Wundt versus Galton--two approaches to gathering psychophysical measurements.

In the Wundtian paradigm, commonly used in visual science and psychophysics, each new observer is treated as a replication. Any conclusions regarding the processing of visual information can be extended to other observers only by indirect inference from successful replication. Differences between individuals require a new theory for each observer, whereas in the neo-Galtonian paradigm, commonly used in the rest of psychology, these differences are treated as noise, or as grounds for a between-groups analysis. The Wundtian paradigm was arguably the original method of experimental psychology. But what Wundt, often seen as the forefather of modern psychology, studied in his 1879 laboratory consisted mostly of simple sensory judgments (such as of the relative brightness of two targets), which would now come under the general heading of psychophysics (Danziger 1990). He was profoundly influenced by the earlier sensory physiologists, such as Helmholtz and Fechner, themselves considered forefathers of the related discipline of psychophysics. Galton was the first to use group statistics in psychology in his infamous 1869`Hereditary Genius' (Wozniak 1999). In the 1930s, with the emergence of behaviourism (Watson 1913; Skinner 1931) coupled with advances in statistics (Fisher 1935) the neo-Galtonian method, applying Fisherian hypothesis testing both to differences between individuals and to general mental phenomena, came to dominate the field of psychology (Danziger 1990). The term neo-Galtonian is itself denigratory, implying the commercial demand for standardised testing in psychology was met by the surreptitious application of racist ideology. Whereas psychology abandoned the procedure of trained introspection (Wundt 1896/1897) with the study of individual differences (Cattell 1906) and behaviourism (Watson 1913), psychophysics, with its association with the exact sciences, remained largely immune to this revolution. Thus, ironically, psychophysicists' response to Fisherian and Neyman ^ Pearson statistics (Neyman 1934; Pearson 1938) was the formulation of signal detection theory, which saw individual observers as statisticians, but was tested in experiments employing the Wundtian paradigm (eg Tanner and Swets 1954). Meanwhile, a quiet probabilistic revolution was taking place in other areas, including medical research (Gigerenzer et al 1989). We estimate that the number of papers employing the clinical method (of which the Wundtian paradigm is an example) has fallen from almost 50% in the 1961 Lancet, to less than 20% in 1997, and those in a special section entitled``Case Report''. The main body of the journal is now devoted to original articles using the statistical (neo-Galtonian) method. In the neo-Galtonian paradigm, a general hypothesis …