Thank you all so much for your engagement with the Epi 222 course! I just finished going through all the rest of the homework assignments, and I very much appreciated everyone’s in-depth thinking about the material and how it applies to your own work.
A few comments for each assignment, that I hope will be thought provoking and worth the read!
Measurement assignment:
You all did a good of digging in on thinking critically about the reliability and validity of measures. I want to acknowledge that this can be challenging, since there is no such thing as a perfect measure, and once you start “looking under the hood” it can make you wonder how you can ever measure anything. I encourage you to continue to be thoughtful about your approach to measurement, without letting perfect be the enemy of the good. That said, particularly paying attention to appropriateness of applying measures to different populations and communities, and the extent to which measures address issues of equity, is critical.
Meghan Ferrara’s post provided a good overall example of how to do a good job with reliability and validity testing, including the use of factor analysis and convergent validity using scales measuring similar concepts, as well as test-retest for reliability.
Kazem Fallahzadeh’s example of establishing content and face validity for cardiovascular risk awareness provided a good example of how do to do a thorough process in that area particularly:
"The first phase was modification of questionnaire by expert panel to obtain satisfactory content validity. In this phase an expert panel assessed content validity of the questionnaire by examining whether the items were representative of the content they were intended to measure. Items were examined for representatives of the scale domain, appropriateness and relevance. The content validity index (CVI), a widely used technique in scale development determined item and questionnaire clarity, homogeneity and relevance on a 4-point Likert scale (ranging from 1=an irrelevant item to 4=an extremely relevant item). A CVI of ≥0.80 is recommended. The expert panel concluded that out of the 85 items that were evaluated, 69 met the CVI≥0.80 criterion and were retained.
The second phase was modification of questionnaire by patient focus group to obtain satisfactory face validity of the 69-item questionnaire resulting from the expert review. Face validity is assessed by end users deciding whether the questionnaire appears to measure what the researchers who developed it claim. Participants were asked to complete the 69-item questionnaire as well as to provide feedback on whether the items correctly measured the intended scales, appropriately stated the intent of the questionnaire and matched the individual’s situations. In addition, participants were asked to respond to questions about clarity, content, appropriateness, format, biases of questions and presentation of information. As a result of the focus group review of the 69-item questionnaire, 6 items were removed, 2 items were added, and several items were modified leaving a final total of 65 items with satisfactory face validity."
Doing Health Disparities Research assignment:
When thinking about what generation your work fits into, it is important to be clear about what the ultimate health outcome is. For example, for cervical cancer screening (Hunter Holt’s area of research) the outcome could be cervical cancer, or alternatively complications from over-treatment. Having too many (or too few) pap smears is one of the potential pathways in route to those outcomes that could be investigated in second generation research, or intervened up on in third/fourth generation research. You could also further peel back the layers of the onion to think about why, for example, someone has too many or too few pap smears by thinking, for example, about health CARE disparities. Similarly, when thinking about disparities in prescribing statins for prevention of kidney disease (Kazem’s work), this is also second generation (whereas disparities in kidney disease itself is second generation). I know this distinction can seem academic/pedantic, but I think it is important to keep your eye on what the health outcome you care about is, and distinguish that from the mechanisms that contribute to getting to that outcome.
A few of you also noted the ways that your work didn’t fit into the framework of research generations – which I agree with! For example, Nicolas Arger commented on doing basic science research, which doesn’t fit into those categories, but can address (or increase) disparities depending on how it done, interpreted, and disseminated. Similarly, Jonathan Amatruda talked about his work on kidney diagnostics, which has potential to impact disparities, but doesn’t fit in the generational frame.
With respect to the question about engaging with the social determinants of health, I really appreciated that many of you brought up the role of demedicalizing/moving activities outside of the formal medical system as a way to facilitate access and improve experience of care – for example, demedicalizing medication abortion (Jennifer Karlin) and promoting adherence to medications for chronic kidney disease using social networks and technology (Elizabeth Black). I recommend Jennifer Karlin’s post for a good overall review of the generations, as well as for the application of demedicalization to the second question.
For the final assignment on Strategic Science, I appreciate your wrestling with this issue. Carol Tran had a very thought-provoking post about people using science to promote problematic theories in the social and public policy sphere.
And I will end with two eloquent responses to the first question – one by Safyer McKenzie-Sampson and the other by Jonathan Amatruda’s. Thank you all for your engagement in this course, and for your excellent final papers and presentations! I look forward to seeing all the wonderful work you will all do.
Safyer:
"I further would like to assert that there is no true “objectivity” in scientific inquiry, as we all have our unconscious and conscious biases, which inevitably seeps into the way we do our work. For example, my work centers the voices and experiences of Black women who experience racism and adverse birth outcomes; it is obvious that my lived experience as a Black woman influences how I work and what research questions I ask. This lack of objectivity is also what affords me access in Black communities to carry out my research—I am trusted because I can relate to my research participants in a way that other researchers cannot. In my opinion, this adds richness to my work! I’m not advocating for race-concordance in all scientific inquiry, but rather that scientists take a step back and acknowledge their positionality and become more honest and forthcoming about how it influences the reasons and the ways in which they carry out research. Moreover, I’d love to see scientific inquiry move towards more research justice frameworks, which allows for advocacy and science to come together in a way that respects the communities for which we aim to advocate."
And Jonathan:
"Medicine is unique in that it demands that doctors be scientists, sociologists, and fellow human beings at the same time. For most who go into medicine, sitting at this nexus galvanizes a deep concern for the well-being of others. It is difficult to care for patients without realizing that society drives illness as much as biology. Though doctors have clearly been thinking about these issues since the days of Virchow or before, only relatively recently has health disparities research taken a more prominent role in academic medicine. This line of study is promising and urgent, but comes with unique risks. Because social medicine naturally leads into policy, the lines between science and advocacy can blur. Science is the endeavor to discover truth in the universe, a process which is inherently amoral (at least by the standard of human society). Advocacy, however, makes normative conclusions, all of which are based on guiding ethical principles for how society should run. In the ideal world, these two systems can harmonize. But even the most well-meaning researchers may end up letting the goals of their advocacy threaten the objectivity of their work. This is a significant hazard, even when the goals are pure. Practicing science without objectivity is in some ways akin to running an underpowered clinical trial—it risks generating conclusions that don’t represent reality and may mislead more than inform. To avoid this trap, we must maintain skepticism, for our own work and that of our peers. A diverse scientific community, with multiple contrasting viewpoints, will maintain skepticism, rigor, and encourage the constant questioning of our assumptions. In this way, the system keeps itself in check. Finally, this is a two-way street: just as science empowers advocacy, advocacy should prompt new and important avenues of scientific inquiry, which can go on to shape policy and ultimately improve society."