Sociology
LEARNING OBJECTIVES
The learning objectives of these papers are to substantiate and describe mediators of a biomedical disparity. Do not write about mental health outcomes like depression or anxiety, even if they have a biomedical component.
You will describe and quantify a health disparity and then attempt to explain it.
Paper 2 will explain it with 3 out of the following 5 mediators:
● EDUCATION ● TRUST & COMPLIANCE ● THE OFFICE VISIT ● SOCIAL NETWORKS/SOCIAL SUPPORT ● INSTITUTIONS
You choose the mediators.
PAPER DRAFT This paper should be roughly 1500 words long, excluding footnotes/endnotes/references. Content is more important to me than length. See this template. This is not an “essay”-type paper, so please follow the headings and structure in the template. Please write concisely.
SUGGESTIONS Sometimes you simply cannot find any published discussions of a mediator. For example, if you’re interested in trans/cis gender disparities in recovery from radiation therapy from soft-tissue sarcoma and by some miracle you are able to quantify a disparity, you may not find any research on, say, whether stress helps explain that disparity.
When this happens, please compensate for the lack of research by spending more time explaining why the mediator could explain the disparity. For example, you might more fully describe how stress mediators in general work. Alternatively, you could argue by analogy. For example, if you found research that showed that stress differences might drive some of the trans/cis gender disparity in colon cancer, you could cite that research and simply assert that the mechanism for soft-tissue sarcoma may be similar.
Sometimes we have no choice but to speculate. That’s fine. But if relevant research is out there (e.g., there’s a ton on what explains race disparities in various kinds of cancer or heart diseases) and you don’t mention any of it, it looks like you didn’t take the assignment seriously.
You should mention relevant mechanisms discussed in class. For example, when talking about economic disparities we focus on how they “do what they do” mostly in childhood. It would therefore make sense for you to demonstrate awareness of that. Or if we read that “stress reduction” is probably the least plausible thing that “social support” does to affect SES health disparities,1 then mention that. You wouldn’t want to write a paragraph that ignores that research.
Students are invited to discuss papers with me in office hours at any stage of development. Because office hours can get very busy, doing this the Friday before the paper is due may not be practical.
FORMATTING File type. Please submit a PDF.
Style. I encourage you to download or copy the template and use it. Please apply (these are not typos; please do not email me to ask about them):
● 1.25-inch margins (left and right; top and bottom are options) ● A sans-serif font ● 11- or 12-point font ● 1.2 line spacing
References. You may use any reference style that appeals to you, including copying citations from Google Scholar or having a citation manager insert them. References do not need to be perfect. (Mine aren’t.) That is why we have copy editors and computer software. Generally you would only want to do either footnotes or endnotes. If you’re not sure what to do, consider Chicago, APA, or MLA style. Strive to be reasonable rather than perfect.
Caution: If you add quotations (which is desirable), please be sure to add a specific page number to the citation where the quotation appears.
1 Matthews, Karen A., Linda C. Gallo, and Shelley E. Taylor. 2010. “Are Psychosocial Factors Mediators of Socioeconomic Status and Health Connections? A Progress Report and Blueprint for the Future.” Annals of the New York Academy of Sciences 1186(1):146–73.
GRADING A rubric is available on Moodle to see how paper components are weighted.
Students should not expect extensive feedback on their writing or analysis. (I am happy to discuss papers more fully in office hours or by written request.) In the past, students typically lose points for one or more of the following reasons:
● Failure to follow instructions; ● Not demonstrating understanding that “quantify a disparity” means providing a
number to describe the gap between the groups (e.g., a relative-risk difference or ratio, or an odds-ratio);
● Failure to demonstrate awareness of relevant course materials (especially on mediators);
● Failing to identify easy-to-find and commonly discussed research applicable to a student’s disparity;
● Not having headings and/or subheadings that make it possible to easily identify the components of the paper;
● Too few references; ● References that are unintelligible, bogus; and ● References to internet garbage rather than government papers or peer-reviewed
scientific literature.
Template
I. OUTCOME AND DISPARITY
After reading this section, I should understand what your biomedical outcome is, why it matters, and what disparity you are investigating. Notice that the following example is very concise. It’s not written in a traditional “essay-style” format.
Notice in particular that the disparity is quantified. I didn’t just say that African Americans had X% risk of the outcome and White Americans had Y% risk. Instead, I tell you what the gap is.
This section should have 2-3 high-quality citations. Websites are generally not high quality. Peer-reviewed research is generally high quality.
Example Venous thromboembolism is “the leading cause of in-hospital mortality.”2 Two variants arise from blood pooling in a vein: Deep vein thrombosis (“DVT”) is a blood clot in a vein—often deep in the leg3—and venous thromboembolism (‘VTE’) is a blockage to an organ resulting from a blood clot that becomes dislodged and travels to another part of the body.
The incidence of DVT is 30% to 60% times (relative risk of 1.3 to 1.6) higher for African Americans than White Americans across gender and age.4 A study of California residents in the 1990s reported that African Americans had the highest incidence of VTE of the race groups studied, followed in turn by Caucasians, Hispanics, and Asian/Pacific Islanders.5 VTE recurrence rates appear to be higher for African American women while “African-American men and Caucasian men have a similar incidence of recurrent VTE.”6
Owens et al. controlled for an enormous slate of comorbidities and factors associated
6 Id. at S16.
5 White, R. H., & Keenan, C. R. (2009). Effects of race and ethnicity on the incidence of venous thromboembolism. Thrombosis Research, 123, S11−S17.
4 AAOS; see also Montagnana et al. at 489.
3 American Academy of Orthopaedic Surgeons, https://orthoinfo.aaos.org/en/diseases–conditions/deep-vein-thrombosis, accessed 1/17/2018 (hereafter “AAOS”).
2 Montagnana, M., Favaloro, E. J., Franchini, M., Guidi, G. C., & Lippi, G. (2010). The role of ethnicity, age and gender in venous thromboembolism. Journal of Thrombosis and Thrombolysis, 29(4), 489−96, 489.
with knee operations and found that the odds of African Americans having DVT compared to White patients was between 39% and 95% higher.7
Though very brief, this section describes what venous thromboembolism is (it is problems caused by various kinds of blood clots) and what the consequences are (it’s the leading cause of in-hospital mortality). I have described and quantified a race disparity and substantiated it with three different sources.
II. MEDIATORS
This section has THREE subsections (A, B, and C), one for each of the three mediators you examine.
A. MEDIATOR 1 [replace with mediator name (e.g., STRESS)]
Each subsection (A, B, and C) has the following THREE subsections (underlined).
(1) MEDIATOR → OUTCOME [ replace with names, such as STRESS → DVT ]
For each mediator you select, first explain how the mediator affects the outcome. For example, if you choose STRESS as a mediator, I want to understand how stress affects deep vein thrombosis. Draw heavily from course materials and do some additional research. For example, I might draw on course materials and argue that stress increases cortisol levels and then learn that high cortisol levels are associated with blood clots or vein thickness, or something along those lines.
Typically you will want to discuss multiple things that the mediator may be doing. But please try to avoid invoking other mediators we cover in class, because that just kicks the can down the road. For example, suppose your mediator is socioeconomic status (SES). Of course, more SES helps you get more education. So it’s tempting to discuss that one of things that SES is “doing” to affect DVT is giving you more education. And it’s not wrong.
But you want instead to focus on pathways that we emphasize in class that don’t invoke other mediators we discuss. Focus on what this mediator is doing that is distinct from what other mediators may be doing.
7 Owens, J. M., Bedard, N. A., Dowdle, S. B., Gao, Y., & Callaghan, J. J. (2018). Venous thromboembolism following total knee arthroplasty: does race matter?. The Journal of Arthroplasty, 33(7), S239-S243.
(2) GROUP → MEDIATOR [ replace with names, such as RACE → STRESS ]
In this section, please do research (and draw on course materials, if they’re relevant) to show that there are group differences with respect to your mediator. For example, if STRESS is your mediator, you would argue that there are race disparities in stress and you would find research to substantiate that link.
(3) LIMITATIONS
In this section you will discuss limitations to the research supporting the causal links you discuss above. For example, you might critique the research on stress and explain why more research is needed to feel confident saying that cortisol affects DVT. And you could discuss limitations to how stress is measured or operationalized in studies linking race to stress. Check your notes—in lecture we often discuss limitations to studies on the social determinants of health. Consider health selection and confounding problems.
In this section you might also mention moderator effects that you identify or speculate may be important.
Each mediator section should have 3-5 high-quality citations, which may include course materials.
2. MEDIATOR 2 [replace with mediator name]
Do the same analysis for each of your other mediators, using the same structure.