(I could not attach papers due to size restrictions)
1.Choose a paper describing the development or validation of a measure of relevance in health disparities research (please give the full citation and/or upload the paper if that's possible).
Hager ER, Quigg AM, Black MM, et al. Development and validity of a 2-item screen to identify families at risk for food insecurity. Pediatrics. 2010;126(1):e26-e32. doi:10.1542/peds.2009-3146.
http://www.ncbi.nlm.nih.gov/pubmed/20595453
2.What was the definition of the construct?
Household food insecurity is defined as the inability to obtain adequate food because of constrained resources, at some time during the year.
3.How did the authors provide evidence on the validity of the measure? Could you think of additional approaches to validating the measure?
The authors took an existing 6-item measure of food insecurity (that is valid and reliable) and sought to establish validity of a 2-item screener (i.e., 2 of the 6 items) to be used in a clinical setting. To establish validity of this screener, they collected data from a large number of participants (n > 30,000), administered the full measure, along with other health measures and indices, and examined the sensitivity and specificity of the 2 items compared to the full measure. They used a 2x2 table to calculate how many families were correctly identified by the screener as being food secure divided by families identified as food secure from the 6-item measure. Since the 2 questions are essentially yes/no questions, they examined sensitivity and specificity when participants provided affirmative responses to 1 question only, both questions, or question 1 and/or 2. The last approach provided the best strategy, with a sensitivity of 97% and specificity of 83%.
4.How did the authors provide evidence on the reliability of the measure? Could you think of additional approaches to evaluating the reliability of the measure?
The authors did not provide any information about the reliability of the screener. I am guessing it is because it is only 2 items, but they also did not provide reliability information of the full scale that was administered. At the very least, I think they should have provided Cronbach’s alpha for the full scale.
5.Describe the implications of a lack of measurement validity or reliability for future research applications.
I think this is a very important issue in health disparities research, which is based on group comparisons of sorts (whether absolute or relative). Not only do measures need to be valid in the general population, but it should also be valid for any specific populations that are being studied that may be qualitatively different from the general population. For example, I often think about validity issues when conducting research with Asian Americans, who are often immigrants and come from diverse cultural backgrounds. Some constructs, such as depression, may have a qualitatively different meaning and may be experienced differently in this group. E.g., “feeling blue,” while it can be literally translated, does not translate conceptually. Finding group differences based on some kind of scale or measure that has been validated in one group but not another may yield false findings. Research on validity is needed to ensure that items are functioning the same across groups. In terms of reliability, I think that using unreliable scales can be a waste of resources and creates unnecessary participant burden, not to mention that it is not sound research.
Part 2:
1.Find a paper describing a health disparity (please give the full citation or upload the paper)
Thomas JF, Temple JR, Perez N, Rupp R. Ethnic and Gender Disparities in Needed Adolescent Mental Health Care. Journal of Health Care for the Poor and Underserved. 2011;22(1):101-110.
http://www.ncbi.nlm.nih.gov/pubmed/21317509
2.Summarize the construct and measurement of the dimension of disparity (e.g., race, SES) and the outcome measured (e.g., self-rated health).
The construct examined in this study was depressive symptoms experienced during the past week, examined across male and female gender and race/ethnicity (African American, Hispanic, and White, but aggregated to White and non-White). The outcome was what the authors called unmet need, defined as “students who score in the moderate or high range on a depression screening, but who have never received treatment for depression, disclosed symptoms of depression, or been diagnosed with depression.”
3.What is the evidence for the validity and reliability of the measures?
The authors used the Center for Epidemiologic Studies Depression Scale (CES-D) to measure depressive symptoms among high school students. In the Methods section, they provide information about previously determined cut off scores, how cut offs differ for males and females, and a brief mention of specificity and sensitivity as it relates to the cut off scores. They also mention that “The CES-D has demonstrated strong internal consistency, adequate test-retest reliability, and is a valid measure of depression for high school students.” However, they do not provide data on the reliability with their particular sample. They also do not mention whether the CES-D has been previously determined to be valid and reliable with non-White students.
The measure of unmet need is a bit murkier to me. First, the definition of unmet need (provided in response to #2) encompasses a lot of aspects that has to do with treatment for depression, disclosing symptoms of depression (presumably to anyone), and being diagnosed (but perhaps not receiving treatment). The authors state that they used three yes/no items to assess these aspects of unmet need, but do not provide any information about the validity. For example, there could be recall bias here or a lack of understanding that would make these questions not valid.
4.What is the reference category used for the disparity measure? Why does this reference category make sense (or not) for this research question?
Based on the authors’ hypothesis, non-White is the reference category with respect to race/ethnicity, as the authors hypothesize that non-Whites are more likely than Whites to have an unmet need. This makes sense because they authors discuss that racial/ethnic minority adolescents are disproportionately of lower income and are less likely to be insured, compared to White counterparts. The authors do not, however, explicitly mention a reference category for gender.
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 is quantified as differences in depression and unmet need (expressed as percentages) and odds ratios of reporting a met need (which is strange because the focus of the paper is on unmet need). In this case, both absolute and relative measures are provided. I think both are indeed preferred for this research question, but I would examine unmet need rather than met need as the DV when conducting logistic regression analyses. I’m not sure why the authors flipped it around this way, since they premise the paper on the argument that racial/ethnic minorities are more likely to have an unmet need, but instead of examining the question in that way, they framed it as whether whites are more likely to have a met need compared to racial/ethnic minorities.