[State-trait characteristics of various measured variables of psychophysiologic activation, cognitive performance and...

The distinction between state and trait sources of variation in psychological variables, familiar in differential psychology ever since the work of Cattell, is translated into a linear model of state-trait analysis. On the basis of this model indices of relative state- and trait variance of a measure (independent of errors of measurement) and the psychometric reliability of the same measure (independent of true state variance) can be determined. The model is applied to a larger set of data from in-field research employing 149 male high-school students (average age 17 years) from the city of Hamburg. Following a time-sampling plan of data acquisition, subjects were requested over a period of three successive weeks and on the average twelve times a day to indicate (on a portable computerized behavioral data recorder) their current mood and feeling and had to answer short psychometric tests (in-field psychometrics); in addition, two peripheral psychophysiological measures of activation (heart rate, finger temperature) were recorded. The psychophysiological measures depended on state variations to about two thirds of their true score variance, the different cognitive tests to about 50 to 80 percent of their true score variance on trait variations. Different scales for recording current mood and feeling states gave different state-trait variance percentages. The method also allows for determining intraindividual psychometric state reliability coefficients. Fields of application of the method in different areas of psychology are indicated.