• Go ahead, make my day

    Never mind what it is about, one of Paul Krugman’s recent posts includes this paragraph on instrumental variables (IV):

    Instrumental variables is a statistical technique that you use to avoid having your results contaminated by reverse causation — say, if stimulus funds were directed to states with especially severe unemployment problems, you might find a spurious negative correlation between stimulus and unemployment. What you need to get around this is some variable that is correlated with stimulus but not affected by the job changes; in effect, you use this other variable to create a predicted stimulus level, then look at how employment is affected by the predicted level, not the actual level. If I’ve just lost you, never mind.

    If, by virtue of my blogging on IV, a single reader isn’t lost who might otherwise be, I’ll be very happy. If that reader is you, maybe you’d be so kind to tell me in the comments. It’d make my day.

    While I’m asking, if any readers use this blog as part of college/university course curriculum or otherwise wish to express their appreciation, that will help. No, I’m not having a bad day. But you can make it a very good one.



    Comments closed
    • Count me in. I had a vague idea how IV studies worked before reading here. Now I understand them well enough to discuss one at dinner with someone who does stats for NEJM articles and not sound like a total idiot.


    • Austin – I find your thoughts and commentary very insightful and helpful while I work on finishing my dissertation. In regard to IV’s, I’ve come to conclude they are like gnomes. I hear a lot about their magical abilities and several reported sightings, but they are hard to find.

    • Austin, within just the last 2 weeks, a former student and 2 colleagues all told me how much they enjoy your blog. We are all amazed at how efficient at writing you must be.

    • As a young research analyst, I find your posts not only intelligent and informative, but also very accessible. You’re on my daily morning reading list, and I appreciate your insight.

    • A few months ago I added Paul Krugman to my Google Reader, mostly because you cite him so often. So, when I read this article, the first thing I thought of was how you have tried to explain exactly what an I.V. was.

    • I am a new assistant professor and I have found your posts to be extremely interesting and useful to deepening my understanding of health policy. I know that I have passed around your powerpoint slides on costs to other academics.

      I do not know if this helps or not but I find this a very useful blog.

    • Belated–but I want to say thank you for all the diligent work you have done explaining IV–and in particular its use in working through the issue of health outcomes in Medicaid vs. no insurance. (This weekend I finally made it through all your posts on this and the majority of the various links.) My understanding of statistics is minimal, and yet you have made the value of IV studies much more clear to me.