Statistical techniques in ophthalmic journals
M. S. Juzych, D. H. Shin, M. Seyedsadr, S. W. Siegner and L. A. Juzych
Wayne State University School of Medicine, Detroit, MI 48201.
Over the years, the use of statistics to evaluate experimental data in
ophthalmology has increased. The present study sought to assess the
frequency and types of statistical techniques used in ophthalmic journals.
We reviewed 974 original articles from the Archives for 1970, 1980, and
1990; the American Journal of Ophthalmology for 1990; and Ophthalmology for
1990. Of the 592 articles reviewed for 1990, 391 (66.0%) contained
statistics, with measures of central tendency most commonly used (385
articles [65.0%]), followed by dispersion (298 [50.3%]), t test (120
[20.3%]), and contingency tables (98 [16.6%]). A reader familiar with 10
statistical techniques would have "statistical accessibility" to 526
(88.9%) of 1990 articles. A statistically significant difference was found
in the percentage of articles containing statistical methods among the
journals (P = .0003; Archives, 75.3%; Ophthalmology, 66.8%; and American
Journal of Ophthalmology, 55.2%).
