TTest_upper() returns the aggregated value for the upper end of the confidence interval for two independent series of values.
This function applies to independent samples student's t-tests.
If the function is used in the load script, the values are iterated over a number of records as defined by a group by clause.
If the function is used in a chart expression, the values are iterated over the chart dimensions.
Syntax:
TTest_upper (grp, value [, sig [, eq_var]])
Return data type: numeric
Arguments:
- value: The sample values to be evaluated. The sample values must be logically grouped as specified by exactly two values in group. If a field name for the sample values is not provided in the load script, the field will automatically be named Value.
- grp: The field containing the names of each of the two sample groups. If a field name for the group is not provided in the load script, the field will automatically be given the name Type.
- sig: The two-tailed level of significance can be specified in sig. If omitted, sig is set to 0.025, resulting in a 95% confidence interval.
- eq_var: If eq_var is specified as False (0), separate variances of the two samples will be assumed. If eq_var is specified as True (1), equal variances between the samples will be assumed.
Limitations:
Text values, NULL values and missing values in the expression value will result in the function returning NULL.
Examples:
TTest_upper( Group, Value )
TTest_upper( Group, Value, sig, false )
