Someone asked how do you find peer-reviewed articles that use simple statistical tests so your students can see the structure of papers that report on statistical analyses. I do this a lot for this blog and for lectures that I give, so I wrote up some suggestions.
Good question. I think it’s great to see how statistics appear “in the wild” and I try to liberally sprinkle my lectures and my blog entries with examples from actual journal articles.
First things first. If you want to use an article, you should try to use one published under an open source license. Those of us who use peer-reviewed articles in our blogs and in our classes have greatly benefited from the wide availability and liberal re-use requirements that open source journals provide. Two very big sites for open source journals are
Biomed Central: http://www.biomedcentral.com/ and
PLOS (Public Library of Science): https://www.plos.org/
Both sites have a robust search feature and it’s easy to find articles that use a particular kind of statistic. Hint: when looking for a term like ANOVA, use the advanced search feature that limits the search to papers where ANOVA appears in the abstract. That’s a pretty good way to assure that your statistical test will be prominent in the paper itself.
So here’s a paper that uses a three factor ANOVA.
and here’s an article that uses a paired t-test
You’ll find hundreds of papers that use elementary statistics this way. The trick is finding a paper that does not require a lot of specialized scientific or medical knowledge. Sometimes you can get a hint just by looking at the title, but often you are forced to skim through the papers yourself to get a feel for the complexity of the topic.
I tend to stay away from genetics papers and immunology papers because I find they often are very technical and tough to read. Papers in Nursing journals, however, tend to be easier to follow because they usually talk about big things like patient care and not small things like biomolecular mechanisms. I also like papers that talk about the economic impact of medical interventions because everyone knows what money is.
This is a crude stereotype, however, and some nursing and economic impact papers are highly specialized and some genetics and immunology papers are accessible. I just have better luck with some topics than others.