choice-metrics.com

Dear Michiel,

so far I´ve used dummy coding for my estimations of the above mentioned model.
Is there any indication to use effects coding instead of dummy coding?
I´ve seen that some papers are using effects coding for categorical variables but am not sure if that might also make sense for my model.
I suppose no, because effects coding prevents from choosing the reference level arbitrarily, but as I have a fixed status quo dummy coding with SQ levels as reference levels seems to make most sense?!

Thankful for any information from your point of view!
Thanks and best,
J.

Sorry, this is the respective code:

Design
;alts = alt1*, alt2*, SQ*
;rows = 18
;block = 2
;eff = (mnl,d)
;alg = mfederov
;require:
SQ.amount = 3, SQ.period = 3
;reject:
alt1.amount = 3 AND alt1.period = 3,
alt2.amount = 3 AND alt2.period = 3
;model:
U(alt1) = b1[-0.0001] * contrib[1.8,2.5,3.2](6,6,6)
+ b2.dummy[0.0003|0.0002|0.0001] * amount[0,1,2,3] ? 0 = 300, 1 = 600, 2 = 900, 3 = unlimited amount
+ b3.dummy[0.0003|0.0002|0.0001] * period[0,1,2,3] ? 0 = 12, 1 = 42, 2 = 72, 3 = unlimited period
/
U(alt2) = b1 * contrib
+ b2 * amount
+ b3 * period
/
U(SQ) = b0[0]
+ b1 * contrib_sq[1.6]
+ b2 * amount
+ b3 * period
$

There is no real difference between dummy and effects coding as shown by Daly & Dekker & Hess (2016), so using dummy coding is perfectly fine and will make it easier to interpret parameters.

Daly, Dekker and Hess (2016) Dummy coding vs effects coding for categorical variables: Clarifications and extensions. Journal of Choice Modelling. https://www.sciencedirect.com/science/article/abs/pii/S1755534516300781

Michiel

Thank you very much for re-confirming, Michiel!