Describe the assessment of reliability in a measurement model

What will be an ideal response?

Recall from Chapter 9 that an unreliable construct cannot be valid. So first we should assess the reliability of the constructs in the measurement model. As in Chapter 9, the coefficient alpha can be used to assess reliability. In addition, we compute composite reliability (CR), which is defined as the total amount of true score variance in relation to the total score variance. Composite reliability corresponds to the conventional notion of reliability in classical test theory (see Chapter 9). As general guidelines, composite reliabilities of 0.7 or higher are considered good. Estimates between 0.6 and 0.7 may be considered acceptable if the estimates of the model validity are good.