Boson!

 

In a blog about knowledge, I would be remiss not to join in on the fervor and scores of articles and posts about the apparent discovery of the Higgs boson. I do not consider myself to be even close to an expert in physics, or a novice follower of physics for that matter.  Like many people I have studied physics in high school, heard of the famous hadron collider and watched a few documentaries about string theory and Einstein; and that is where my knowledge on the subject begins and ends.  But I am really excited about this announcement!!  I am excited not because I fully understand what this boson really is (I don’t), but because we are still learning about our universe and making momentous discoveries!  For me, this discovery shines a fresh light on what is still unknown to us, the homo sapiens of planet earth; and with the unknown comes the opportunity for new exploration, new adventure and learning. It also shines light on the value that we, as a society, are willing to place on the pursuit of knowledge.  I am certain that this scientific discovery may have some very immediate and practical applications for the world; but that isn’t necessarily what we are celebrating this week.  This week, the articles I have read and the conversations I have heard are all celebrating the discovery because it is a discovery, and because we have invested so much time (48 years) and resources (from scientists and private investors alike) into gaining this new knowledge.  We care about wrestling with the complexities, we care about untangling what we know from what we do not, and we care about sharing that information with each other.  So, I am very happy to be a part of a society that got excited about a boson; there is great value in this excitement alone, and I hope that we will continue to wrestle with our questions with the same tenacity that lead to this celebration of science.  So..... Here’s to homo sapiens, Here's to CERN and Here’s to the boson!!

This is a great article that explains more about the Higgs boson and this particular discovery: The Economist on Higgs Boson

Painting By Numbers - Questions in Quantitative Program Evaluation

"I must begin, not with hypothesis, but with specific instances, no matter how minute." 

"Genius is the error in the system."

"The beholder's eye, which moves like an animal grazing, follows paths prepared for it in the picture." 
~Paul Klee

I am currently working on a new survey project, which will, conceivably, become an annual survey to gather information about client’s experiences of the health services they receive.  For now, we are still in the learning (or pilot) phase of this particular survey process.  As we begin data analysis and gather feedback from the community regarding their experiences of the survey, I am flooded with thoughts about how we measure Public Health work more broadly; and the data cleaning process as it applies to the decisions we make every day to balance feasibility, effectiveness, and inclusivity.

Often in public health and public policy, we reach for easy-to-grab statistics, neatly packaged numbers, percentages that speak to a larger story.  But I worry that the picture we may paint with our research tools and use of statistics can become more of a paint-by-numbers exercise and less of a nuanced and unique picture of the landscape.  Thus, we risk telling the same, expected story over time. 

In this particular case, we are left with the decision to address what seems to be systematic error in our “dirty data”, and exclude surveys from individuals who had less capacity to complete the survey appropriately (i.e., they didn’t do it according to the instructions).  Is this poor survey design?  Perhaps - It seems there is reason to consider whether the response categories fully fit the clients' reality; or whether it better fits the institutional, anticipated framework (similar to confirmation bias).  Is this a demonstration of the links between education and health?  Maybe – That certainly seems like an overall trend we are noticing in our analyses linked to demographics.  Frequently, however, it seems that we are excluding  those responses that were more creative and more individualized.  For example, some individuals responded by designing their own, alternative, Likert scales; and/or wrote qualitative descriptions rather than utilize the closed-ended response categories.  Of course, these trends are not consistent across any particular survey item or set of items, so we can’t simply exclude certain pieces of the survey.  Instead, in our analysis, we may simply lose the responses from those few individuals who didn’t want to answer our questions in a prescriptive manner.

By design, our use of quantitative data and statistics is intended to give us an aggregate picture, to summarize trends, to make decisions for the program as a whole, not necessarily for each individual.  Additionally, there are some creative qualitative methods that we may employ to capture the more individualized responses, but even these efforts will not help us capture that information and those perspectives that are simply missing on different parts of the survey.  Further, these responses will be missing from the easily digestible summary statistics. So, I am left with the questions:  Are we unintentionally marginalizing a group of people in our data analysis process; Do we even have enough information to identify who the group is that we are excluding; Is it appropriate to measure satisfaction with such broad strokes; Are we giving a voice to the voiceless or silencing those voices that do not provide the anticipated response categories;  When we utilize quantitative information to inform and improve health systems are we confident that we are including the perspectives of those who are often not heard?  If the numbers are small enough, and they don't change the aggregate picture, we don’t worry too much about these issues – But, should we worry more about the act of exclusion in this process (at least from a philosophical, non-statistical perspective)? Should we simply dismiss these outliers, whether or not they skew our summary statistics?

To be inclusive in our information gathering processes is not easy and certainly not simple to standardize.  An obvious solution is to design a participatory evaluation process that deliberately includes the voices of the more marginalized clients.  With limited resources and capacity, such a participatory process is often not feasible on a consistent basis; and it may not be ideal for measuring trends consistently over time. In Public Health, we often leave the needs of the individual to the clinician or direct service provider that can most fully assess individual needs, but we miss this individualized perspective at the policy and program-design level.  So, as I deal with “data cleaning” for this survey project, part of me wishes for a messier, less prescribed painting on the canvas – One that is harder to instantly discern but filled with depth and nuance.