Abstract

We argue that tests of measuring situated competence can be viewed differently from cognitive and socio-cultural perspectives. Also we suggest analytical methods for each perspective. From cognitive perspective, competence is the main issue and situational factors are not of substantive interest and they are treated as noise or measurement errors. When specificity is a part of measurement error and two measures share the same measurement method, shared specificity causes a corelation between the measurement errors(or uniquenesses). From mathematical point of view, both measurement error and latent variable(trait factor) can be predictors of measured variables. Then a corelation between measurements can be reparameterized as a latent variable. However, this latent variable is interpreted as a method factor representing a consistent use of a certain measurement method to measure trait factors. This method factor represents the situation imposed in the measurement process. Specification of either measurement error corelations or situation factors depends on whether the investigator has cognitive or socio-cultural perspective in designing the situational test.