Guest Post: The Confidently-Quantified Self

Check out Measured You Store for great deals on tracking gadgets and apps!

statistics course for self-trackersThe following post was written by Dr. Alan Dabney. Professor Alan Dabney received his Ph.D. in biostatistics from the University of Washington in 2006. He joined the faculty in the statistics department at Texas A&M University later that year and received tenure in 2011. Dr. Dabney conducts research in the analysis of “big data,” particularly the kind that originate from biological applications; for a list of his research publications, please see his Google Scholar profile. In addition to his research activities, Dr. Dabney is an award-winning teacher of both undergraduate and graduate students in the statistics department at Texas A&M. He is also active in the creation of non-standard educational media that is broadly accessible. Examples include his upcoming “graphic novel,” The Cartoon Introduction to Statistics, and his featured role in W.H. Freeman’s Stat-Clips video lecture series. He can be reached at adabney (at) stat.tamu (dot) edu.

I just got home from attending my very first QS event, the 2013 QS Europe conference in Amsterdam! I attended the conference because I am a statistician, and because I am enthusiastically passionate about self-learning. In my Breakout session and in subsequent interactions with the conference participants, I learned a couple of interesting things.

First, self-trackers are statisticians, in the purest sense. Or, we might say, self-trackers are data scientists, in that we carefully collect, analyze, interpret, and draw conclusions on the basis of data. It just happens that self-trackers collect data on themselves. As a community, I believe self-trackers offer an accessible vision of honest, objective introspection and assessment, something gravely lacking in this crazy world.

Second, while completely relevant to the general practice of data science, statistics as a formal practice is not relevant to the average self-tracker. This is because the tools of the statistical trade are hidden from the layman behind archaic jargon and traditions. There is therefore an opportunity to start from scratch and re-envision the guts of statistical science in a way that is accessible to the average self-tracker.

I propose a Coursera class, which I’m happy to teach, on something like “personal analytics.” In it, we learn how to design, implement, analyze, interpret, and draw confident conclusions from self-tracking data. Without jargon, we would learn to interpret data with a healthy dose of skepticism. All data is equal to truth (the thing we care about and wish we knew) plus error (random or otherwise), and the trick to being a confident data analyst is to learn how to separate truth from error, based on data.

I’d see the course as being very, very, very light on math and programming. As much as possible, I would keep the discourse non-technical and non-jargon-y. Our case studies and homeworks would be motivated by real-world self-tracking data. And since we’re starting from scratch, I’d envision us building the thing together, so that it optimally reflects our interests. I’m ready to get started if there is sufficient interest. Perhaps a target launch in September 2013?

If you are interested in a personal analytics course, please email me (my contact information is above). In the email, please provide the following information:

1. Name and contact information.

2. Would you be willing to share data with the class for group analysis?

3. Would you be interested in helping to design / administer the course? If so, in what capacity?

Look forward to hearing from you!

Related Posts Plugin for WordPress, Blogger...
Print Friendly
Measured Me Recommends:
Best Apps for Self-Tracking: rTracker and Track & Share
Product of the Month:
Inner Balance HRV and Stress Sensor



Buy directly from HeartMath or shop on Amazon

3 Responses to Guest Post: The Confidently-Quantified Self

  1. Ian Clements says:

    Yes; I’m interested.

    2. Would you be willing to share data with the class for group analysis? Yes

    3. Would you be interested in helping to design / administer the course? If so, in what capacity? Helping to design, yes; administer the course, probably not

    I have 5+ years of daily QS tracking in an Excel spreadsheet, covering 200+ independent lifestyle variables (exercise, CAM, supplements, drinks, foods, drugs); 100+ dependent variables (biochemical, body composition).

  2. Measured Me says:

    Hello Ian,

    I just forwarded your responses to Alan. Please contact him directly for any additional information about the course.

  3. Sameer Jain says:

    I am very interested. There are various things I’m interested in quantifying, but the uncertainty around how to minimize the cognitive burden of the tracking, as well as how to draw valid conclusions from the data given the vast # of confounding variables, has kept me from pursuing this.

    1. Name and contact information.
    Submitted through form.

    2. Would you be willing to share data with the class for group analysis?
    Yes

    3. Would you be interested in helping to design / administer the course? If so, in what capacity?
    Design – yes, me and my colleagues in PwC’s mobile health group would be interesting in contributing any input we can into the design of the course. Among these is the co-organizer of QS NYC; I’d be happy to get him involved as well.

Leave a Reply

Your email address will not be published. Required fields are marked *


8 + 7 =