Still grappling with demographics

Last year I wrote about my changing perspective on demographic variables. My grappling has continued since then.
I think of it as an academic puberty of sorts.

I remember the many crazy thought exercises I subjected myself to as a teenager, as I tried to forge my own set of beliefs and my own place in the world. I questioned everything. At times I was under so much construction that it was a wonder I functioned at all. Thankfully, I survived to enter my twenties intact. But lately I have been caught in a similar thought exercise of sorts, second guessing the use of sociological demographic variables in research.

Two sample projects mark two sides of the argument. One is a potential study of the climate for underrepresented faculty members in physics departments. In our exploration of this subject, the meaning of underrepresented was raised. Indeed there are a number of ways in which a faculty member could be underrepresented or made uncomfortable: gender, race, ethnicity, accent, bodily differences or disabilities, sexual orientation, religion, … At some point, one could ask whether it matters which of these inspired prejudicial or different treatment, or whether the hostile climate is, in and of itself, important to note. Does it make sense to tick off which of a set of possible prejudices are stronger or weaker at a particular department? Or does it matter first that the uncomfortable climate exists, and that personal differences that should be professionally irrelevant are coming into professional play. One could argue that the climate should be the first phase of the study, and any demographics could be secondary. One might be particularly tempted to argue for this arrangement given the small sizes of the departments and hesitation among many faculty members to supply information that could identify them personally.

If that was the only project on my mind, I might be tempted to take a more deconstructionist view of demographic variables altogether. But there is another project that I’m working on that argues against the deconstructionist view- the Global Survey of Physicists.

(Side or backstory: The global survey is kind of a pet project of mine, and it was the project that led me to grad school. Working on it involved coordinating survey design, translation and dissemination with representatives from over 100 countries. This was our first translation project. It began in English and was then translated into 7 additional languages. The translation process took almost a full year and was full of unexpected complications. Near the end of this phase, I attended a talk at the Bureau of Labor Statistics by Yuling Pan from Census. The talk was entitled ‘the Sociolinguistics of Survey Translation.’ I attended it never having heard of Sociolinguistics before. During the course of the talk, Yuling detailed and dissected experiences that paralleled my own into useful pieces and diagnosed and described some of the challenges I had encountered in detail. I was so impressed with her talk that I googled Sociolinguistics as soon as I returned to my office, discovered the MLC a few minutes later. One month later I was visiting Georgetown and working on my application for the MLC. I like to say it was like being swept up off my feet and then engaging in a happy shotgun marriage)

The Global Survey was designed to elicit gender differences in terms of experiences, climate, resources and opportunities, as well as the effects of personal and family constraints and decisions on school and career. The survey worked particularly well, and each dive into the data proves fascinating. This week I delved deeper into the dynamics of one country and saw women’s sources of support erode as they progressed further into school and work, saw the women transition from a virtual parity in school to difficult careers, beginning with their significantly larger chance of having to choose their job because it was the only offer they received, and becoming significantly worse with the introduction of kids. In fact, we found through this survey that kids tend to slow women’s careers and accelerate men’s!

What do these findings say about the use of demographic variables? They certainly validate their usefulness and cause me to wonder whether a lack of focus on demographics would lessen the usefulness of the faculty study. Here I’m reminded that it is important, when discussing demographic variables, to keep in mind that they are not arbitrary. They reflect ways of seeing that are deeply engrained in society. Gender, for example, is the first thing to note about a baby, and it determines a great deal from that point in. Excluding race or ethnicity seems foolish, too, in a society that so deeply engrains these distinctions.

The problem may be in the a priori or unconsidered applications of demographic variables. All too often, the same tired set of variables are dredged up without first considering whether they would even provide a useful distinction or the most useful cuts to a dataset. A recent example of this is the study that garnered some press about racial differences in e-learning. From what I read of the study, all e-learning was collapsed into a single entity, an outcome or dependent variable (as in some kind if measure of success of e-learning), and run by a set of traditional x’s or independent variables, like race and socioeconomic status. In this case, I would have preferred to first see a deeper look into the mechanics of e-learning than a knee jerk rush to the demographic variables. What kind of e-learning course was it? What kinds of interaction were fostered between the students and the teacher, material and other students? So many experiences of e-learning were collapsed together, and differences in course types and learning environments make for more useful and actionable recommendations than demographics ever could.

In the case of the faculty and global surveys as well, one should ask what approaches to the data would yield the most useful analyses. Finding demographic differences leads to what- an awareness of discrimination? Discrimination is deep seeded and not easily cured. It is easy to document and difficult to fix. And yet, more specific information about climate, resources and opportunities could be more useful or actionable. It helps to ask what we can achieve through our research. Are we simply validating or proving known societal differences or are we working to create actionable recommendations? What are the most useful distinctions?

Most likely, if you take the time to carefully consider the information you collect, the usefulness of your analyses and the validity of your hypotheses, you are one step above anyone rotely applying demographic variables out of ill-considered habit. Kudos to you for that!

Total Survey Error: nanny to some, wise elder for some, strange parental friend for others

Total Survey Error and I are long-time acquaintences, just getting to know each other better. Looking at TSE is, for me, like looking at my work in survey research through a distorted mirror to an alternate universe. This week, I’ve spent some time closely reading Groves’ Past, Present and Future of Total Survey Error, and it provided some historical context to the framework, as well as an experienced account of its strengths and weaknesses.

Errors are an important area of study across many fields. Historically, models about error assumed that people didn’t really make errors often. Those attitudes are alive and well in many fields and workplaces today. Instead of carefully considering errors, they are often dismissed as indicators of incompetence. However, some workplaces are changing the way they approach errors. I did some collaborative research on medical errors in 2012 and was introduced to the term HRO or High-Reliability Organization. This is an error focused model of management that assumes that errors will be made, and not all errors can be anticipated. Therefore, every error should be embraced as a learning opportunity to build a better organizational framework.

From time to time, various members of our working group have been driven to create checklists for particular aspects of our work. In my experience, the checklists are very helpful for work that we do infrequently and virtually useless for work that we do daily. Writing a checklist for your daily work is a bit like writing instructions on how you brush your teeth and expecting to keep those instructions updated whenever you make a change of sorts. Undoubtedly, you’ll reread the instructions and wonder when you switched from a vertical to a circular motion for a given tooth. And yet there are so many important elements to our work, and so many areas where people could make less than ideal decisions (small or large). From this need rose Deming, with the first survey quality checklist. After Deming, a few other models arose. Eventually, TSE became the cumulative working framework or foundational framework for the field of survey research.

In my last blog, I spoke about the strangeness of coming across a foundational framework after working in the field without one. The framework is a conceptually important one, separating out sources of errors in ways that make shortcomings and strengths apparent and clarifying what is more or less known about a project.

But in practice, this model has not become the applied working model that its founders and biggest proponents expected it to be. This is for two reasons (that I’ll focus on), one of which Groves mentioned in some detail in this paper and one of which he barely touched on (but likely drove him out of the field).

1. The framework has mathematical properties, and this has led to its more intensive use on aspects of the survey process that are traditionally quantitative. TSE research in areas of sampling, coverage, response and aspects of analysis is quite common, but TSE research in other areas is much less common. In fact, many of the less quantifiable parts of the survey process are almost dismissed in favor of the more quantifiable parts. A survey with a particularly low TSE value could have huge underlying problems or be of minimal use once complete.
2. The framework doesn’t explicitly consider the human factors that govern research behind the scenes. Groves mentioned that the end users of the data are not deeply considered in the model, but neither are the other financial and personal (and personafinancial) constraints that govern much decision making. Ideally, the end goal of research is high quality research that yields a useful and relevant response for as minimal cost as possible. In practice, however, the goal is both to keep costs low and to satisfy a system of interrelated (and often conflicting) personal or professional (personaprofessional?) interests. If the most influential of these interests are not particularly interested in (or appreciative of) the model, practitioners are highly unlikely to take the time to apply it.

Survey research requires very close attention to detail in order to minimize errors. It requires an intimate working knowledge of math and of computer programming. It also benefits from a knowledge of human behavior and the research environment. If I were to recommend any changes to the TSE model, I would recommend a bit more task based detail, to incorporate more of the highly valued working knowledge that is often inherent and unspoken in the training of new researchers. I would also recommend a more of an HRO orientation toward error, anticipating and embracing unexpected errors as a source of additions to the model. And I would recommend some deeper incorporation of the personal and financial constraints and the roles they play (clearly an easier change to introduce than to flesh out in any great detail!). I would recommend a shift of focus, away from the quantitative modeling aspects and to the overall applicability and importance of a detailed, applied working model.

I’ve suggested before that survey research does not have a strong enough public face for the general public to understand or deeply value our work. A model that is better embraced by the field could for the basis for a public face, but the model would have to appeal to practitioners on a practical level. The question is: how do you get members of a well established field who have long been working within it and gaining expertise to accept a framework that grew into a foundational piece independent of their work?