Richard A. Zeller

Measurement in the social sciences : the link between theory and data

1. Introduction to measurement 2. Factor analysis 3. Reliability 4. Validity 5. Evaluating systematic error 6. Integrating reliability and validity Appendix: multiple indicators Bibliography Index.

Osteoporosis Knowledge, Health Beliefs, and DXA T-Scores in Men and Women 50 Years of Age and Older.

PURPOSE To compare the knowledge of osteoporosis, revised health belief model variables (RHBM), and DXA (dual energy x-ray absorptiometry) T-scores among men and women 50 years of age and older. DESIGN This was a secondary analysis that used 218 healthy community-based women 50 to 65 years of age and 226 healthy community-based men ≥50 years of age. Data for women were collected for 18 months during 2001–2003, and data for men were collected for 18 months during 2004–2006. The variables were knowledge of osteoporosis, RHBM variables (susceptibility, seriousness, benefits and barriers of calcium intake and of exercise, health motivation, self-efficacy of calcium intake and exercise), and DXA T-scores. METHOD The women and men completed an osteoporosis questionnaire prior to having a DXA screening test. A total of 190 women and 187 men had DXA scans. FINDINGS More than half of the sample had abnormal bone density scans. Knowledge of osteoporosis was low for women and even lower for men. Seven of the nine health belief variables were associated with gender. Women perceived osteoporosis to be serious and that they were susceptible. Men did not perceive osteoporosis to be serious or that they were susceptible. Men were more health motivated and more confident of their ability to engage in exercise.

Tailored interventions to enhance osteoporosis prevention in women.

PURPOSE The purpose of this pilot study was to determine if tailored nursing interventions based on personal knowledge of bone mineral density from a dual-energy x-ray absorptiometry cause increases in knowledge of osteoporosis, health beliefs, or osteoporosis-prevention behaviors in postmenopausal women 50–65 years of age, 6 months after the intervention. METHOD The design for this pilot study was a two group quasi-experimental design. The treatment group received a tailored intervention; the control group did not. Outcome data were gathered at 6 months after dual-energy x-ray absorptiometry. The tailored intervention was designed and given to each woman via telephone using her dual-energy x-ray absorptiometry results and osteoporosis questionnaire data that addressed her knowledge of osteoporosis and osteoporosis-prevention behaviors of calcium intake, exercise, smoking, and alcohol use. A written mailed copy of the intervention followed the telephone interview. Six months after the intervention, the women were mailed another osteoporosis questionnaire to determine if the tailored intervention made a difference in the outcome variables. SAMPLE A total 124 women between the ages of 50 and 65 (101 control, 23 treatment) comprised the sample. RESULTS There was no difference in knowledge between groups. On the average, there were significantly more perceived barriers to calcium in the tailored group (mean = 13.48) than in the nontailored group (mean = 11.55) (t = 2.147; df = 122; p = .034). There were significantly more perceived barriers to exercise in the tailored group (mean = 14.39) than in the nontailored group (mean = 12.21) (t = .144; df = 122; p = .034). Daily calcium intake increased in both the tailored and the nontailored groups. The tailored intervention increased womens daily calcium intake from 614.28 to 1039.10 mg (t = −2.896; df = 22; p = .008). The nontailored group daily calcium intake increased from 587.91 to 916.30 mg (t = −3.541; df = 100; p = .001); there was no significant difference between the groups. Weight-bearing exercise behaviors decreased from 96.04 minutes to 59.2 minutes in the tailored group but increased slightly in the nontailored group from 81.47 to 87.26 minutes of exercise. CONCLUSION Tailored interventions increased womens perceived barriers to calcium and exercise. Both groups increased calcium intake. The mixed findings of increased perception of barriers to calcium and exercise and decreased exercise behaviors indicate the need for further study. This important intervention has implications for orthopaedic nurses and healthcare professionals involved in health promotion and prevention of osteoporosis.

Combining Qualitative and Quantitative Techniques to Develop Culturally Sensitive Measures

The purpose of this chapter is to propose and illustrate a strategy designed to develop valid, reliable, and culturally sensitive measures of risk perceptions, social support, and coping strategies as applied to HIV/AIDS behavior and mental health issues for special populations. The specific concern concentrates on the identification and measurement of salient social dimensions for those who face a high risk of HIV/AIDS infection. The measurement effort will address:

Depression and codependency in women

Seven million American women are depressed, and 40 million Americans, primarily women, have been labeled as codependent. This study aimed to identify the prevalence of codependency in women undergoing treatment for depression, examine the relationship between codependency and depression, and determine which of the symptoms of codependency are most highly predictive of depression scores. Depression and codependency were measured in a sample of 105 depressed women by using the Beck Depression Inventory and the Codependency Assessment Tool. Descriptive statistics, Pearsons Product Moment Correlation, and multiple regression were used for analysis. Of these depressed women, 36% were moderately to severely codependent. Depression and codependency were strongly related, with the significant gamma = .92 (P < .001). Of the codependency subscales, Low Self-Worth and Hiding Self correlate most strongly with depression; Other Focus/Self-Neglect added the least-independent--explanatory power. Thus, future research should be directed toward the relationship of codependency to power, alienation of self, and personality disorders.

Development and testing of the codependency assessment tool

Codependency constitutes a significant health risk, particularly for women, because codependent women are often involved in abusive and potentially harmful relationships. Individuals who are identified as codependent can engage in therapy and gain knowledge and freedom from such relationships. However, there is no reliable and valid measure of codependency that is consistently used to identify these individuals. This article describes the development and testing of the Codependency Assessment Tool, a multivariate tool that conceptualizes codependency as a construct comprising five factors: (1) Other Focus/Self-Neglect, (2) Low Self-Worth, (3) Hiding Self, (4) Medical Problems, and (5) Family of Origin Issues. The instrument has excellent reliability and validity. Its test-retest reliabilities = .78 to .94; Cronbachs alpha = .78 to .91. Criterion validity was determined to be established by using known groups; construct validity was established by comparing the codependency dimensions with depression.

The Utility of Intervening Constructs in Experiments

The interpretation of experimental manipulations and the precise relation. ship of that interpretation to the observed effects are two major problems facing experimental procedures designed to test social theory. The results of the present study indicate that careful measurement of intervening attitudinal constructs is useful in dealing with these problems. Not only is the predictive power of the experimental model increased, but theoretical assumptions which usually remain untested in experimental research are clarified and explicitly tested. In particular, attention to dimensionality, validity, and reliability of intervening attitudinal constructs avoided serious misinterpretation of the intervening process in the experimental situation and the associated inference to the applicability of the theory.

Codependency in male and female helping professionals

The purpose of this study is to determine and compare the prevalence of codependency in a group of 77 female and a group of 72 male helping professionals. Relatively low rates of codependency were observed in this sample, and the five-factor structure of the Codependency Assessment Tool (CODAT) was replicated. Males showed slightly higher codependency than females on the total CODAT and on the Hiding Self and Family of Origin Issues subscales. The CODAT was shown to be relevant to both males and females. Further research is needed to determine if a general population of male and female subjects would yield similar results.

SOCIAL JUDGMENTS OF RESPONSIBILITY IN ELDER SELF-NEGLECT CASES

Abstract In this study, 1,038 case judgments made by 26 adult protective service (APS) workers in the United States were examined using hypothetical elder self-neglect vignettes. The independent variables used to construct realistic vignettes were based on the relevant literature, the law in the jurisdiction being studied, and other practical considerations. The factorial survey method was used in the vignette construction phase, and multiple regression was used for analysis. The independent variables of age, household income, disability, family presence and involvement, and the form and consequence of the self-neglect revealed statistically significant findings, but the main effect rested with the social perception of disability in assigning responsibility. Over 25% of the variance was explained in the model, the majority of which could be attributed to the victims disability. The findings indicate that judgment of responsibility was almost entirely due to the perception of the victims disability. Resp...

The Effects of Violating the Normality Assumption Underlying r

Using randomly generated samples from populations having varying rhos, different shapes of distributions, and varying sample sizes, the effects of violating the population normality assumption underlying r were examined. Results indicate that the normality assumption underlying r is robust (i. e., that violation of the population normality assumption does not seriously alter the interpretation of r). Violation of the population normality assumption appears, therefore, to be insufficient reason to deny r a place as a major tool for sociological analysis.