Summary statistics can be used to summarize data sets and relationships between data in a simple way. Within Sift there are curretnly three differeten statistical tests:

1. One Sample T Test
2. Two Sample T Test
3. Paired T Test

The Summary Statistics dialog is found on the toolbar and under the Analysis menu.

The summary statistics dialog is split into two sections: the test information, and the results.

Depending on the test you select the test information and results displayed might be different.

To run a one sample t test you will need to input the following:

• Test Name
• Statistical Test: One Sample T Test
• Significance Level: Options of 0.1, 0.05, 0.025,0.01,0.005,0.001
• Data Selection
  • Group: One Metric must be selected
  • Workspaces: Select all workspaces you want included in your test
  • Use Excluded Data: If checked, any excluded data will be included in your test
  • Use All Workspaces: If checked, all workspaces will be used in your test
  • Population Mean: Mean of the population you are comparing against

Once the test has run the results information will be displayed:

• Result Name: Drop down to ensure you are looking at the right test
• Test Information
  • Statistical Test: What test was performed
  • Significance Level: What significance level was used to run the test
  • Population Mean: What Population Mean was used
  • Group1: Name of group used in test. If expanded the selected workspaces will be shown
• Group Results
  • For the group the following results will be shown
    • Mean
    • Standard Deviation
    • Standard Error
    • N
• Test Results
  • Degrees of Freedom
  • T Stat
  • Null Hypothesis : Either will say Do Not Reject Null Hypothesis, or Reject Null Hypothesis

To run a two sample t test you will need to input the following:

• Test Name
• Statistical Test: Two Sample T Test
• Significance Level: Options of 0.2,0.1, 0.05,0.01,0.002
• Data Selection
  • Group: Two Metrics must be selected
  • Workspaces: Select all workspaces you want included in your test
  • Use Excluded Data: If checked, any excluded data will be included in your test
  • Use All Workspaces: If checked, all workspaces will be used in your test

Once the test has run the results information will be displayed:

• Result Name: Drop down to ensure you are looking at the right test
• Test Information
  • Statistical Test: What test was performed
  • Significance Level: What significance level was used to run the test
  • Group1: Name of group one used in test. If expanded the selected workspaces will be shown
  • Group2: Name of group two used in test. If expanded the selected workspaces will be shown
• Group Results
  • For each group and overall the following results will be shown
    • Mean
    • Standard Deviation
    • Standard Error
    • N
• Test Results
  • Degrees of Freedom
  • Pooled Standard Deviation
  • T Stat
  • Null Hypothesis : Either will say Do Not Reject Null Hypothesis, or Reject Null Hypothesis

To run a Paired t test you will need to input the following:

• Test Name
• Statistical Test: Paired T Test
• Significance Level: Options of 0.2,0.1, 0.05,0.01,0.002
• Data Selection
  • Group: Two Metrics must be selected
  • Workspaces: Select all workspaces you want included in your test
  • Use Excluded Data: If checked, any excluded data will be included in your test
  • Use All Workspaces: If checked, all workspaces will be used in your test

Once the test has run the results information will be displayed:

• Result Name: Drop down to ensure you are looking at the right test
• Test Information
  • Statistical Test: What test was performed
  • Significance Level: What significance level was used to run the test
  • Group1: Name of group one used in test. If expanded the selected workspaces will be shown
  • Group2: Name of group two used in test. If expanded the selected workspaces will be shown
• Group Results
  • For each group and overall the following results will be shown
    • Mean
    • Standard Deviation
    • Standard Error
    • N
• Test Results
  • Degrees of Freedom
  • Mean of Differences
  • Standard Deviation of Differences
  • Standard Error of Differences
  • T Stat
  • Null Hypothesis : Either will say Do Not Reject Null Hypothesis, or Reject Null Hypothesis