PubMed · 7710451
Conventional statistics and useful statistics.
Abstract
Differences between conventional statistical methods and more useful, modern methods are demonstrated using a statistical analysis of data from therapeutic research in rheumatology. The conventional methods, t-test and graphs of mean values and the boxplot, detect almost no differences between treatment groups. A more recent procedure for analysing group differences is the Wilcoxon-Mann-Whitney test. The associated graphs are based on the cumulative distribution function of the two treatment groups and the synthetic Receiver Operating Characteristic (ROC). Special differences, namely baseline dependencies, can be visualized in this way.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
V W Rahlfs. 1995. Conventional statistics and useful statistics.. https://pubmed.ncbi.nlm.nih.gov/7710451/
Cite the original work for its findings. Save a collection to share your selection of sources.