Protocol Implementation
Morning and afternoon sets of n=32 mice (16 cages) were run to complete the n=64/replication.
Semi-blind testing was employed by using cage numbers rather than strain identity as identifiers (caveat regarding coat color).
Testing order (which mouse genotype,sex, etc) was based on a constrained randomized sequence (although cage-pair mates were always tested in adjacent positions). This sequence varied across site.
Same daily test routine for each squad of 2 or 4 mice was used on the first 5 tests.
Documents describing protocols are available on the web site.