A while back Jon Sutton at The Psychologist asked my opinion on the end of history illusion. This was sparked by an interesting Science paper by Quoidbach, Gilbert and Wilson. Blogger and mathematician Jordan Ellenberg had written a blog post arguing that the paper makes a mistake: "a somewhat subtle mistake, but a bad mistake, and one which kills a big chunk of the paper".
Jon wanted a second opinion, and after a bit of reading I replied that Ellenberg's criticisms were valid. I meant to blog about it at the time but got caught up in other things. Consequently I missed the BPS research digest piece on it.
The reason for writing this blog post is because the flaw that Ellenberg spotted is quite interesting in its own right and because both the description by Ellenberg and the description in the Research Digest article probably don't explain it clearly enough for some readers to appreciate. Ellenberg's piece is (I hasten to add) crystal clear but relies on a reader being comfortable with the formal, mathematical approach he takes (which many psychologists won't be). The Research Digest description just gives the brief gist (with a link to Ellenberg for the full picture). Here is my belated attempt at a psychologist-friendly interpretation with no formal notation - and as little maths as possible.
According to the end of history illusion people underestimate how much they will change in the future. For example, someone asked to predict how their personality would change in the next ten years would come up with a prediction closer to their original position than their actual position. Quoidbach et al. tested this mainly by asking people to predict future values on some psychological variable (e.g., a personality test score) and then showing that actual change is much greater than the difference between the original and predicted scores. This seems highly plausible, but Ellenberg pointed out that the difference in the predicted and original scores is a different quantity from the expected (absolute) change in scores.
Why is this? Perhaps the easiest way to understand is to work through a simple example. Imagine that my extraversion score is 50 on a scale that goes from 0 (extremely introverted) to 100 (extremely extraverted). A researcher then asks me to predict my extraversion score in 10 years time. I, being a keen observer of human nature (bear with me on this if you know me - it is just an example), am aware that personality is not fixed and judge that I am likely to change quite a bit - say 15 points - on the scale. However, I might get more extraverted or I might get more introverted (depending on how life treats me over the next ten years). Given that I'm in the middle of the scale, I could end with a score of 35 or a score of 65. Thus I predict that my extraversion score after 10 years will be (35 + 65)/2 = 50. It looks as though I've predicted zero change, when what I've done is give the best prediction I can (one that minimizes my prediction error). Had I instead been asked to give the absolute change I expected, my answer would have been different. It would have been (15 + 15)/2 = 15 (not zero).
Although the example is simple it captures the essence of the problem. Commenters on Ellenberg's blog looked again at the raw data that Quoidback et al. provided. According to their analyses the end of history illusion largely disappears when analyzed correctly (though only some of the data sets support such a reanalysis). Thus if the end of history illusion effect exists (and the basic premise seems highly plausible) it is quite probably a much smaller and more fragile effect than originally thought. That makes sense to me - because I'm not sure that such a bias could be both pervasive and large in the face of the counter-evidence available to people about past change in themselves and change in others.
My continued interest in the effect is slightly different. There seems to be a cognitive illusion at work here - one that makes the difference between the original score and predicted score appear to be a good measure of an entirely different quantity - the expected absolute change in score ...
Aug
17
I Will Not Ever, NEVER Run a MANOVA
I have been thinking to write a paper about MANOVA (and in particular why it should be avoided) for some time, but never got round to it. However, I recently discovered an excellent article by Francis Huang that pretty much sums up most of what I'd cover. In this blog post I'll just run through the main issues and refer you to Francis' paper for a more in-depth critique or the section on MANOVA in Serious Stats (Baguley, 2012).
Jan
19
A brief introduction to logistic regression
I wrote a brief introduction to logistic regression aimed at psychology students. You can take a look at the pdf here:
A more comprehensive introduction in terms of the generalised linear model can be found in my book:
Baguley, T. (2012). Serious stats: a guide to advanced statistics for the behavioral sciences. Palgrave Macmillan.
A more comprehensive introduction in terms of the generalised linear model can be found in my book:
Baguley, T. (2012). Serious stats: a guide to advanced statistics for the behavioral sciences. Palgrave Macmillan.
May
18
Serious Stats: Obtaining CIs for Spearman's rho or Kendall's tau
I wrote a short blog (with R Code) on how to calculate corrected CIs for rho and tau using the Fisher z transformation.
May
13
Serious stats: Type II versus Type III Sums of Squares
I have written a short article on Type II versus Type III SS in ANOVA-like models on my Serious Stats blog:
https://seriousstats.wordpress.com/2020/05/13/type-ii-and-type-iii-sums-of-squares-what-should-i-choose/
https://seriousstats.wordpress.com/2020/05/13/type-ii-and-type-iii-sums-of-squares-what-should-i-choose/
Sep
5
Egon Pearson correction for Chi-Square
I have just published a short blog on the Egon Pearson correction for the chi-square test. This includes links to an R function to run the corrected test (and also provides residual analyses for contingency tables).
The blog is here and the R function here.
The blog is here and the R function here.
Sep
15
Provisional programme: ESRC funded conference: Bayesian Data Analysis in the Social Sciences Curriculum (Nottingham, UK 29th Sept 2017)
Bayesian Data Analysis in the Social Sciences Curriculum
Supported by the ESRC’s Advanced Training Initiative
Venue: Bowden Room Nottingham Conference Centre
Burton Street, Nottingham, NG1 4BU
Booking information online
Provisional schedule:
Organizers:
Thom Baguley twitter: @seriousstats
Mark Andrews twitter: @xmjandrews
Supported by the ESRC’s Advanced Training Initiative
Venue: Bowden Room Nottingham Conference Centre
Burton Street, Nottingham, NG1 4BU
Booking information online
Provisional schedule:
Organizers:
Thom Baguley twitter: @seriousstats
Mark Andrews twitter: @xmjandrews
Jun
13
STOP PRESS Introductory Bayesian data analysis workshops for social scientists (June 2017 Nottingham UK)
The third and (possibly) final round of the workshops of our introductory workshops was overbooked in April, but we have managed to arrange some additional dates in June.
There are still places left on these. More details at: http://www.priorexposure.org.uk/
As with the last round we are planning a free R workshop before hand (reccomended if you need a refresher or have never used R before).
There are still places left on these. More details at: http://www.priorexposure.org.uk/
As with the last round we are planning a free R workshop before hand (reccomended if you need a refresher or have never used R before).
May
25
Serious Stats blog: CI for differences in independent R square coefficients
In my Serious Stats blog I have a new post on providing CIs for a difference between independent R square coefficients.
You can find the post there or go direct to the function hosted on RPubs. I have been experimenting with knitr but can't yet get the html from R Markdown to work with my blogger or wordpress blogs.
You can find the post there or go direct to the function hosted on RPubs. I have been experimenting with knitr but can't yet get the html from R Markdown to work with my blogger or wordpress blogs.
Add a comment