Experimental design

[mimii.zai]

Hello all!
I am new to choice experiments and was wondering if someone could explain what each of the following mean:
D error 0.0922
A error 8.208901
B estimate 0
S estimate 15362.43345

Fixed prior value
Sp estimates
Sp t-ratios

[Michiel Bliemer]

The Ngene manual explains so I recommend reading the manual, but in short:

D-error is the normalised determinant of the variance-covariance matrix; the lower this value, the more precise the parameter estimates will be.
A-error is the trace of the variance-covariance matrix, the lower this value, the more precise the parameter estimates will be.
B-estimate indicates average probability balance across choice tasks, a value of 1 means perfect balance (i.e., each alternative is chosen with equal probability) while a value of 0 means complete imbalance whereby one alternative is always preferred over other alternatives.
S-estimate indicates the sample size needed to estimate all parameters statistically significant with 95% confidence; note that this value highly depends on the priors used, one should ignore this value if priors are not reasonably reliable, they typically should come from estimations in a pilot study.
Sp-estimates are sample size estimated for each parameter separately so you can see which parameters will be the most difficult to estimate; again, these values are only meaningful when reliable priors are used.

Michiel