ANOVA: An Easily Accessible Executive Analytics Tool
Often the administrator is presented with two sets of results or observations to concur with or to make a decision upon. The single factor analysis of variance, or ANOVA to statisticians, informs the administrator whether the sets of data are similar or statistically different. ANOVA is readily available with minimal effort and at no cost to the executive as it is one of the embedded analytical tools in Microsoft Excel. The analytical procedure calculates a p statistic that indicates dissimilar sets of results or observations when lower than 0.05 or conversely indicates similarity of results when equal or greater than 0.05. The critical p of 0.05 is a generally accepted reference for testing statistical significance. The following hypothetical scenario that often plays out in real life illustrates the utility of this tool.
In a meeting with the administrator, Geena, the human service director said, “Ernesto, manager of the East Office, has a performance review coming up, but I’m bothered that his results have dropped lately. I’ve compared the East Office to the West Office in terms of each unit’s percent (%) completed initial assistance need forms (of new clients). As shown,
the eight month average completion of 22.9% at the West Office is higher than the 15.2% completion at the East Office. The results in other units reflect the drop in performance.”
The administrator entered the results in Excel and ran the ANOVA tool. The calculated p is equal to 0.001. (The Excel ANOVA analysis is shown at the end of this article.) “The West Office’s results are statistically higher than that of the East Office results since the p calculated by the ANOVA tool is lower than 0.05. Have you checked the client loading?” “No” said the director, “But I have the client load ratios.”
The administrator entered these sets of ratios in the ANOVA tool, resulting in a calculated p of 0.01. “Statistically, West Office staff assist more clients than East Office staff, suggesting why West Office staff completion rates were lower. The lower completion rate results do not prove that Ernesto’s performance has declined.”
The director enthusiastically said, “Thank you for the demonstration of the ANOVA tool. I am now thinking about finding out the role client loading has on the initial assistance form completion rate. With this tool, we might be able to improve our processes and reach the goal of 100% completion rate.”
The ANOVA analysis produced by Microsoft Excel analytical tools add-in:
