In Schachter (1959) experiment, a nominal scale was used to assess affiliation. Subjects selected a response item from one of three categories: wait alone, with others, or no preference. (See figure 6-3 for another example.) Nominal scales place responses into categories but do not provide the researcher with any information about differences in magnitude between the items.
The next level of measurement is called an ordinal scale, which is a rank ordering of items. The magnitude of each value is measured in the form of rank. Placing first, second, or third in a track event is an example of an ordinal scale. The winner’s time in faster than the runner-up’s time; the runner-up’s time is faster than the time of the third-place finisher. Here, an ordinal scale gives us some idea about the relative speed of the three racers, but it does not tell us the precise speed of any single racer. A researcher interested in studying the effects of violent cartoons on children’s aggressiveness might rank order different cartoons by how violent they are. “Gundum Wing” might be ranked higher than “Power Puff Girls,” but we would not know how much higher. To quantify the magnitude of differences between events, we turn to higher levels of measurement.