Data Algorithms

Chapter 20. Cochran-Armitage Test for Trend

The Cochran-Armitage test for trend (CATT) is used in analyzing germline data. For example, variants in a VCF (variant call format) file generated by DNA sequencing can be labeled as germline data. The CATT is a statistical method of directing chi-squared tests toward narrow alternatives. If R is a set of response variables and E is a set of experimental variables, then the CATT is sensitive to the linearity between R(s) and E(s) and detects trends. The CATT can be expressed another way: if B is a binary outcome of some events {PASSED, FAILED} and C is a set of ordered categories {C₁, ..., C_n}, then the CATT can be used as a linear trend in proportions on B across levels of C. To apply the CATT, we build a contingency table: two rows with outcome values {PASSED, FAILED} and n columns as {C₁, ..., C_n}. The contingency table for the CATT is explained in the next sections.

According to Wikipedia:

The Cochran-Armitage test for trend, named for William Cochran and Peter Armitage, is used in categorical data analysis when the aim is to assess for the presence of an association between a variable with two categories and a variable with k categories. It modifies the Pearson chi-squared test to incorporate a suspected ordering in the effects of the k categories of the second variable. For example, doses of a treatment can be ordered as “low,” “medium,” and “high,” and we may suspect that the treatment benefit cannot become smaller as the dose ...