Dunnett’s Test

Basic Concepts

Dunnett’s test is used when we want to compare one group (usually the control treatment) with the other groups. In this case, Dunnett’s test is more powerful than the other ANOVA post-hoc tests described on this website since fewer tests are performed.

The test is similar to Tukey’s HSD, except that instead of testing

image1193

we test whether

image272

where n = size of the group samples, o = mean of the control group, j is the mean of any other group and td is the (two-tailed) Dunnett’s critical value given in the Dunnett’s Table. The table contains the values td(k, dfW, α)  where k = the number of groups (treatments) including the control.

The test can also be used when the group means are not equal via

image273

although, since Dunnett’s Table is based on equal group sizes, the above formula is only accurate if the group sizes are not too different.

Worksheet Function

Real Statistics Function: The following function is provided in the Real Statistics Resource Pack:

DCRIT(k, df, α, interp) = the critical value td for k independent variables, the given degrees of freedom, and the value of alpha. If interp = TRUE (default) the recommended interpolation is used; otherwise linear interpolation is used.

DPROB(x, k, df, iter, interp, txt) = estimated p-value for the Dunnett’s test at x for the distribution with k groups, degrees of freedom df, iter = # of iterations in estimating the p-value from the Dunnett’s table of critical values (default = 40) and interp = TRUE (default) for recommended interpolation and FALSE (linear interpolation).

Note that when txt = FALSE (default) if the p-value is less than .01, DPROB is rounded down to 0, while if the p-value is greater than .1 it is rounded up to 1. When txt = TRUE, then the output takes the form “< .01” or “> .1”.

Data Analysis Tool

Real Statistics Data Analysis Tool: We now show how to use Dunnett’s Test data analysis tool to address Example 1.

Example 1: Assuming that Method 1 is the control group in Example 2 of ANOVA Basic Concepts, compare the mean of this method with the means of the other methods using Dunnett’s test. The data is repeated in Figure 1.

We can perform this test using the Single Factor ANOVA tool. When the dialog box shown in Figure 1 of ANOVA Analysis Tool appears, fill in the Input Range with A3:D11, make sure that the Column headings included with data is checked, and choose the Dunnett’s option. After clicking on the OK button, the output shown on the right side of Figure 1 is displayed.

Dunnett's post-hoc test

Figure 1 – Dunnett’s Test

We see that there is a significant difference between Method 1 (the control) and 2 as well as between Method 1 and Method 3.

Note that cell I11 contains the formula =DCRIT(COUNT(I4:I7),H11,L2). Thus the critical value in Dunnett’s Table when k = 4, df = 28, and α = .05 is 2.483.

Examples Workbook

Click here to download the Excel workbook with the examples described on this webpage.

Reference

Howell, D. C. (2010) Statistical methods for psychology (7th ed.). Wadsworth, Cengage Learning.
https://labs.la.utexas.edu/gilden/files/2016/05/Statistics-Text.pdf

10 thoughts on “Dunnett’s Test”

  1. Hi Rodney
    I’m using the One Factor ANOVA and ticking of Dunnett’s test, but I only get an output for D-test and not T-test like you. I’ve tried doing it with your example workbook with the same results. What am I doing wrong?

    Kind regards,
    Anna

    Reply
    • Hello Anna,
      You aren’t doing anything wrong. I changed the label from T Test to D-TEST in one of the last releases. The values haven’t changed, just the label. Sorry that this caused you trouble, but I appreciate your flagging this error so that I could correct it.
      I have now corrected the webpage so that hopefully it won’t cause confusion for others.
      Charles

      Reply
  2. Has there been any progress in understanding why the RealStats add-in appears to be incompatible with Excel 2007? I have carefully followed all of your advice on installation etc., tried getting it to work over and over again, but I always get the message that “Excel experienced a serious problem with the ‘realstats-2007’ add-in… you should disable this add-in and check to see if an update is available.” If there is anything new that I could try, please let me know. BTW, the data analysis add-in works fine, so I suspect it’s some security issue in realstats.

    Reply
    • Hi Rodney,
      Microsoft has introduced a number of capabilities since Excel 2007. I have used many of these capabilities in the Real Statistics code. Since most users no longer use Excel 2007 and because Microsoft no longer supports Excel 2007, I have no plans to make the latest releases of Real Statistics compatible with Excel 2007. I still maintain an old release of the Real Statistics software that is compatible with Excel 2007.
      Charles

      Reply
  3. Hi Charles,
    I’m currently using the Data Analysis Tool in Excel for Mac and it seems that the functions for obtaining the critical values and p-values aren’t working. The analogous functions for Tukey’s HSD do work, but for some reason I can’t get them for Dunnett’s Test, and I need to run Dunnett’s Test to follow up Welch’s ANOVA

    Reply
      • Hello Antonio,
        Probably Dunnett’s test can be performed for k > 21, but the largest table of critical values that I have found is limited to k <= 21. Charles

        Reply
    • Hi Antonio,
      I just checked that the DCRIT and DPROB functions are working in the Windows version of the Real Statistics software.
      When you say that these functions are not working on the Mac are you referring to the cases where k is less than or equal to 21 or only where k is larger than 21?
      Charles

      Reply

Leave a Comment