NAME: NCAA Basketball Tournament Data TYPE: Census SIZE: 1706 Observations, 5 Variables (3 numeric, 2 character) DESCRIPTIVE ABSTRACT: This data set contains every NCAA Basketball Tournament game ever played. The tournament has been held every year since 1939. SOURCES: Years 1939-1990 were obtained from "The Encyclopedia of the NCAA Basketball Tournament", by Jim Savage, Dell Publishing, NY,NY (1990). Years 1991-1995 were obtained from selected major U.S. newspapers. VARIABLE DESCRIPTIONS: Year Columns 1-4 Winning School Columns 6-24 Winning Score Columns 25-27 Losing School Columns 29-47 Losing Score Columns 48-50 SPECIAL NOTES: The last game listed for each year is the championship game. STORY BEHIND THE DATA: Author had some free time, and was interested in trends over time in the scores of games, as well as how individual schools have fared over the years. PEDAGOGICAL NOTES: 1) Excellent for teaching course in SAS (from my experiences), in terms that it allows for interesting subsetting examples. (All games lost by NORTH CAROLINA, all championship games, all games where winning team scored over 100, etc.) 2) Allows for comparisons (by decade, before and after rules changes, etc) 3) The Total score of the game is a variable whose distribution is well modelled by a Normal distribution (does well on Pearson goodness of fit test, not well on smooth test), so allows for neat sampling examples, where each student could take there own sample and inferential procedures can be used 'in repeated sampling'. 4) A subject that virtually anyone can relate to. SUBMITTED BY: Larry Winner Department of Statistics, University of Florida 103 Griffin/Floyd Hall, Gainesville, FL 32611 winner@stat.ufl.edu