A Grco-Latin square in its pure form may not be very useful, but extension and replications of
Question:
A Grco-Latin square in its pure form may not be very useful, but extension and replications of it often prove to be quite useful. Consider the following design:
where rows, columns and Greek letters are blocking factors.
(i) What would you call this design? Give an appropriate linear model and outline the ANOVA table, giving sources of variation, d.f., and sums of squares. (ii) Give the SAS statements for analyzing data from such an experiment. (iii) Suppose you find out that for the experiment under consideration the "columns' represent animals. More specifically, columns 1-4 represent animals from one breed and columns 5-8 animals from a different breed. The researcher is interested in finding out whether differences among treatments are breed- specific. Is the design given above appropriate for investigating this question? If yes, explain why; if no, indicate what you would have done differently. (iv) For the design in (iii) give an appropriate linear model and outline the ANOVA table, giving sources of variation and d.f. (v) Give the SAS statements for the analysis suggested in (iv).
Step by Step Answer:
