Sunday 1 December 2013

Cochran's Q Test Calculator with Post-hoc Analysis

Please click to add a treatment.

Cochran's Q test is a non-parametric statistical test used to test whether a number of dependent treatments have identical effects, and is used for dichotomous data (expressible as numbers with a value of either 0 or 1). This test can be regarded as an extension of the McNemar test.

Should the calculated p-value be less that 0.05, a cursory post-hoc analysis is carried out by performing pairwise McNemar tests on the treatments. Any resulting p-value less than 0.05 is highlighted by two asterisks.

Simply click on the link near the top to add text boxes. Each text box stores a single treatment and needs to be filled in with comma separated numbers, which take value 0 or 1. All treatments need to have the same number of samples, which is equal to the number of blocks.

Alternatively, you can choose two file entry methods:-

  1. Select multiple single column CSV files to populate the text boxes by repeatedly pressing the Choose File button - there must be one distinct (and differently named) file for each text box i.e. one file per group. Each file can have a different number of samples.
  2. Select a single multi-column CSV file by pressing the Choose File button once, where the number of columns equals the number of groups - all groups need to have the same number of samples.

Results pending...


Logistic Regression Calculator and ROC Curve Plotter

This blog post implements a Logistic Regression calculator for a binary output. Consider a binary outcome response variable \(Y\...