The different scales correlate so highly, in part, because the original selection of the items for inclusion in each scale was based on a comparison of normals with different clinical groups.
The items, then, were selected based on their differentiation of normals from various psychiatric populations, rather than on their differentiation of one psychiatric population from another. Although the psychiatric groups varied from the normals on several traits, this manner of scale construction did not develop accurate measurements of these different traits.
Rather, the scales are filled with many heterogeneous items and measure multidimensional, often poorly defined attributes.
This approach has also led to m any items being shared with other scales. In contrast, an approach in which specific psychiatric groups had been compared with one another would have been more likely to have resulted in scales with less item overlap and with the ability to measure more unidimensional traits.
The different scales correlate so highly, in part, because the original selection of the items for inclusion in each scale was based on a comparison of normals with different clinical groups.
The items, then, were selected based on their differentiation of normals from various psychiatric populations, rather than on their differentiation of one psychiatric population from another. Although the psychiatric groups varied from the normals on several traits, this manner of scale construction did not develop accurate measurements of these different traits.
Rather, the scales are filled with many heterogeneous items and measure multidimensional, often poorly defined attributes.
This approach has also led to m any items being shared with other scales. In contrast, an approach in which specific psychiatric groups had been compared with one another would have been more likely to have resulted in scales with less item overlap and with the ability to measure more unidimensional traits.
การแปล กรุณารอสักครู่..
