EN
PEL GUSOK)

REPEATED-MEASUREMENT AND OTHER
EXPERIMENTAL DESIGNS

19.1 INTRODUCTION

In previous chapters the basic ideas involved in the design, analysis, and
interpretation of one-way classification experiments, and factorial experi-
ments, were discussed in considerable detail. In psychology and education
considerable use is made, both through choice and necessity, of other
experimental designs. "The purpose of this chapter is to introduce the
reader in an elementary way to a number of other experimental designs in
common use. Some of these designs involve assumptions, and present
problems, that are not involved in designs previously discussed. Some
awareness and understanding of these assumptions and problems are es-
sential to the proper use of these designs.

Research workers in psychology and education make freguent use of
experimental designs in which measurements are repeated a number of
times on the same subjects. "These designs, and the assumptions un-
derlying their use, are described in some detail in this chapter. Ran-
domized block designs, designs with nested factors, and Latin sguare
designs are also described.

The treatment of these designs in this chapter must of necessity be ele-
mentary. Some of these designs can be combined and extended in a vari-
ety of ways leading to experiments of much complexity. Because of the
ready availability of computers for data analysis, the current trend in psy-
chology and education is toward more complex designs. Ouestions can be
raised regarding the merits of this trend toward complexity. A much more
advanced treatment of the topics discussed in this chapter will be found in
books by Winer (1971), Myers (1972), and other authors.