How Do I Perform a One Sample t-Test in Excel Using SigmaXL?

Hypothesis Testing – One Sample t-Test

1. Open Customer Data.xls, select Sheet 1 tab (or press F4 to activate last worksheet). Click SigmaXL > Statistical Tools > 1 Sample t-Test & Confidence Intervals. If necessary, check Use Entire Data Table, click Next.
2. Ensure that Stacked Column Format is selected. Select Overall Satisfaction, click Numeric Data Variable (Y) >>, select Customer Type, click Optional Group Category (X) >>.
3.  Historically, our average customer satisfaction score has been 3.5. We would like to see if this has changed, with the results grouped by customer type. Null Hypothesis H0: μ=3.5; Alternative Hypothesis Ha: μ≠3.5
4. Enter 3.5 for the Null Hypothesis H0: Mean value. Keep Ha as Not Equal To.




5. Click OK. Results:




6. Note the p-values. Customer Type 2 shows a significant change (increase) in Satisfaction Mean (p-value < .05), whereas Customer Types 1 and 3 show no change (p-value ≥ .05). Also note the confidence intervals around each mean match the results from Descriptive Statistics.
7. In the Measure Phase we determined that Overall Satisfaction for Customer Type 2 has non-normal data but this does not imply that the p-value for the 1 Sample t-test is wrong. The Central Limit Theorem applies here: the distribution of averages tends to be normal, even if the individual observations are not-normal. With a sample size of 42, the t-test is fairly robust against skewed data.

Click here to download a Free 30-Day Trial.