mmbcv 1.0.0
- Added
heterogeneity_test() for testing equality of
transition-specific effects.
- Added
linear_trend_test() for testing GLS linear
trends.
- Added
order_restricted_test() with chi-bar-square
inference.
- Added support for coefficient subsets and user-specified transition
scores.
- Created separate help pages for the three tests.
- Reorganized internal testing functions into separate source
files.
mmbcv 0.4.0
- Added
heterogeneity_test() for an omnibus Wald test of
equality among selected transition-specific coefficients.
- Added
linear_trend_test() for one- or two-sided
generalized least-squares trend tests with equally or unequally spaced
transition scores.
- Added
order_restricted_test() for monotone increasing
or decreasing alternatives using a chi-bar-square reference
distribution.
- All three tests accept integer, logical, or character coefficient
selection through
index and work with any conformable
covariance estimator returned by MMBCV().
- The order-restricted test calculates analytic chi-bar-square weights
for one or two inequalities and simulation-based weights for higher
dimensions. It returns the weights, weight Monte Carlo standard errors,
and p-value Monte Carlo standard error, and accepts precomputed weights
when desired.