dataFrame.columns.stat — zTest()
Description
The zTest() method of the stat object performs a two-proportion z-test between selected binary columns.
Signature
dataFrame.columns(...columnNames).stat.zTest({ mode: 'twoSided', pooled: false })Arguments
...columnNames( string[] )- The name of the columns from which to compute the Z-test.
- Column 1: A binary grouping variable identifying the reference and test groups.
- Column 2: A binary outcome variable containing success and failure values.
options(object)- Z-test options.
Options
mode(string)- The alternative hypothesis used for the test.
oneSidedtwoSided(default)
pooled(boolean)- Indicates whether a pooled proportion should be used when estimating the standard error.
true: pooled version usedfalse: pooled version not used (default)
Returns
stat(object)- A statistic object containing the z-test results.
Z_score(number)- The computed z-test statistic.
p_value(number)- The p-value associated with the test statistic.
confidence(number)- A confidence score computed as
1 - p_value.
Notes
- The method compares the proportions observed in two groups.
- The null hypothesis states that both proportions are equal.
- The
pooledoption controls whether a pooled estimate of the proportion is used when computing the standard error. - Small p-values indicate evidence against the null hypothesis.
- Positive z-scores indicate that the first proportion is greater than the reference proportion.
- Negative z-scores indicate that the first proportion is lower than the reference proportion.
- The confidence score is calculated as
1 - p_value. - The test assumes sufficiently large sample sizes for the normal approximation to be valid.
Example
// computes a z-test for 2 binary columns from a dataFrame
var stat = dataFrame.columns('total', 'success').stat.zTest();
// log the stat details
notebook.log(stat);