RES/710 v5
Course Project Worksheet – Using Excel for
Statistical Testing
Last week’s assignment required a review of various statistical tests. Earlier this week, in Discussion 1,
you explored which test might be most appropriate given your chosen variables. Now is your time to
practice running the statistical test and analyzing the data. Complete each of the items below:
1. Which statistical test (T-Test, ANOVA, Chi-Square, or Regression) did you choose for your
variables?
[Enter your response here.]
2. Explain why you chose this test.
[Enter your response here.]
3. Next, using Excel (refer to the handouts from this week for step-by-step instructions), run the
statistical test you chose above and paste the output below.
[Paste Excel output here.]
4. Analyze the results.
[Enter your response here.]
References
[List references according to APA guidelines.]
Copyright 2021 by University of Phoenix. All rights reserved.
RES/710 v5
Measures of Association and Tests of
Significance in Excel
T-Test in Excel
First perform an f-test to see if your data has equal or unequal variances by selecting Data > Data
Analysis Tool > F-Test Two-Sample for Variances.
Be sure to select the Labels box if your highlighted area includes labels.
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 2 of 10
Your output should look something like this:
Here we have unequal variances so we would choose t-Test: Two-Sample Assuming Unequal
Variances.
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 3 of 10
Next, highlight the cells we are using to perform the t-test. We are hypothesizing that there will be no
difference in the mean, so we set that to 0.
Sample output:
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 4 of 10
ANOVA in Excel
To run an ANOVA in Excel, select Data > Data Analysis, and then choose the appropriate ANOVA test:
Choose the variables you want to include (input range), select the Labels box if needed, and set your
alpha (usually at .05).
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 5 of 10
Sample output:
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 6 of 10
Regression in Excel
Select Data > Data Analysis > Regression:
Next, select your Y range (dependent variable) and your X range (independent variables). Select the
Labels box if appropriate.
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 7 of 10
Sample output:
Chi-Square in Excel
Running Chi-Square in Excel is a little trickier. First, create a frequency table of the variables.
Highlight the table, select Copy, then go directly below and select Paste Special or Paste Values.
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 8 of 10
Relabel this table Expected and delete data inside the table (keep the total columns).
Next, we need to calculate the expected values.
Using this same method for all, multiply the column total by the row total, then divide by the grand total.
Column total * row total / Grand total (in bottom right corner of table) = Expected frequency
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 9 of 10
Continue until all expected frequencies are filled.
Next, we need to calculate the p. Type p to label your output, and then fill in the =chitest function by
highlighting the actual values in the top table and the expected values in the bottom table.
Copyright 2021 by University of Phoenix. All rights reserved.
Measures of Association and Tests of Significance in Excel
RES/710 v5
Page 10 of 10
This results in a probability value that you will compare to your alpha (.05, or .01) to decide whether you
have a statistically significant difference. (Here we fail to reject our null hypothesis since .687>.05,
meaning that there is no statistically significant difference between our observed and expected
frequencies.)
Copyright 2021 by University of Phoenix. All rights reserved.
