ABSTRACT
In Chapter 3 we explained the design of the present experiment. This design called for observations to be made before and after the experimental treatment. ‘Observations’ here does not simply mean ‘watching’, but rather ‘gathering empirical data’; and just as physical scientists use instruments such as ammeters or weighing scales to provide measures of things we cannot see, so behavioural scientists have their own instruments. Psychometric tests are less directly validated (i.e. proven to measure what we want them to measure) and less reliable (i.e. proven to give us the same measure whenever the same amount is measured) than are most instruments used in the physical sciences. Hence, we have to pay much more attention to the problem of selecting our tests. It’s easy to open a can with an efficient can opener: it’s only when there is no can opener that we have to consider what instrument to use to try and get at the beans!
