H567/02 Section B - Useful research and psychology as a science

What makes research useful?

Basic research investigates principles without requiring an immediate practical product. Applied research addresses a practical problem. Basic work can later support applications, so “no immediate treatment” does not mean useless. Conversely, a promising application is not established simply because a researcher describes a real problem.

Different positions give different weight to theoretical understanding, practical benefit and possible costs. One view prioritises findings that change everyday behaviour or policy. Another values carefully established knowledge even when its eventual uses are uncertain. A critical position asks who benefits, whether the effect generalises and whether implementation could harm people. These positions can complement one another: useful practice needs evidence, and practical problems can generate theoretical questions.

Judge usefulness through a chain: finding → justified inference → feasible application → outcome evidence. Check validity and reliability before generalising; then ask about cost, accessibility, duration of benefit and unintended effects. A statistically significant laboratory result is not automatically a worthwhile improvement outside the laboratory.

From evidence to an application

Grant et al. (1998) manipulated silent or noisy studying and testing of meaningful material. Recall and recognition were better when study and test contexts matched. For students facing a quiet examination, studying in a quiet setting is therefore a reasoned application of contextual cues. The result is not “noise always damages memory”: noisy study followed by noisy testing was also a matching condition. The short experiment did not establish the size or duration of benefit across all subjects, ages and revision programmes. OCR core-studies guide, pp21–25.

Chaney et al. (2004) tested a Funhaler with young children with asthma. Incentive toys linked to appropriate breathing made use more engaging; parents reported better adherence after the two-week Funhaler period than with their previous device. This connects a behavioural mechanism to a concrete health problem in home settings. However, the standard device preceded the Funhaler, and parent reports could reflect practice, novelty or expectations. A short pilot does not establish long-term clinical improvement. “Useful for adherence” and “proven to improve every health outcome” are different claims. OCR guide, pp31–33.

Compulsory mental-health research also illustrates conceptual usefulness. Rosenhan (1973) reported experiences of pseudopatients admitted after claiming auditory symptoms and then behaving normally. The account challenged how diagnostic labels influenced interpretation and treatment in the institutions studied. It can prompt scrutiny of assessment procedures and patient experience without proving that all present-day diagnoses are invalid. Its historical setting, deception and dependence on observations limit the conclusions. OCR mental-health guide, pp4–11.

These applications connect cognitive, developmental and mental-health evidence to component 01 design: specify the target outcome and compare a proposed change with an appropriate baseline. A new trial of quiet revision might measure delayed recall while controlling study time and materials; a change to a medical device requires suitable professional oversight. Research-informed proposals remain proposals until their actual outcomes are evaluated.

What does a scientific approach require?

A scientific approach uses systematic evidence to evaluate explanations and remains open to correction. Empirical evidence comes from observation or measurement. Objectivity means reducing the influence of personal expectations on collection and interpretation. Operational definitions specify how variables are measured or manipulated so claims can be tested.

Falsifiability means that an explanation makes claims which possible evidence could contradict. It is not the ability to prove a theory permanently true. A prediction that aggression will increase after a particular model can be challenged by appropriate comparison data. An account that reinterprets every possible response as supporting the same hidden motive is harder to challenge.

Replicability concerns whether others can repeat a sufficiently documented procedure. Replication results test how dependable an effect is across repetitions or changed conditions. Standard instructions help reproducibility; they do not guarantee that results will match. Reliable scoring means consistency, while validity concerns whether the measurement and inference capture what is intended. A reliably scored, culturally biased test can still be invalid.