Part 1:
- Choose a paper describing the development or validation of a measure of relevance in health disparities research (please give full citation)
Heather Bradley , Amy Tsui , Michelle Hindin , Aklilu Kidanu & Duff Gillespie (2011) Developing scales to measure perceived HIV risk and vulnerability among Ethiopian women testing for HIV, AIDS Care: Psychological and Socio- medical Aspects of AIDS/HIV, 23:8, 1043-1052, DOI: 10.1080/09540121.2010.543880
2. What was the definition of the construct?
Perceived risk is one’s self-assessed likelihood of becoming HIV-infected based on HIV knowledge and behavior. Perceived vulnerability is felt susceptibility to HIV infection even in the absence of risk behavior.
3. How did the authors provide evidence on the validity of the measure? Could you think of additional approaches to validating the measure?
Besides including these measurement items in a survey, they also conducted in-depth interviews and added items to the scales based on these findings, which increased the Cronbach’s alpha from 0.66 to 0.74 for the perceived vulnerability scale and from 0.87 to 0.89 for perceived risk. All items were validated with a post-enumeration survey. They calculated Pearson’s correlations between the scales and comparable constructs and behaviors.
Both scales had high construct validity, though perceived risk had a higher correlation with HIV status than perceived vulnerability. Conversely, perceived vulnerability is more highly correlated with HIV salience. This suggests that perceived HIV risk and perceived HIV vulnerability should be measured separately.
4. How did the authors provide evidence of reliability of the measure? Could you think of additional approaches to evaluating the reliability of the measure?
After the initial survey, a psychometric analysis of the item correlation of reliability was performed. Items correlating less than 30% with either perceived risk or perceived vulnerability were dropped form further analysis. Furthermore, internal reliability was assessed among items using Cronbach’s coefficient alpha.
As described in the paper, interviews were cross-sectional, and so they were unable to examine test-retest reliability for scale items, or how vulnerability perceptions may change over time. Future interviews and survey administration among the same participants would have provided proof of the reliability of the measures.
5. Describe the implications of a lack of measurement validity or reliability for future research applications.
Usually, a single question is asked to measure perceived HIV risk, and thus, assess their likelihood that they are infected with HIV. Their estimations may reflect either perceived behavioral risk or felt vulnerability to disease. People seeking HIV prevention services in response to perceived behavioral risk may have more risk than those solely motivated by vulnerability.
The findings of this study indicate that the typically used single item measuring perceived HIV risk may be simultaneously and inadequately measuring perceived risk and perceived vulnerability, suggesting that these very different constructs should be measured separately. Understanding how these constructs may independently and jointly motivate people to seek HIV prevention services will help target outreach and counseling messages towards those most in need of those services. The absence of an improved scale measuring these constructs may cause public health outreach and policy to misdirect their efforts.
Part 2:
- Find a paper describing a health disparity (give full citation)
Owusu-Edusei, K., Jr., et al. (2013). "The association between racial disparity in income and reported sexually transmitted infections." Am J Public Health 103(5): 910-916.
2. Summarize the construct and measurement of the dimension of disparity (e.g. race, SES) and the outcome measured (ex: self-rated health)
The objective of the study was to examine the association between racial disparity in income and reported race-specific county-level bacterial STIs, focusing on disparities between blacks and whites. They defined 2 race-income county groups (high and low race-income disparity) on the basis of the difference between Black and White median household incomes. This data was collected through the 2000 census. The outcome, disparities in STI rates across groups was reported by public health surveillance records.
Consistent with previous literature, STI rates for Blacks was substantially higher than for Whites, with racial disparities in income associated with racial disparities in chlamydia and gonorrhea rates.
3. What is the evidence for the validity and reliability of the measures?
Though the authors created their own measures without specific reliability and validity measures, they did perform both a continuous and dichotomous regression models for their race income disparity measures. They also included control variables for demographic and socioeconomic factors, including percentage of ethnicity per county, birth rate, death rate, male-to-female population, age, population density, crime rate, and suburban commute.
4. What is the reference category used for the disparity measure? Why does this reference category make sense (or not) for this research question?
The reference category was created as follows: The study created a continuous variable (race-income disparity measure) as the difference between the median household income for Whites and Blacks separately in comparison to the national average. Next, they defined the 2 race-income county groups (high vs. low race-income disparity) on the basis of Black and White median income at the county level in the 48 contiguous U.S. states. The high race-income disparity group comprised of those counties in which Black median household income (median $29,919) was lower than the national average and White median household income (median $46,046) was above the national average ($41,994).
For the purposes of this study, which was specifically looking at race-income equality, the creation of this reference category is justifiable rather than using absolute income measures which would have focused on income’s effects on STI rates, rather than how both race and income mediate these effects.
5. How is the disparity quantified? Is this an absolute or relative measure or are both provided? Describe which type of measure you would prefer for this research area, or , if both, why?
The disparity was quantified as a relative measure of median income disparity by race, which found higher STI rates in the high race-income disparity group than the low race-income disparity group that cannot be explained by differences in overall (black or white) income. The authors argue that if absolute levels of income (rather than income disparity) were the main determinant of STI morbidity, STI rates in Blacks would be expected to be lower in the higher race income disparity group than in the low race-income disparity income group, which was not found to be true. As described above, I agree that creating a relative measure for this study is justifiable to show how both race and income mediate STI rates.
The authors conclude that racial disparities in household income may be a more important of racial disparities in reported STI morbidity than are absolute levels of household income. This association of racial disparities in income with STI disparities by race can promote further examination of the mechanisms through which race-income disparities may be associated with or exacerbate disparities in STI rates. These insights can inform strategies aimed at reducing these disparities.