practitioners and experts in our research center, and the
purification procedures, suggest that the refined ECC
instrument retains good content validity [23].
D. Construct Validity
Construct validity is used to examine how well the results
obtained from using the measure fit with the theories that lie
behind the design of the instrument. There are two common
forms of construct validity, convergent and discriminant
validity [21], [24]. This research uses two methods to
confirm the construct validity of ECC: (1) the correlation
matrix approach to investigate construct validity of the
developed ECC measure [25]; (2) correlation coefficients
between each dimension of ECC scale and average score of
the ECC scale. In the correlation matrix methods, this
research examines convergent validity by determining
whether relationships between scales of the same factor are
higher than zero and large enough to execute discriminant
validity analysis [23], [26]. We also examine discriminant
validity by counting the number of times that an item had a
higher correlation with items of other factors than with those
of its own factor [25]. The correlation matrix shows only five
violations for 398 potential comparisons.