The ANOVA: Two-Factor Without Replication Tool If you want the richer analysis available from the least squares approach to ANOVA, you're better off with LINEST() or the Data Analysis add-in's Regression tool. The single factor tool can be a quick and handy way of running a one-way ANOVA, especially if you're primarily interested in F ratios and probability levels. The Data Analysis Add-in's ANOVA ToolsР”В®xcel's Data Analysis add-in includes three tools that perform ANOVAs: They have real consequences for the probability statements that you want to quantify using the analysis of variance.

These labels-nested versus crossed, random versus fixed-are not just fussy distinctions without a difference. Factors whose levels exhaust the levels of interest, such as Male patients versus Female patients, are termed fixed factors. In that case, Hospital is a random factor. If Hospital is a factor in an experiment, it is very likely that the experimenter wants to generalize the findings to other hospitals not included in the experiment.

Random factors that comprise levels that are considered random selections from a larger population.

Factors whose levels do appear at every level of another factor are termed crossed factors. An example is hospitals that provide, or do not provide, a particular treatment.

Nested factors that have levels that do not appear at every level of another factor.

In the first article in this two part series, Nested and Random Factors in ANOVA, I discussed two special types of factors in the analysis of variance: Statistical Analysis Using Excel LiveLessons (Video Training)

Split plot design in xlstat