Include Top

Home | Login | Live Help is Live Help

Included Menu

How Do I Run a Two Sample Comparison Test in Excel Using SigmaXL?

We will now do a full comparison test of Customer Satisfaction for Customer Type 1 and 2. This test checks each sample for normality, equal variance (F-test and Levene’s), 2 sample t-test (assuming equal and unequal variance), and Mann Whitney test for equal Medians. Depending on the normality, variance, and sample size results, the appropriate p-values are highlighted in yellow.

Two Sample Comparison Test

  1. Open Customer Data.xls, click on Sheet 1 Tab.
  2. Click SigmaXL > Statistical Tools > 2 Sample Comparison Tests. Click Next. Check Stacked Column Format. Select Overall Satisfaction, click Numeric Data Variable (Y) >>, select Customer Type, click Group Category (X) >>.

  3. 2 Sample Comparison Test

  4. Click OK. Select Customer Type 1 and 2.

  5. Two Sample Comparison
  6. Click OK.

  7. Two Sample Comparison Results

    Customer Type 2 has non-normal data. This makes Levene’s test the appropriate test for unequal variance. Levene’s test indicates that Customer type 2 has a significantly lower variance, or standard deviation. The lower standard deviation translates to a consistent level of satisfaction.

    Since Levene’s test indicates unequal variance, the appropriate t-test assumes unequal variance. The t-test indicates that Customer Type 2 has a significantly higher mean satisfaction.

    Clearly the next step would be to determine a root cause or best practices to reduce the variability in overall satisfaction and increase the mean for all customer types.

Click here to download a Free 30-Day Trial.