Sample Size Calculator and Data Analysis Tool for Non-Inferiority and Equivalence tests in SMART

Input for Non-Inferiority Test:

Significance Level:

Desired Power:

Standardized Effect Size:

Input for Equivalence Test:

Significance Level:

The Standardized Equivalence Margin:

The Standardized Effect Size:

Sample Size (includes all AIs) to Start:

Input for Data Analysis:

Data Description:

The input data must be xls or xlsx format. It should contain precisely five variables (columns) with the specific variable names: sl, Trt.1, R, Trt.2, Y; where sl: denotes serial number (ex: 1,2,3,...), Trt.1: denotes first stage treatments (must be A or B), R: denotes responder status (must be 1 or 0, 1: responder, 0: non-responder), Trt.2: denotes second stage treatments (must be C or D if first stage treatment was A; or E or F if first stage treatment was B), Y: denotes the continuous primary outcome.

Reading user's Data:

Significance Level:

The Non-Inferiority/Equivalence Margin ( θ ):

The Test for Data Analysis:

Non Inferiority

Equivalence

Comparison Type:

Distinct-Path

Shared-Path

Choosen Input Parameter Values:

Note: The required sample size is same for a comparison of distinct-path or shared-path AIs with respect to a fixed standardized effect size. However, a different combination of the non-inferiority margin, mean difference of two AIs and the variance(mean difference of two AIs) can give same standardized effect size.

Note: The required sample size is same for a comparison of distinct-path or shared-path AIs with respect to the fixed standardized equivalence margin and standardized effect size. However, a different combination of the equivalence margin, mean difference of two AIs and the variance(mean difference of two AIs) can give same pair of standardized equivalence margin and standardized effect size, respectively.

Data Analysis Results: