Tools › Quality › Design of Experiments

Design of Experiments

1 tool in the Quality area.

ANOVA and DOE

Analysis of variance and two-level factorial experiments from pasted data: one-way ANOVA with the ANOVA table, F and p-value, R², the Brown–Forsythe (modified Levene) and Bartlett tests of equal variances, group means with confidence intervals, Tukey HSD and Fisher LSD pairwise comparisons; two-way ANOVA with replication and the interaction term, the cell-means table, main-effect and interaction plots and Tukey comparisons of the factor means; and 2^k factorial designs (two to five factors) with effects and coefficients by the contrast method, sums of squares, the ANOVA for a replicated design or Lenth's pseudo standard error with the half-normal plot for a single replicate, the fitted regression model in coded units and the residuals. Residual-versus-fitted and normal probability plots for every analysis. The default case is the plasma etching experiment of Montgomery (Design and Analysis of Experiments, Example 3.1), whose printed ANOVA table and Tukey and LSD values the page reproduces; the battery-life (Example 5.1) and filtration-rate (Example 6.2) examples load with the other two analyses.

4 worked examples reproduced, listed on the tool page.

Other Quality sections

Control · Measurement · Improvement · Inspection and Metrology