JMP platforms and statistical methods

Distribution:
Histogram, summary statistics, confidence intervals for promotions

Graph Builder:
Comparative dot plots, summary statistics, confidence intervals for means, scatter plot, linear regression

Multivariate:
Correlations, scatter plot matrix

Fit Y by X:
One-factor ANOVA, one-variable regression

Fit Model:
Multiple variable regression, stepwise (forward selection and backward elimination)

Objective

Use ANOVA and multivariable regression to describe and compare the yield of different seed varietals in an experiment while controlling for the effects of other environmental factors.

The primary objectives of the analysis are to:

1. Determine if sufficient statistical evidence exists demonstrating differences in yield between the seed varietals and to quantify those differences.
2. Determine if other growing conditions variables (soil quality, fertilizer, % sun, rainfall, and number of irrigations) have a significant effect on yield and can be accounted for in the estimates comparing yield between the seed varietals.
3. Determine if the effect those other variables have on yield is different based upon the seed varietals.

Problem statement

Agricultural companies continuously work on developing seed varietals that improve the performance of a crop, such as yield, plant health, resistance to disease and pests, temperature tolerance, among others.

A seed breeding research team for one such company is interested in comparing the yield between three corn seed varietals they produce (standard, high yield, and high yield plus).

Data from 161 fields throughout the midwestern U.S. was gathered: 47 fields used the standard seed, 70 used the high yield varietal, and 44 used the high yield plus varietal.