5
$\begingroup$

I want to model mean colony size as a function of predator density. Within each treatment level the smaller colonies (3-5 individuals) are a lot more common than the large ones (up to 600 individuals). Which error distribution should I use?

$\endgroup$
  • $\begingroup$ Could you describe distribution of colonies sizes in greater detail? Is it symmetric or skewed etc.? $\endgroup$ – Tim Dec 14 '15 at 13:03
  • $\begingroup$ most of the treatment replicates look similar to this graph I found on google statit.com/images/skewed_histogram.gif $\endgroup$ – golgi Dec 14 '15 at 13:13
7
$\begingroup$

This is a partial answer only, but the graphical content makes a comment a poor alternative. In a comment, the OP talks about using sqrt(log()) as a transformation.

I'd advise against that on the general grounds that it is a very unusual and idiosyncratic transformation, so you will face puzzlement at all levels up to supervisors, examiners and paper reviewers. (Translate to the terminology of your own education and research set-up.)

I'd also advise that it really stretches out low colony sizes. Here is a plot for sizes 1(1)600.

enter image description here

Notice how on this transformed scale the interval from size 1 to size 5 is more than half the entire range from size 1 to size 600.

Implication: this transformation will create outliers for small colony sizes. The better fit observed could be an artefact of over-transforming.

I wouldn't go further than a log link. At least that is standard and easier to think about.

$\endgroup$
11
$\begingroup$

There's not enough information to offer a definitive answer.

Colony size would be a count, I gather (rather than say an area, or a mass).

It's better not to think of them as representing an error term but rather the conditional distribution of the response.

Some possible models include negative binomial or log series distribution -- these are two distributions in the exponential-family which could be suitable for such counts, though they're not the only possibilities (some care must betaken with the zero-case; for example the negative binomial used in GLMs starts from 0, but colony sizes are presumably bounded below by 1).

In some cases you may be able to get some use from the Poisson but I expect that typically it won't be nearly skewed enough.

(The log series distribution may not be implemented in most software.)

You may need to consider more complex models that straight exponential family. You might need to consider truncated or shifted distributions, or if zeroes are possible, perhaps zero-inflated or hurdle models; you may also need mixed effect GLMs for example.

$\endgroup$
  • $\begingroup$ Thanks for your help. I've found that transforming the data like this: lm(sqrt(log(y))~x, ) makes it fit the model better but I wrote the code for the graph with CIs based on an untransformed model: qplot(x, y, data= data, geom= c("point", "smooth"), method= "lm", )+theme_bw() Any ideas of how I could change it so the new model is incorporated? I could put the method as glm with a log identity in but the sqrt makes the model fit much better. $\endgroup$ – golgi Dec 14 '15 at 15:17
  • 1
    $\begingroup$ Its a count, sort of, but it is not really a sum of independent contributions, so the usual arguments leading to Poisson regression do not apply. $\endgroup$ – kjetil b halvorsen Jan 18 '16 at 19:23
9
$\begingroup$

About transformation, see :

O’Hara, Robert B., and D. Johan Kotze. 2010. « Do Not Log-Transform Count Data ». Methods in Ecology and Evolution 1 (2): 118‑22. doi:10.1111/j.2041-210X.2010.00021.x.

Summary :

1. Ecological count data (e.g. number of individuals or species) are often log-transformed to satisfy parametric test assumptions.

2. Apart from the fact that generalized linear models are better suited in dealing with count data, a log-transformation of counts has the additional quandary in how to deal with zero observations. With just one zero observation (if this observation represents a sampling unit), the whole data set needs to be fudged by adding a value (usually 1) before transformation.

3. Simulating data from a negative binomial distribution, we compared the outcome of fitting models that were transformed in various ways (log, square root) with results from fitting models using quasi-Poisson and negative binomial models to untransformed count data.

4. We found that the transformations performed poorly, except when the dispersion was small and the mean counts were large. The quasi-Poisson and negative binomial models consistently performed well, with little bias.

5. We recommend that count data should not be analysed by log-transforming it, but instead models based on Poisson and negative binomial distributions should be used.

$\endgroup$
  • 2
    $\begingroup$ +1 If this is you summarizing your own paper - hence the "we" in 4 and 5 - (then hey, hi and welcome ...) you should probably indicate that you're summarizing your own work, not quoting the paper without attribution (which it sort of looks like). [If instead that summary is a direct quote, you could indicate it by putting > at the start of each quoted paragraph, but the paper doesn't seem to have exactly those items, so I don't think it's that] $\endgroup$ – Glen_b Dec 17 '15 at 0:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.