Premium
Statistical Tests for Signals in Categorial Temporal Data
Author(s) -
Can Ann R.,
Meeker William Q.
Publication year - 1996
Publication title -
biometrical journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.108
H-Index - 63
eISSN - 1521-4036
pISSN - 0323-3847
DOI - 10.1002/bimj.4710380105
Subject(s) - statistics , categorical variable , wald test , statistic , test statistic , mathematics , statistical hypothesis testing , data set , set (abstract data type) , computer science , programming language
This paper describes methods for using categorical temporal data to detect differences in behavior between a treated group and a control group. The first‐level output from the data is typically a set of many different correlated test statistics comparing the two groups. In previous work, a decision was made by counting the number of significant individual tests and calibrating with bootstrap simulation. This article goes further, suggesting two possible alternative statistics: the sum of the squared individual test statistics and a Wald‐like combination of the individual test statistics. All three overall comparison statistics are defined and a method for computing critical values from simulated distributions using a bootstrap method is given. The use of all three methods is then demonstrated on each of three data sets. Finally, a simulated power study reveals that the Wald‐like statistic is much better than the other two, leading to the suggestion of its use in place of the other two statistics.