Pseudoscience (1)

  I’ve been putting off post(s) on this topic because I just would like to ignore it but it’s just like a bad toothache that won’t go away and sooner or later you have to deal with it. The reason I’m starting today is a co-authored paper by Assistant Prof. Andreas Madestam, Boccioni University and Assistant Prof. David Yanagizawa-Drottwhich, Harvard Kennedy School which was mentioned in the Harvard Gazette and concludes that taking children to see a Fourth of July parade turns them into Republicans so all you progressives beware this insidious right wing plot to indoctrinate your offspring. You still have time before tomorrow to pack your bags and drive to some left wing commie blue state campground.

Abstract:

Do childhood events shape adult political views and behaviour? This paper investigates the impact of Fourth of July celebrations in the US during childhood on partisanship and participation later in life. Using daily precipitation data to proxy for exogenous variation in participation on Fourth of July as a child, we examine the role of the celebrations for people born in 1920-1990. We find that days without rain on Fourth of July in childhood have lifelong effects. In particular, they shift adult views and behaviour in favour of the Republicans and increase later-life political participation. Our estimates are significant: one Fourth of July without rain before age 18 raises the likelihood of identifying as a Republican by 2 percent and voting for the Republican candidate by 4 percent. It also increases voter turnout by 0.9 percent and boosts political campaign contributions by 3 percent. Taken together, the evidence suggests that important childhood events can have persistent effects on political beliefs and participation and that Fourth of July celebrations in the US affect the nation’s political landscape.

   Among the footnotes is a suggested comparison to “the effect of the Hajj pilgrimage on adult (Muslim) pilgrims attitude, beliefs and practices” and arguments that “July the Fourth and similar public rituals can be rationalized as a common-knowledge generating co-ordinating mechanism that allows people to submit to a social or political authority.” So your kids turn into gun toting fundamentalist fanatics who submit to conservative ideology and only watch Fox news.

 There isn’t enough information in the truncated form of the paper to make a complete analysis but I do have some questions. The authors stated that the effect is in the shape of an inverted U with respect to age. This makes some sense. Children under 5 won’t understand the emotional implications other than an immediate response and teenagers are probably pretty jaded. Let’s say that we are left with 10 Fourth of July events which have maximal effect out of 15 in the study (3-18) Are the effects additive or non linear? The equation given in the paper has eight co variables to control for such things as age cohort, geographical location, race, gender, family income, education etc. What is the degree of imprecision in each variable given that the total imprecision is the sum of the individual imprecision. How do you assign someone to a particular race? Self identification or external assignment? After the authors control for these factors they do a regression analysis to calculate the degree of correlation which they state is statistically significant.

   A correlation is not causality. If you plot the degree of obesity geographically versus the degree of conservatism in the States, there is a strong correlation. You can’t infer that being porky makes you conservative or being conservative makes you fat.

  They don’t distinguish between variables (race, gender, education etc.) that are extraneous and those that are confounding. "An extraneous variable is a variable that MAY compete with the independent variable in explaining the outcome of a study.  A confounding variable (also called a third variable) is an extraneous variable that DOES cause a problem because we know that it DOES have a relationship with the independent and dependent variables. A confounding variable is a variable that systematically varies or influences the independent variable and also influences the dependent variable." The authors state that the co variables are not individually statistically significant and the use of an F-test demonstrates that the co variables are not jointly statistical significant with respect to the effect caused by an assumed and not measured exposure to F. of J. events. Note also that the education and family income variables are current at the time of study and not those existing at the time of the F. of J. events so the time dependency is not accounted for in their data.

  The caveats were not mentioned in media reports. They assumed that the effect on the individuals did not vary with age in their calculations although they stated the response was in an age related inverted U. If the effects were independently calculated for each age, the data was too “noisy” i.e. the imprecision was too high for statistically valid results. They also stated that the degree of adult conservative affiliation was self reported so I would assume that a primary variable had a lot of wiggle room in it which belies all their mathematical cleverness. This is called in stats parlance reactivity. Reactivity is defined as an alteration in performance that occurs as a result of being aware of participating in a study. In other words, reactivity occurs sometimes because research study participants might change their performance because they know they are being observed. You also have interpretative validity. Interpretive validity is present to the degree that the researcher accurately portrays the meanings given by the participants to what is being studied.

  They stated that the study was an investigation into “group identity formation.” (Clarissa – comments?) I would say that I made a very cursory analysis but from the Fox video the paper gives certain parties grist for their viewpoints without the important nuances and caveats which are critical. Given other recent news reports such as the post I gave on the paper, debt gives students higher self esteem, what else is new. It would also be interesting to do a control study in Canada with Canada Day. Is this effect specific to America? I would say that this represents pseudoscience given all the assumptions, variables and methodological approach. Comments?