The most useful statistical test that nobody knows

Image: Plaque commemorating Fisher on Inverforth House. Peter O’Connor via, CC BY-SA 2.0

Do you know Fisher’s method for combining P-values?  If you do, move along; I’ve got nothing for you. If you don’t, though, you may be interested in what’s surely the most useful statistical test that – despite the fame of Fisher himself – nobody knows about.

Fisher’s method is the original meta-analysis.  When I was a grad student, and nobody had heard of meta-analysis (or cell phones, or the internal-combustion engine), I had a supervisory committee member who liked to make strong statements.  One of his favourites was “A bunch of weak tests don’t add up to a strong test!”  Many of his strong opinions were right, but that one was wrong: of course they do!  Combining weak tests to make a strong one is precisely the point of meta-analysis.  It was Fisher’s method, when I showed it to him, that made my skeptical committee member eat his words.  But despite the intervening decades in which meta-analysis has become mainstream, Fisher’s method still seems little-known.  That’s a shame, because it’s easy and powerful.

Fisher’s method is a way to take multiple, independent tests of a null hypothesis and fuse them into a single strong test.  Those multiple tests can be from different experiments, of course, but more usefully they can be tests of different predictions from the null.  For example, the null hypothesis that insecticide X has no effect on plant growth can generate predictions suitable for t-tests (growth differs between treated vs. control plants), G-tests of association (treated vs. control plants differ in frequency of meeting a size threshold), or regression (plant growth rate changes with insecticide dosage).  Sure, it’s unlikely that you’d run all these yourself – but you might easily find each in the literature, along with others you couldn’t have imagined running.  Wouldn’t it be nice to be able to combine their independent results?

What makes Fisher’s method work is that no matter what the form of the prediction or the test applied to it, if the null hypothesis holds* then P values are distributed uniformly on [0,1].  Fisher’s insight was that this means twice the sum of the negative logs of those P values will have a χ2 distribution.  More precisely, for k tests:


Here’s a simple worked example.  Imagine four insecticide experiments:

  • two t-tests, P = 0.11 and P = 0.12
  • a G-test, P = 0.21
  • a regression, P = 0.08

Nothing significant, right?  Wrong.  Fisher’s method gives a test statistic of 16.8, with 8 degrees of freedom and a combined P = 0.03.  This shouldn’t shock you: while none of the individual tests have P below the (absolutist) threshold of 0.05, it’s unlikely that four experiments would get four smallish values in the absence of any real effect.

So it’s dead easy: find a bunch of published experiments testing the hypothesis, pick out their P values, run a very simple calculation, and bam – meta-analysis on the cheap!

Now, Fisher’s method isn’t a panacea.  It’s just as vulnerable to file-drawer effects as any other meta-analysis.  And its focus on P-values is a shortcoming as well as an advantage.  Fisher’s method offers no way to combine effect sizes.  Of course, it’s difficult to see how you could combine effect sizes across experiments that measure group differences, frequency differences, and slopes – awkward situations like this were exactly Fisher’s motivation.  Therefore, a significant result in Fisher’s method means you’ve detected a credible pattern, but not that you’ve measured its strength.  That’s fine, since we should always be remembering that a significant P value is the beginning, not the end, of an analysis.

So more people should know about Fisher’s method.  Perhaps it’s unbearably nerdy to have a favourite statistical test; but Fisher’s method is a strong contender as mine.  And when have I (blogging or otherwise) worried about being too nerdy?

© Stephen Heard ( June 7, 2016

*^And if all test assumptions are met, of course. Fisher’s method is powerful; it isn’t magic.


11 thoughts on “The most useful statistical test that nobody knows

  1. alexmikov

    Your right.. Its Nerdy to have a favourite statistical test… but then again I keep a copy of Sokal and Rolf, Biometry by my bed. This test was taught to me and used a lot at uni in 1984/85 but i was unaware of its history. Thanks.


  2. Nick Collins

    Probably should mention the condition that the p values must be independently determined. That is, you shouldn’t combine the p values for your 3 plant responses if the measurements were carried out on the same set of plants. Otherwise you need to think about some sort of Bonferroni correction.


    1. ScientistSeesSquirrel Post author

      Quite right, Nick!. That’s what I meant in the 1st sentence, 3rd paragraph, by “multiple independent tests of a null”. I think your language, “p values independently determined”, means the same thing – right?


  3. Andrew D. Steen (@drdrewsteen)

    Wow! This test gives me a way to use the ca. 1-gazillion independent, mostly >0.05 p-values I hadn’t figured out how to use in the paper I’m in the middle of writing up. Thanks much.
    Your post inspired me to read more on methods for combining p-values. Whitlock (2005, J. Evol Bio) argues that a weighted-z test is more powerful than Fisher’s method when sample sizes are unequal.


  4. phaneron0

    I find the paper by Art Owen more comprehensive KARL PEARSON’S META-ANALYSIS REVISITED

    By the way when I taught intro stats at Duke (2007/2008) the students seemed to have little difficulty undertaking and appreciating Fisher’s method (well at least appreciate why looking at individual p-values can be very mislead when multiple similar studies).

    Keith O’Rourke


  5. Pingback: Presenting statistical results: pointers in scientific writing | Scientist Sees Squirrel

  6. Pingback: Apresentando resultados estatísticos: Apontamentos em escrita científica – Mais Um Blog de Ecologia e Estatística

Comment on this post:

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.