Psychometric properties of a scale to measure menopause-related symptoms in two ethnicities.

OBJECTIVES: To evaluate the psychometric properties of a scale we constructed to measure menopause-related symptoms in Hispanic and White non-Hispanic women. METHODS: Items were generated from the literature and clinical experiences. Hispanic (n = 123, 51.34 +/- 5.14 years) and White (n = 210, 53.36 +/- 4.80 years) women completed the scale and other instruments used to validate respective constructs. A subgroup of 69 women completed our scale a second time. RESULTS: Following exploratory factor analysis using the entire sample, confirmatory factor analysis was conducted for Hispanic and White women, separately. The same four-factor model (vasomotor, psychological symptoms, relationship(s), and other symptoms) fit data for both ethnic groups. The coefficient alpha for internal consistency was 0.852 and 0.849 for the initial and second administration, respectively, and the test-retest reliability was significant (p < 0.001). There were differences in factor scores among groups with different menopausal status. Factor scores were significantly correlated with scores of a global quality-of-life instrument (p < or = 0.001). CONCLUSIONS: Our study suggests that the scale is appropriate for use for Hispanic and White non-Hispanic women and is capable of detecting differences among women with different menopause-related symptoms. Menopause-related symptoms measured by the scale were correlated with global quality of life.

Analysis Of Quasi-Experimental Time-Series Designs.

Multivariate procedures, based on the general linear hypothesis, are outlined for inferring treatment effects in quasi-experimental time-series designs. These procedures, which take into account the dependent nature of the data obtained in time-series experiments and which require comparing the fitted regression curves for the pre-treatment and post-treatment observations, provide exact tests of significance for the various hypotheses of interest.