School
University of GuelphDepartment
Marketing and Consumer StudiesCourse Code
MCS 3030Professor
Tanya MarkChapter
12Chapter 12 Research Methods
Pg. 235237
123 EXPERIMENTAL ANALYSIS
 The simplest experimental design, twogroup posttestonly randomized experiment, is usually analyzed
with the simple ttest
 The factorial experimental designs usually analyzed w/ ANOVA model
 Randomized block designs use special form of the ANOVA blocking model that uses dummycoded
variables to represent the blocks
123A The 2group posttestonly Randomized Experiment
Analysis need to meet following requirements…
Has 2 groups
Uses postonly measure
Has 2 distributions (measures),e ach w/ an average and variation
Assesses treatment effect as the statistical (no chance) difference between the groups
3 options to estimate treatment effect:
1. Compute an independent ttest
2. Compute 1way ANOVA between 2 independent groups
3. Use regression analysis to regress the posttest values into a dummycoded treatment variable
123B Factorial Design Analysis”
See page 236
123C Randomized Block Analysis
