Skip to main content
. 2019 Sep 4;7:248. doi: 10.3389/fpubh.2019.00248

Table 4.

Convergent and discriminant validity of item selection for the Duke-PH subscales.

Items Social determinants subscale Health determinants subscale
Social determinants subscale items
 Money for basic needs 0.31** 0.34**
 Living conditions 0.50** 0.24**
 Support versus stress 0.36** 0.39**
 Education 0.34** 0.19**
 Employment 0.29** 0.11*
 Discrimination 0.34** 0.23**
 Social interactions 0.27** 0.25**
Health determinants subscale items
 Perceived health 0.25** 0.36**
 Illness burden 0.40** 0.37**
 Health care quality 0.16** 0.23**
 Pain 0.20** 0.47**
 Depressed feelings 0.40** 0.34**
 Health care visits 0.06NS 0.27**
 Health confinement 0.19** 0.40**

n = 450.

Spearman rank-order correlations between item scores and their own subscale scores (shown in bold font) are item remainder correlations that do not include the index item score in the subscale score (for indication of convergent validity). Correlations between item scores and the other subscale scores (shown in regular font) include all item scores in the subscale scores (for indication of discriminant validity when compared with scores in bold font).

*

p <0.05,

**

p <0.001, NS = Not statistically significant.