Hello All:
I have a 2 x 4 factorial plus a control (9 treatments) in a RCBD. The first factor is qualitative (product application frequency) and the second factor is quantitative (product rate: 0.075, 0.15, 0.3 and 0.6). I am wondering what would be the most effective analysis. Currently I run the 2 x 4 factorial (omitting the control) to test for main effects and interactions of frequency and rate. Then I run some single d.f. orthogonal contrasts to partition trt SS to compare specific trts or groups of treatments. I have two questions:
Thank you,
James
I have a 2 x 4 factorial plus a control (9 treatments) in a RCBD. The first factor is qualitative (product application frequency) and the second factor is quantitative (product rate: 0.075, 0.15, 0.3 and 0.6). I am wondering what would be the most effective analysis. Currently I run the 2 x 4 factorial (omitting the control) to test for main effects and interactions of frequency and rate. Then I run some single d.f. orthogonal contrasts to partition trt SS to compare specific trts or groups of treatments. I have two questions:
1. I have seen some references that suggest nesting the factorial structure (frequency x rate) within another factor (product: applied or not applied). In this study, the entire factorial structure would only be within the "applied" level of product. The control would be the only treatment in the "not applied" level. Is this a valid analysis for a factorial plus control design?
2. Can I use the control plot (which is the equivalent as a "0" rate) along with the levels of the rate factor to use orthogonal polynomials for trend analysis? [(e.g., to test for linear, quad., cubic, quartic response to rate (0, 0.075, 0.15, 0.3 and 0.6)]I can post examples of my SAS code if needed. I'm just looking for some general input for now. Please let me know if I need to supply more information to help you better answer these question.
Thank you,
James
