choice-metrics.com

This is an odd question.

Is there away to specify a fixed subset of sets within a design, and then use Ngene to find optimal additional sets to add.

For example, assume 8 sets are pre-specified (from an efficient design of 8 sets).
I want an efficient 12 set design, where 8 of the sets must come from the initial design, that is, I want to find the best 4 additional sets to add.

Is there a way to do this - I've not spotted one in the manual.
(I appreciate that ideally one would just generate a 12 set design!)

thanks
Dan

I do not see how Ngene could do that right now.

Ngene can read in designs and evaluate them, and also use them as a starting point in some algorithms, but at the moment there is no way of telling Ngene that it should only change 4 out of 12 choice sets.
But I like the idea of being able to do this. Perhaps this could be a possible future feature.

Thanks Michiel

We'll look at what we can do in Excel for now and see what is possible in the next version of Ngene
(although this request is far less important to me than the wish to have interactions on dummy or effects coded attributes)

best wishes
Dan

In version 1.1 we introduced interactions between dummy or effects coded variables, so that feature is already in You may want to update to this new version (is free) if you did not already do so.