Testing Equivalence with Repeated Measures: Tests of the Difference Model of Two-Alternative Forced-Choice Performance