ABSTRACT
It is often hard to convince doctors and therapists that good research can be carried out with individuals or with small groups of subjects. Even academic psychologists (though not neuropsychologists or behavioural psychologists) find this idea difficult to accept. Although their discipline began with the intensive study of individuals (e.g., Wundt and Ebbinghaus used themselves as subjects for many years), there is a widespread belief amongst psychologists that findings from small-N studies are neither valid nor generalisable. This belief can be traced back to the 1920's when new statistical procedures were developed (e.g., Fisher, 1925). These procedures required large numbers of subjects, the use of randomisation to eliminate secondary variables, and the use of control groups. In many quarters, this kind of research has become the trademark of acceptable scientific experimentation.
