f I had to prioritize one, I would choose discriminant validity as more important in structural equation modeling. From my research paper on using Kahoot in learning, I ...
f I had to prioritize one, I would choose discriminant validity as more important in structural equation modeling. From my research paper on using Kahoot in learning, I measured constructs such as learning motivation, engagement, and perceived effectiveness. These constructs are conceptually related, so it is easy for them to overlap. If discriminant validity is weak, the model cannot clearly distinguish between these constructs. For example, motivation and engagement might be treated as the same factor, which makes the interpretation unclear.
Theoretically, discriminant validity ensures that each construct is unique and not redundant. Even if convergent validity is achieved (items load well on a factor), the model is still problematic if different constructs are not clearly separated.
In practice, weak discriminant validity can lead to multicollinearity, unstable estimates, and misleading conclusions. Researchers may wrongly interpret results or overestimate relationships between variables.
Therefore, while both types of validity are important, discriminant validity is more critical because it ensures the distinctiveness and clarity of constructs, which is essential for meaningful analysis