different hospital wards or shifts.

(b) A suitable open-ended question gives participants the freedom to express their subjective experiences in their own words (e.g., 'What aspects of your daily duties at the hospital cause you the greatest amount of stress?'). This yields qualitative data because the responses consist of rich descriptive text providing depth and contextual insight into the reasons behind stress.

(a) [2 marks total]:
- 1 mark: Identifies a methodological strength of Likert rating scales (e.g. produces numerical/quantitative data, easy to analyse/compare, high replicability).
- 1 mark: Contextualises to measuring nurse/workplace stress (e.g. comparing stress ratings across hospital departments).

(b) [2 marks total]:
- 1 mark: Provides a clear, appropriate open-ended question relevant to nurse workplace stress.
- 1 mark: Explains that it generates qualitative data because responses are descriptive/textual/in the participant's own words rather than predefined numerical categories.

(a)
1. It contains a predetermined set of core questions/topics that are prepared in advance to guide the session.
2. The interviewer has the flexibility to ask follow-up questions, probe responses, or change the order of questions based on what the participant says.

(b)
- Similarity: Both structured and semi-structured interviews are self-report techniques involving direct face-to-face (or telephone/online) verbal interaction between the researcher and the interviewee to gather primary data about attitudes, beliefs, or behaviours.
- Difference: In a structured interview, every participant is asked the exact same questions in the exact same order with no deviation, providing high standardisation and replicability. In contrast, in a semi-structured interview, the interviewer can deviate from the set schedule to pursue novel responses and probe for elaboration, which provides greater depth and qualitative richness but lowers standardisation.

(a) 1 mark for each outlined feature (up to 2 marks maximum):
- Pre-set / predetermined list of core/fixed questions or guide/schedule (1);
- Flexibility to ask unplanned follow-up questions / probes / clarify answers based on responses (1);
- Questions can be omitted or their order changed according to the flow of conversation (1);
- Can generate a combination of qualitative and quantitative data (1).

(b) Similarity (2 marks):
- 1 mark for identifying a valid methodological similarity (e.g. both are self-report methods / both involve real-time direct verbal interaction / both gather primary data / both can ask open/closed questions);
- 1 mark for elaboration/justification.

Difference (2 marks):
- 1 mark for identifying a valid methodological difference (e.g. flexibility/standardisation, use of fixed vs probing questions, ease of replicability);
- 1 mark for elaboration/justification contrasting both methods.

Guidance Notes:
- For difference: Do not credit simply stating that one is structured and one is not without explaining what this means methodologically (e.g. identical standardised questions vs adaptable/probe questions).
- For similarity: Do not accept vague statements such as 'both are experiments' (0 marks).

### Model Plan: Field Experiment on Flexitime and Job Satisfaction

#### 1. Aim and Hypotheses
- Aim: To investigate whether office workers operating under a flexitime work schedule report significantly higher job satisfaction than office workers operating under a standard fixed-hours schedule.
- Directional (One-tailed) Hypothesis: Office workers on an 8-week flexitime schedule (allowing core hours of 10:00–14:00 with flexible start/finish times totalling 37.5 hours/week) will score significantly higher on a 20-item job satisfaction questionnaire (scored 20–100) than office workers remaining on a fixed 09:00–17:00 schedule.
- Null Hypothesis: There will be no significant difference in job satisfaction scores between office workers on a flexitime schedule and those on a fixed-hours schedule; any observed difference will be due to chance.

#### 2. Variables and Experimental Design
- Independent Variable (IV): Work schedule format, with two conditions:
- Condition A (Experimental): Flexitime arrangement (must complete 37.5 hours per week, core compulsory attendance 10:00–14:00, arrival flexible between 07:00 and 10:00, departure flexible between 14:00 and 19:00).
- Condition B (Control): Fixed schedule (must work standard 09:00–17:00 with a 30-minute lunch break, totalling 37.5 hours per week).
- Dependent Variable (DV): Operationalised job satisfaction score measured via an adapted Job Descriptive Index (JDI) consisting of 20 Likert-scale items (1 = Strongly Disagree to 5 = Strongly Agree, total score 20–100, where higher scores reflect greater job satisfaction).
- Design: Independent measures design. Two matched departments of comparable task difficulty and compensation within the same tech firm are randomly allocated to either Condition A or Condition B to prevent contamination and order effects.

#### 3. Sample and Setting
- Setting: Natural office working environment of a mid-sized software company across two distinct floors/departments.
- Sample: \(N = 60\) full-time software support specialists (30 per condition, matched for age, gender distribution, and average tenure).
- Sampling Technique: Opportunity sampling within the partner company, screening out any staff currently on probationary periods or part-time contracts.

#### 4. Controls and Standardisation
- Both conditions maintain the exact same total contracted weekly hours (37.5 hours) and base salary.
- Both groups perform identical core duties and use the same ticket-handling software.
- Baseline job satisfaction is measured one week prior to the study to confirm no pre-existing satisfaction differences between groups.
- The intervention runs for a continuous period of 8 weeks to allow the novelty effect of flexitime to settle.

#### 5. Ethical Considerations
- Informed Consent: Participants are provided with full information sheets detailing that work schedule patterns are being researched, guaranteeing that participating or declining will have zero impact on performance reviews or job security.
- Right to Withdraw: Workers can opt out at any time without penalty.
- Confidentiality: All questionnaire data are pseudonymised with unique numeric participant IDs; individual responses are never shared with line managers or executive leadership.
- Debriefing: A full debrief sheet is issued at week 8 explaining the exact comparative aims of the study.

#### 6. Data Analysis
- Quantitative data (total satisfaction score per participant, 20–100) will be collated.
- Descriptive Statistics: Calculate the mean and standard deviation (or median and range) of job satisfaction scores for the flexitime group versus the fixed-hours group.
- Visual Representation: Construct a bar chart showing the mean job satisfaction score for Condition A compared with Condition B, complete with clearly labelled axes.

Generic Marking Grid for Paper 4 Section B (Plan a study – 10 marks total, AO2 Application):

- Level 4 (8–10 marks):
- The response is a well-developed, clear, and comprehensive plan tailored specifically to the scenario.
- Contains detailed operationalisation of both general research features (aim/hypotheses, sample and sampling technique, ethical safeguards, descriptive statistical analysis) and method-specific features for a field experiment (clear IV with 2+ operational levels, precise operationalisation of DV, experimental design, controls, field setting).
- The procedure is described with sufficient precision and coherence to allow direct replication by another researcher.

- Level 3 (5–7 marks):
- The plan is appropriate and mostly complete, with good application to the scenario.
- Covers key features (IV, DV, sample, design, procedure) but may lack detail or precision in one or two areas (e.g. vague sampling technique, incomplete control of extraneous variables, or generalised ethical considerations).
- Replicable in most respects with minor ambiguities.

- Level 2 (3–4 marks):
- The response shows basic planning relevant to the topic but has substantial omissions or lacks operational detail.
- IV or DV may be identified but not clearly operationalised; procedure is outlined in broad terms with limited experimental control.
- Difficult to replicate without significant assumptions.

- Level 1 (1–2 marks):
- Fragmented or very brief plan with minimal application to the scenario.
- Major methodological elements are absent, inaccurate, or confused.

- Level 0 (0 marks):
- No creditworthy response.

1 mark for identifying a relevant feature of validity (e.g. control over extraneous/confounding variables / high internal validity).
1 mark for explaining this strength in the context of Tamsin's study (e.g. ensuring recall performance is caused by the music condition rather than environmental distractions).

1 mark: Identification of a valid strength related to validity (e.g. control of extraneous variables / establishing cause and effect / standardisation).
1 mark: Explanation applied to the scenario (e.g. ensuring factual text recall differences between the jazz music and silent conditions are caused by the IV rather than extraneous noise or lighting differences).

Creditworthy points:
- High level of control over extraneous variables (1 mark) means Tamsin can ensure all participants read under identical lighting and noise conditions, increasing internal validity (1 mark).
- Standardisation of the study material/test (1 mark) ensures every participant faces the exact same 10-item factual questions, ensuring the test accurately measures memory recall across both conditions (1 mark).

Do not credit general definitions of validity that are unlinked to the scenario for the second mark.

1 mark for identifying an aspect of ecological validity (e.g. natural environment / participants behaving naturally / low demand characteristics).
1 mark for contextualised explanation showing how this applies to helping behaviour or the train station setting.

1 mark: Identification of a valid point about ecological validity / natural setting / realism.
1 mark: Explanation linked directly to the context of Marcus's study (dropping papers / bystander helping in a train station).

Creditworthy points:
- The study is conducted in a real-life public setting (train station concourse) (1 mark), meaning the situation of someone dropping papers is a familiar, authentic event so behaviour can be generalised to real-world helping situations (1 mark).
- Commuters/bystanders are unaware they are part of a psychological investigation (1 mark), so they are unlikely to show demand characteristics or altered social desirability when deciding whether to help the person in formal or torn clothes (1 mark).

Do not credit general references to 'more valid' without explanation of why/how ecological validity is achieved.