ABSTRACT
This chapter is about certain spatial types of measurement of people, as carried out in the social and biological sciences. The global issue for us is perceptual classification and we specialize on spatial, for instance, geometric or topological, aspects. We offer more detail below, but for now, we simply note that classification is the assignment of an object, typically presented to one or more of our sensory modalities, to one of some number (perhaps infinite!) of classes or designated sets, often names. Because in general terms, this is neither more nor less than a function or mapping of a set of things (stimuli, memories, etc.) into another (the names or classifications), this makes this concept extremely broad. Thus, classification covers many specific activities of psychological interest. Some important ones are the following:
A famous form of classification is that of categorization, where a bunch of objects is partitioned so that each object is separately assigned to a single (typically) unique category. There are many kinds of categorization but one of special importance is usually given its own title.
Identification is where each category has exactly one member (this may be rare in the real life). We can perform an experiment where each person has exactly one name and each name is given to exactly one person, but of course, in the real world, the name of “John” is given to thousands of people. In most of the interesting psychological situations, classification is made difficult by any number of factors. This could be because learning of the assignments is incomplete, such as in a recognition memory experiment. In this case the presented stimulus may be very easy to perceive, but the learned patterns may have become quite noisy, making the recognition task challenging. Other situations include some kind of added noise, randomness, or even very brief exposure periods.
Yes–no signal detection is the specification by an observer of whether a signal has been presented. There are only two responses: “YES” versus “NO” could be used either with faulty memory (the celebrated “Old” versus “New” memory-recognition experiment) or with simply “hard-to-detect” physical stimuli.
Typically with psychological scaling, the observer gives ratings on one or more aspects of the stimuli. For instance, one might simply be asked to report a number that seems to represent the psychological magnitude of an aspect of a stimulus, such as how happy a particular face looks (see below for more on this). Or, as is often the case in multidimensional scaling (explained a bit later), the observer might be asked about the psychological similarity of two or more objects. An old anthology, but one still held in high esteem, with several chapters that are relevant to our present enterprise is Volume I of the Handbook of Mathematical Psychology (Luce, Bush, & Galanter, 1963). The philosophy of science of the present authors regarding measurement and modeling of people and an expansion of some of themes, can be located in Townsend and Thomas (1993).
