Features of Science [AL]

Features of Science: A-Level Depth

This benchmark extends the foundational treatment of scientific features (B1492) to cover the deeper philosophical and historical analyses required at A-Level: Popper's hypothetico-deductive method and the problem of induction, Kuhn's paradigm framework in greater detail, the replication crisis as a contemporary challenge to psychology's scientific status, and the characteristics of good scientific reporting. These frameworks do not merely describe science from the outside but illuminate the logical structure of scientific reasoning and the social processes through which scientific knowledge is constructed and revised.

Popper and the Hypothetico-Deductive Method

Karl Popper (1959) proposed that science proceeds via the hypothetico-deductive method: scientists formulate bold hypotheses and then attempt to falsify them through rigorous testing. If a test fails to falsify a hypothesis — if the predicted outcome is observed — this provides support for the hypothesis but does not prove it. No number of confirmations can prove a general scientific law, because a single disconfirming instance is sufficient to falsify it. This is Hume's problem of induction: logical reasoning from particular observations to general laws is never deductively valid — observations can only ever confirm that a law has held so far, not that it always will.

Popper's solution was to shift from induction to falsification: rather than accumulating confirmatory evidence, scientists should actively seek the most stringent possible tests that could falsify their hypotheses. A hypothesis that survives repeated, determined attempts at falsification is corroborated — not proven, but provisionally accepted as the best available account. This also explains why a single striking counter-example can overturn a well-established theory: Newton's mechanics, confirmed by centuries of observations, was overturned by relativistic evidence that classical mechanics could not account for.

The falsifiability criterion also provides a demarcation between science and pseudoscience or metaphysics. Freudian psychoanalysis and Marxist historical theory, Popper argued, were unfalsifiable — they could be interpreted to explain any possible outcome and thus made no genuine empirical predictions that could be disconfirmed. In contrast, Einstein's theory of general relativity made precise, novel predictions (e.g. the bending of starlight around massive objects) that could have been shown to be false but were confirmed — exemplifying genuine scientific hypotheses.

Kuhn and the Structure of Scientific Revolutions

Thomas Kuhn (1962) offered a historically grounded critique of Popper's rationalist account of science. Kuhn argued that science does not proceed through the continuous falsification of hypotheses by disconfirming evidence — scientists within a dominant paradigm routinely ignore or explain away anomalous findings rather than abandoning the paradigm on the basis of a single counter-example. This is what Kuhn called normal science: puzzle-solving within accepted frameworks without questioning fundamental assumptions.

Only when anomalies accumulate to a point of crisis — when too many puzzles remain unsolvable within the paradigm — does the scientific community begin to question its foundations. A scientific revolution occurs when a rival paradigm emerges that better accounts for the accumulated anomalies. The shift from the old paradigm to the new is not simply a matter of rational evidence evaluation — it involves a gestalt shift in how the whole field is perceived, and Kuhn argued it resembles a religious conversion more than a logical deduction. During the transition, the old and new paradigm are incommensurable — they define their key terms differently and ask different questions, making direct comparison difficult.

Implications for psychology: Kuhn suggested that psychology — with its multiple competing paradigms (behaviourist, cognitive, biological, psychodynamic, humanistic) — was a pre-paradigmatic or multi-paradigmatic discipline, lacking the consensus of a mature science. This raises the question of whether psychology can be considered fully scientific in the Kuhnian sense, or whether the simultaneous coexistence of incommensurable approaches signals a field still seeking its foundational paradigm.

The Replication Crisis in Depth

The replication crisis (Open Science Collaboration, 2015; Nosek et al.) refers to the finding that a substantial proportion of published psychological findings — estimated at 36–61% of social and cognitive psychology results — fail to replicate when independently retested. Contributing factors include:

• Small sample sizes: underpowered studies produce unreliable effect size estimates that may be inflated by chance.
• p-hacking (also called HARKing — Hypothesising After Results are Known): researchers engage in flexible data analysis — stopping data collection when p < 0.05 is reached, selectively reporting outcomes, or post hoc constructing hypotheses to fit the data — inflating Type I error rates far above the nominal 5%.
• Publication bias: journals preferentially publish significant results, creating a published record that over-represents real effects.
• Questionable research practices (QRPs): a range of practices that fall short of outright fraud but compromise the integrity of findings.

Responses to the replication crisis include open science reforms: pre-registration (publicly committing hypotheses and methods before data collection, preventing HARKing); open data (sharing raw data for independent verification); registered reports (editorial commitment to publish based on methods before results are known, eliminating publication bias); and multi-site replications using large, collaborative samples. These reforms aim to align scientific practice more closely with the norms of transparency, objectivity, and replicability that characterise the scientific ideal.

Evaluating Popper and Kuhn

Popper and Kuhn offer complementary but competing accounts of science. Popper describes the logical ideal — science should proceed by falsification. Kuhn describes the historical reality — science actually proceeds through paradigm-bound normal science punctuated by revolutions. Critics of Popper argue that he underestimates the social and psychological dimensions of science; critics of Kuhn argue that his account risks relativism — if paradigm choice is not fully rational, scientific progress may not be cumulative or objective. Imre Lakatos's research programmes theory attempted a synthesis: scientific theories are not falsified by single counter-examples (Popper) but researchers maintain their core assumptions through 'protective belts' of auxiliary hypotheses, revising these rather than the core when anomalies arise. Progress occurs when one research programme proves more fruitful and progressive than another.