A couple months ago I posted an entry on this blog announcing Webology Solutions' Drupal v Joomla Survey. Well now the results are in.
We had nearly 200 reponses from professional web developers versed in Drupal, Joomla, Wordpress and a number of other content management systems. The results are available in two formats. We have posted a pdf file containing summary results for each question asked in the survey. We have also posted the detailed data in a comma separated file available for download (Personally identifying data has been removed). The results of the survey are now posted on our blog. You can download the files from our download section at WebologySolutions.com.
In the following weeks we will be posting analysis of the data, discussing issues such as what functions each CMS supports best and which CMS makes it easier to develop custom applications.
Comments
Drupal claims an install base
Drupal claims an install base of 250,000 sites, Joomla claims (depending on source) 10 to 30 million sites. A survey of 200 random users has no statistical relevance whatsoever.
Sample Size and Margin or Error
While I agree that larger sample sizes are better, to say a sample of 200 has no statistical value is incorrect. When the sample size is small relative to the population you can estimate the margin error without reference to the population size. For a 95% confidence interval, the formula is SQRT((p*(p-1))/n)*1.96. Where ‘n’ is the sample size and ‘p’ is the proportion of people that answered a question a certain way.
The margin of error for n=200 and p=.5 is about 6.3%. Not as good as good as some presidential polls but still valuable. By the way, some polls in the last election sampled a few hundred people to estimate how over 100 million people would vote. So while no survey is perfect, I think our survey does add some value to the Drupal vs. Joomla debate.
One last point, our survey is not of users at large but of professional web developers. I think our respondents’ expertise gives more weight and a unique perspective to our survey.
Project home page traffic
A quick look unique visitors to each project website:
http://siteanalytics.compete.com/drupal....
I some how doubt two projects with similar traffic to their project websites, have a two orders of magnitude difference in installations. But it's possible.
A quick look unique visitors to each project website
http://www.google.com/trends?q=joomla,+d...
You'd need to take into
You'd need to take into account the different local communities Joomla has. For example, the Dutch community site rivals the main project site in visitor count.
I've no idea if Drupal has something similar.