One Sample Tests - Means

From Q
Jump to: navigation, search

The significance tests used for means is determined by whether or not the data is weighted and by the Means setting and the Design effect for weight setting in Statistical Assumptions.

If Means = Non-parametric:

Data is not weighted: t-Test of a Mean
Data is weighted:Complex Samples t-Test of a Mean

If Means = z-Test:

Data is not weighted: Z-Test of a Mean
Data is weighted: Complex Samples Z-Test of a Mean

If Means = t-Test:

Data is not weighted: t-Test of a Mean
Data is weighted: Complex Samples t-Test of a Mean

If Means = Quantum Means or Survey Reporter Means: Quantum and Survey Reporter t-Test of a Mean

Each of these tests has an expected value,e, as an input. For Number - Grid questions: e is computed multiplying the average for the row with the average for the column and dividing by the overall table average (excluding cells for which Not Duplicate is 0). In most other instances it is the average of all the cells or 0 (the expected value is shown in the results of a Planned Tests Of Statistical Significance).

See also