Cognitive Algebra. Exact algebraic rules have been found in many areas of psychology (Chapter 21). Among other uses, these algebraic rules can provide theory control of confounding.
One application appears with the size-weight illusion. Apparent heaviness of a lifted weight is influenced through expectancy based on visual size, as was seen in Figure 5.3 (page 131). The traditional control was by elimination, placing the weights behind a screen so they could not be seen.
Theory control of the visual cue was obtained with the additive model, size + weight, demonstrated by the parallelism of Figure 5.3. The size-expectancy effect is thus fractionated out, leaving a pure measure of the sensory effect of the weight stimulus itself.
This model goes further to dissect the conscious sensation of heaviness into two nonconscious components, one for each stimulus determinant. It does this by shifting focus from the original sensory question to the more general question of how sensory and cognitive cues are integrated. This integration rule, aside from its own interest, gave a superior answer to the sensory question. b
An exemplary book on confounding is the methods book of Underwood and Shaughnessy (1983), especially for its invaluable sets of exercises. Although these exercises are limited by their framework of traditional verbal learning and memory, this book deserves emulation in every field.
The undergraduate text, Experimental Psychology, by Solso, Johnson, and Beal (1998), is based on the theme that students learn best by concrete example. Numerous case examples are used to illustrate concepts of experimental method. These are expanded with 16 short articles reprinted from diverse fields, each with running commentary and with exercises for students. The material on the nature of experimental design and on control is well done and avoids the historical and philosophical didactics that deaden many texts. Although aimed at undergraduates, this book should be a useful resource for graduate classes; see also Ware and Brewer (1988), cited on page 778.
Among undergraduate texts concerned with experimental psychology, Levin and Hinrichs (1995) and Shaughnessy and Zechmeister (1997) are worthwhile. Useful texts more oriented toward social and field research include Jones' (1996) fine book, Aronson, Ellsworth, Carlsmith, and Gonzales (1990), and an admirable book by Light, Singer, and Willet (1990), concerned with observational studies of higher education.
I believe methods texts are a concern of the whole field, and deserve vigorous discussion and debate. Anyone who writes a text puts in enormous efforts, but seldom is there useful feedback. How to improve is a central question—and responsibility—yet virtually nothing is known about how. Many undergraduate methods texts suffer from over-involvement in statistics. At the graduate level, methods texts are almost nonexistent; what the unfortunate students get instead is statistics texts (see Chapter 23).