[ont.events] Statistics Seminar: Two-Level Designs of Resolution III* and V

ruth@utstat.uucp (Ruth Croxford) (02/23/89)

Topic:     Two-Level Designs of Resolution III* and V
Speaker:   Dennis K.J. Lin, Dept. of Statistics, University of Toronto
Date:      11:00 a.m., Wednesday, March 1, 1989
Place:     Sidney Smith Hall, Room 2130, 100 St. George St. University of Toronto
Abstract:

     The problem of how many factors, k, can be accommodated in a two-level 
fractional factorial design of resolution V with a specified number of runs, 
N=2^q say, has been solved only for q <= 8.  (See Box and Hunter, 1961; Addelman,
1965; Draper and Mitchell, 1967).  We investigate the extension of these results
to 9<=q<=12, using resolution III* designs as intermediates.  Resolution III*
designs are resolution III designs in which no two factor interactions are
confounded with one another.  They arise naturally in another context, in the 
search for small composite designs, and were first suggested by Hartley (1959).
Resolution III* designs are here linked to resolution V designs, and some new
resolution V designs are thus obtained.  This is joint work with Norman R. 
Draper.
------------

Coffee and tea will be served in the De Lury Lounge (SS6006) at 10:30 a.m.