[R] Fractional Factorial Design on 4-level factor

[Michael Haenlein]

Dear all,

I am running a simulation experiment with 8 factors that each have 4
levels. Each combination is repeated 100 times. If I run a full factorial
this would mean 100*8^4 = 409,600 runs.

I am trying to reduce the number of scenarios to run using a fractional
factorial design. I'm interested in estimating the main effects of the 8
factors plus their 2-way interactions. Any higher level interactions are
not of interest to me. My plan is to use a standard OLS regression for
that, once the simulations are over.

I tried to use the FrF2 package to derive a fractional factorial design but
it seems that this is only working for factors on two levels. Any idea how
I could derive a fractional factorial design on factors with four levels?

Thanks for your help,

Michael



Michael Haenlein
Professor of Marketing
ESCP Europe

	[[alternative HTML version deleted]]