Simultaneous Comparisons of Multiple Treatments to Two (or More) Control

✍ Scribed by D. R. Hoover

• Publisher: John Wiley and Sons
• Year: 1991
• Tongue: English
• Weight: 483 KB
• Volume: 33
• Category: Article
• ISSN: 0323-3847
• DOI: 10.1002/bimj.4710330803

✦ Synopsis

DIJNNETT (1955) developed a procedure simultnneously comparing k treatments to one control with an exact overall type I error of a when all sampling distributions are normal. Sometimes it is desirable to compare k treatments t o m g 2 controls, in particular t o two controls. For instance, several new thernpies (e.g., pain relievere) could be compared t o two standard therapies (e.g., Aspirin a n d Tylenol). Alternatively, a standard thernpy could be very expensive, difficult to apply and/or have bad side effecte, making i t uaeful to compare each new therapy to both standard therapy a n d no therapy (Placebo).

Diinnett'e method is expanded here t o give cornparisone of mean values for k treatments to mean values for m t 2 controls nt an exnct overall type I error of a when all sampling distributions are normal. Tabled values needed to make exact simultaneous cornparisone at a=.O5 are given for m = 2 . An application ie mnde to a n example from the literature.