University College London · FACULTY OF PSYCHOLOGY

PHDE0067 Chap.2 Psychology as a Science

- one subject, every graph, every model, every mark
7 Chapters4-page Bible
Our own words - no uploaded lecturer files
Updated for this semester
Chapter 2 of 9 · PHDE0067

Psychology as a Science

Psychology as a Science asks how evidence can support or challenge claims about behaviour and mental processes. Constructs become testable through operational definitions, but consistent measurement is not the same as measuring the intended idea. Correlation summarises association without settling direction or third-variable explanations.

Experiments strengthen causal inference through manipulation and control, although every conclusion remains tied to the sample, operation and setting. Observational and alternative methodologies can answer questions that experiments cannot address ethically or adequately. Descriptive statistics organise observed patterns, while inferential reasoning extends beyond them under assumptions.

The chapter trains a calibrated conclusion: identify the design, state what the evidence supports and explain precisely what would be needed for a stronger claim.

In this chapter

What this chapter covers

  • 01

    Philosophy of science and testable claims

  • 02

    Operational definitions

  • 03

    Reliability and construct validity

  • 04

    Descriptive and inferential evidence

  • 05

    Correlation and third variables

  • 06

    Experiments and causal inference

  • 07

    Observation, alternative methods and generalisability

Worked example · free

Evaluate a sleep and recall association

Q [5 marks]. Students report sleep duration and complete a recall task. Longer reported sleep is associated with more recalled words. State the strongest conclusion and one improvement. The practice weighting is not a published University assessment scheme; it partitions claim, limitation and redesign.
The ladder moves from a research question through design and observed pattern to a bounded conclusion; the lower note shows why alternative explanations limit cThe ladder moves from a research question through design and observed pattern to a bounded conclusion; the lower note shows why alternative explanations limit causal inference. question design pattern boundedconclusion measurement quality governs every stepalternative explanations narrow causal claims
The ladder moves from a research question through design and observed pattern to a bounded conclusion; the lower note shows why alternative explanations limit causal inference.
  • 1Describe the finding as an association in the measured sample and task.
  • 2Keep direction, reporting error and third-variable explanations open.
  • 2Propose a feasible controlled comparison with consistent measurement and ethical safeguards.
The study supports association rather than causation. A stronger design would improve control over competing explanations while keeping the conclusion limited to the chosen sleep-related condition and recall measure. The marks are an original practice weighting.
Sia tip — Write the design noun before the conclusion verb so the claim cannot outrun the method.
Glossary

Key terms

Operational definition
A procedure that specifies how a construct is measured or manipulated.
Third variable
An additional factor that may produce or alter an observed association.
Causal inference
A justified conclusion that changing one factor produced a difference in another.
FAQ

Psychology as a Science FAQ

Why does correlation not prove causation?

An association leaves direction unresolved and may reflect another variable influencing both measures. Causal interpretation needs stronger control over how the presumed cause changes and how competing explanations are excluded.

Can a reliable measure still be invalid?

Yes. A measure can produce consistent scores while systematically capturing the wrong construct. Reliability supports stability, whereas validity concerns the meaning attached to the scores and the evidence for that interpretation.

Are experiments always the best method?

Experiments are powerful for controlled causal contrasts, but manipulation may be unethical, artificial or irrelevant to questions about naturally occurring experience. The appropriate method depends on the question and the inference required.

How can a large study still support a narrow conclusion?

Sample size and representativeness answer different questions. More observations can reduce sampling uncertainty while the participants, setting or operational definition remain systematically restricted. Audit the population selected, the task used and the outcome measured before generalising. Then state the target population and identify which feature requires replication.

A direct replication tests the same critical method, whereas a conceptual replication changes the operation while preserving the proposed relation; each strengthens a different part of the evidence claim.

Study strategy

Exam move

Use a fixed evidence audit for every study. Write the research question, construct, operational definition, comparison, outcome and strongest supported conclusion. Add a column for one alternative explanation and another for the smallest design change that addresses it. Practise separating reliability from validity with examples that vary consistency and meaning independently.

For correlations, draw arrows in both directions and add a plausible third variable before writing any causal sentence. For experiments, identify the manipulated factor, outcome, control condition and remaining limit on generalisation. Compare a laboratory experiment with an observational or qualitative approach to the same broad topic and explain what each makes visible.

In timed MCQ work, underline absolute causal language and reject it when the design provides only association. In essay work, connect every criticism to the conclusion it weakens rather than listing generic problems. Finish each practice response with one calibrated sentence that names the sample, measure and boundary of the claim.

Extend the audit with a transfer column that names the population, setting and operation to which the conclusion is supposed to apply. A larger sample can reduce sampling uncertainty while remaining systematically narrow, so always distinguish size from selection. Create four miniature cases: reliable and valid, reliable but invalid, unreliable with a plausible target, and weak on both.

Explain which evidence would move each case. For generalisability, change one feature at a time and predict why the effect might strengthen, weaken or remain stable. A direct replication should preserve the critical method; a conceptual replication should change the operation while retaining the proposed relation. Compare their evidential contributions rather than calling both mere repetition.

During reading, mark every causal verb and locate the design feature that licenses it. If no feature exists, rewrite the sentence as association or prediction. Prepare one qualitative example where meaning or context is the central outcome, then state why experimental control is not the only scientific virtue. The final check asks whether an alternative method would answer the same question or a different one.

This prevents broad claims that one method is always superior and keeps evaluation tied to the research aim.

Working through Psychology as a Science in PHDE0067? Sia is AskSia’s AI Psychology tutor — ask any PHDE0067 Psychology as a Science question and get a clear, step-by-step explanation grounded in how PHDE0067 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.

A+Everything unlocked
Unlocks this Bible + all 32 of your subjects at University College London, plus 1,000+ more Bibles across the AskSia library.
Sia - your PHDE0067 tutor, unlimited, worked the way the exam marks it
The full 4-page Bible + practice bank with worked solutions
Chrome extension - sync your LMS so Sia knows your deadlines
Bilingual EN / Chinese on every Bible and every Sia answer
$0.99 Trial
30-day money-back · cancel in one tap · how it works