ABSTRACT
In the previous chapters, we have discussed various approaches to connect an observed score to a true score. Within the random sampling theoretic framework, true scores are rarely estimated directly. Rather, the strength of the linear relationship between the observed score and the true score is estimated through a reliability coefficient. When this coefficient is high, the observed score is effectively a linear transformation of the true score. However, a high reliability coefficient only indicates that the observed score is an effective stand-in for the true score. It does not indicate that the score can be interpreted as an indication of the quantity of the construct intended to be measured. In other words, demonstrating reliability alone is not adequate. In addition, we need to demonstrate the validity of the test score so that the score can be meaningfully interpreted.
