A Survey of Strong Inference in Ecology Papers: Platt’s Test and Medawar’s Fraud Model

In 1897 Chamberlin wrote an article in the Journal of Geology on the method of multiple working hypotheses as a way of experimentally testing scientific ideas (Chamberlin 1897 reprinted in Science). Ecology was scarcely invented at that time and this has stimulated my quest here to see if current ecology journals subscribe to Chamberlin’s approach to science. Platt (1964) formalized this approach as “strong inference” and argued that it was the best way for science to progress rapidly. If this is the case (and some do not agree that this approach is suitable for ecology) then we might use this model to check now and then on the state of ecology via published papers.

I did a very small survey in the Journal of Animal Ecology for 2015. Most ecologists I hope would classify this as one of our leading journals. I asked the simple question of whether in the Introduction to each paper there were explicit hypotheses stated and explicit alternative hypotheses, and categorized each paper as ‘yes’ or ‘no’. There is certainly a problem here in that many papers stated a hypothesis or idea they wanted to investigate but never discussed what the alternative was, or indeed if there was an alternative hypothesis. As a potential set of covariates, I tallied how many times the word ‘hypothesis’ or ‘hypotheses’ occurred in each paper, as well as the word ‘test’, ‘prediction’, and ‘model’. Most ‘model’ and ‘test’ words were used in the context of statistical models or statistical tests of significance. Singular and plural forms of these words were all counted.

This is not a publication and I did not want to spend the rest of my life looking at all the other ecology journals and many issues, so I concentrated on the Journal of Animal Ecology, volume 84, issues 1 and 2 in 2015. I obtained these results for the 51 articles in these two issues: (number of times the word appeared per article, averaged over all articles)

Explicit hypothesis and alternative hypotheses

“Hypothesis”

“Test”

“Prediction”

“Model”

Yes

22%

Mean

3.1

7.9

6.5

32.5

No

78%

Median

1

6

4

20

No. articles

51

Range

0-23

0-37

0-27

0-163

There are lots of problems with a simple analysis like this and perhaps its utility may lie in stimulating a more sophisticated analysis of a wider variety of journals. It is certainly not a random sample of the ecology literature. But maybe it gives us a few insights into ecology 2015.

I found the results quite surprising in that many papers failed Platt’s Test for strong inference. Many papers stated hypotheses but failed to state alternative hypotheses. In some cases the implied alternative hypothesis is the now-discredited null hypothesis (Johnson 2002). One possible reason for the failure to state hypotheses clearly was discussed by Medawar many years ago (Howitt and Wilson 2014; Medawar 1963). He pointed out that most scientific papers were written backwards, analysing the data, finding out what it concluded, and then writing the introduction to the paper knowing the results to follow. A significant number of papers in these issues I have looked at here seem to have been written following Medawar’s “fraud model”.

But make of such data as you will, and I appreciate that many people write papers in a less formal style than Medawar or Platt would prefer. And many have alternative hypotheses in mind but do not write them down clearly. And perhaps many referees do not think we should be restricted to using the hypothetical deductive approach to science. All of these points of view should be discussed rather than ignored. I note that some ecological journals now turn back papers that have no clear statement of a hypothesis in the introduction to the submitted paper.

The word ‘model’ is the most common word to appear in this analysis, typically in the case of a statistical model evaluated by AIC kinds of statistics. And the word ‘test’ was most commonly used in statistical tests (‘t-test’) in a paper. Indeed virtually all of these paper overflow with statistical estimates of various kinds. Few however come back in the conclusions to state exactly what progress has been made by their paper and even less make statements about what should be done next. From this small survey there is considerable room for improvement in ecological publications.

Chamberlin, T.C. 1897. The method of multiple working hypotheses. Journal of Geology 5: 837-848 (reprinted in Science 148: 754-759 in 1965). doi:10.1126/science.148.3671.754

Howitt, S.M., and Wilson, A.N. 2014. Revisiting “Is the scientific paper a fraud?”. EMBO reports 15(5): 481-484. doi:10.1002/embr.201338302

Johnson, D.H. (2002) The role of hypothesis testing in wildlife science. Journal of Wildlife Management 66(2): 272-276. doi: 10.2307/3803159

Medawar, P.B. 1963. Is the scientific paper a fraud? In “The Threat and the Glory”. Edited by P.B. Medawar. Harper Collins, New York. pp. 228-233. (Reprinted by Harper Collins in 1990. ISBN: 9780060391126.)

Platt, J.R. 1964. Strong inference. Science 146: 347-353. doi:10.1126/science.146.3642.347

10 thoughts on “A Survey of Strong Inference in Ecology Papers: Platt’s Test and Medawar’s Fraud Model

  1. Jeremy Fox

    This is a *great* idea for a post–I’m jealous! And I’m depressed but not surprised by the answer you got.

    There are a few topics in ecology and evolution on which there’s a widely-agreed “checklist” of boxes one has to tick off to confirm some hypothesis and rule out the alternatives. Character displacement is one prominent example, and bet hedging is another: https://dynamicecology.wordpress.com/2014/03/31/the-power-of-checking-all-the-boxes-in-scientific-research-the-example-of-character-displacement/ But such checklists are not only rare, all the boxes in them are rarely ticked off even when they do exist.

    Reply
  2. Dan Zoppellini

    Charley,

    I’m not an Ecologist but I am most certainly interested in Ecology, especially where it ‘meets’ forestry and agriculture. The idea of making clear alternative hypothesis as part of a study seems necessary to me. I had a quick look through Google Scholar in attempt to find Chamberlin (1897) and came across the abstract to this paper:

    http://geology.gsapubs.org/content/18/1/44.short

    Do you think that Johnson is right to say that Chamberlin’s methods are unrealistic?

    Reply
    1. Charles Krebs Post author

      Dan – I think Chamberlin was far ahead of his time and still is. If Chamberlin’s methods and the general approach of hypothetical-deductive science is not the way to go in science, then the question is what is the alternative? I think one could say that Chamberlin’s approach would be unrealistic in a science that is completely new in a nothing-known, completely descriptive stage, but that hardly is the case today with physics, chemistry, and biology, and I would argue that especially in applied sciences like agriculture, forestry and wildlife management Chamberlin is needed more than ever.

      Reply
      1. Dan Zoppellini

        Thank you for your response, Charley. Again, I understand and agree with what you’re saying. To my mind, adopting a multi-hypothesis format can only facilitate explaining observed phenomena or interpreting recorded data. Likewise, without having multiple explanations as to why something has happened, would you not also limit the possible understanding of other relevant associations?

        Reply
        1. Charles Krebs Post author

          Dan – I agree that the list of multiple working hypotheses may be incomplete, missing what is a key causal factor, and I think this happens again and again in ecology and science in general. I think we always have to be alert for this possibility and seek additional alternatives. For a simple example, if you think a particular grassland’s productivity is being affected either by grazing or by nutrient deposition in rainfall, you may not have in your hypothesis group the effects of ground level ozone on plants. So we should think as broadly as possible, but always realize we may have missed the boat. The history of the causes of disease illustrates this so well.

          Reply
  3. Pingback: Friday links: weak inference in ecology, the Ambiguous Pazuma, you vs. lunch, and more | Dynamic Ecology

  4. Jeff Ollerton

    Great post that raises some interesting points about how we do ecology, and science more generally. I thought I’d share an experience with you which relates to the question of multiple hypotheses. A few years ago we published the following paper in Oecologia:

    http://oldweb.northampton.ac.uk/aps/env/lbrg/journals/papers/Ollerton-NicheOverlap-Oecologia2007.pdf

    During the review process one of the referees commented that s/he wanted us to get rid of the multiple, ensuing hypotheses because “it reads like an undergraduate lab report”. Fortunately I didn’t have to defend our approach because the editor directed me to ignore that comment. Clearly he disagreed with the reviewer.

    But it does suggest some resistance within the ecological community to embrace multiple hypotheses. And I wonder why that is? Is it because we like nice neat, easily testable science? Or do ecologists just not see enough of it to realise that it’s a useful way to conduct our research?

    Reply
    1. Charles Krebs Post author

      Jeff – good comments and your experience with this reviewer strikes me as indicating we need a whole blog site on “Chamberlin Awards for Bizarre and Zany Comments in Manuscript Reviews”. There is resistance to multiple working hypotheses from some who say we do not know enough yet to use this approach, but perhaps the underlying flaw is the lack of clear thinking, which is very hard work to be sure. Your story reminds me of one comment from a reviewer of one of my textbooks who complained that I used metric units and I should be using English units so the students would understand….

      Reply
  5. falko

    Interesting post, I agree with most of it.

    But I’d like to stick up for ecologists a bit by posing an alternative explanation for the shortage of multiple alternative hypotheses in ecological studies (fitting, considering the topic of this post, no?). Couldn’t it simply be that the tendency of journals to limit the lengths of manuscripts (added to the pressure on authors to publish more papers) has meant that tests of the alternative hypotheses are spread over multiple manuscripts?

    It may well be that broader ecological research programs are following Platt’s route to ‘strong inference’, but this just isn’t obvious from individual research articles…

    Reply

Leave a Reply

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