choice-metrics.com

Hello all,

I am trying to generate an efficient design to estimate all possible interactions using the normally distributed priors for the main effects; overall there are 100 parameters for the model with all 2-way interactions. I have 5 attributes with 3 levels each, dummy coded. When I input my script(see below), Ngene shows an error: Two identically named attributes do not have consistently specified levels: 'alt1.mobility', 'alt1.mobility'.
Could you please advise me upon this issue?

Thanks in advance!

Design
;alts=alt1, alt2
;rows=200
;block=10
;eff=(mnl,d, mean)
;bdraws=halton(500)
;model:
U(alt1)=b1+t1.dummy[(n,-0.16,0.08)|(n,-1.31,0.67)]*mobility[1,2,3]+
t2.dummy[(n,-1.3,0.66)|(n,-0.96,0.49)]*selfcare[1,2,3]+
t3.dummy[(n,0.51,0.26)|(n,-1.55,0.79)]*usualactivity[1,2,3]+
t4.dummy[(n,-0.37,0.19)|(n,-0.01,0.00)]*pain[1,2,3]+
t5.dummy[(n,-1.31,0.67)|(n,-0.81,0.41)]*depression[1,2,3]+
i1*mobility[1]*selfcare[1]+
i2*mobility[1]*selfcare[2]+
i3*mobility[1]*selfcare[3]+
i4*mobility[2]*selfcare[1]+
i5*mobility[2]*selfcare[2]+
i6*mobility[2]*selfcare[3]+
i7*mobility[3]*selfcare[1]+
i8*mobility[3]*selfcare[2]+
i9*mobility[3]*selfcare[3]+
i10*mobility[1]*usualactivity[1]+
i11*mobility[1]*usualactivity[2]+
i12*mobility[1]*usualactivity[3]+
i13*mobility[2]*usualactivity[1]+
i14*mobility[2]*usualactivity[2]+
i15*mobility[2]*usualactivity[3]+
i16*mobility[3]*usualactivity[1]+
i17*mobility[3]*usualactivity[2]+
i18*mobility[3]*usualactivity[3]+
i19*mobility[1]*pain[1]+
i20*mobility[1]*pain[2]+
i21*mobility[1]*pain[3]+
i22*mobility[2]*pain[1]+
i23*mobility[2]*pain[2]+
i24*mobility[2]*pain[3]+
i25*mobility[3]*pain[1]+
i26*mobility[3]*pain[2]+
i27*mobility[3]*pain[3]+
i28*mobility[1]*depression[1]+
i29*mobility[1]*depression[2]+
i30*mobility[1]*depression[3]+
i31*mobility[2]*depression[1]+
i32*mobility[2]*depression[2]+
i33*mobility[2]*depression[3]+
i34*mobility[3]*depression[1]+
i35*mobility[3]*depression[2]+
i36*mobility[3]*depression[3]+
i37*selfcare[1]*usualactivity[1]+
i38*selfcare[1]*usualactivity[2]+
i39*selfcare[1]*usualactivity[3]+
i40*selfcare[2]*usualactivity[1]+
i41*selfcare[2]*usualactivity[2]+
i42*selfcare[2]*usualactivity[3]+
i43*selfcare[3]*usualactivity[1]+
i44*selfcare[3]*usualactivity[2]+
i45*selfcare[3]*usualactivity[3]+
i46*selfcare[1]*pain[1]+
i47*selfcare[1]*pain[2]+
i48*selfcare[1]*pain[3]+
i49*selfcare[2]*pain[1]+
i50*selfcare[2]*pain[2]+
i51*selfcare[2]*pain[3]+
i52*selfcare[3]*pain[1]+
i53*selfcare[3]*pain[2]+
i54*selfcare[3]*pain[3]+
i55*selfcare[1]*depression[1]+
i56*selfcare[1]*depression[2]+
i57*selfcare[1]*depression[3]+
i58*selfcare[2]*depression[1]+
i59*selfcare[2]*depression[2]+
i60*selfcare[2]*depression[3]+
i61*selfcare[3]*depression[1]+
i62*selfcare[3]*depression[2]+
i63*selfcare[3]*depression[3]+
i64*usualactivity[1]*pain[1]+
i65*usualactivity[1]*pain[2]+
i66*usualactivity[1]*pain[3]+
i67*usualactivity[2]*pain[1]+
i68*usualactivity[2]*pain[2]+
i69*usualactivity[2]*pain[3]+
i70*usualactivity[3]*pain[1]+
i71*usualactivity[3]*pain[2]+
i72*usualactivity[3]*pain[3]+
i73*usualactivity[1]*depression[1]+
i74*usualactivity[1]*depression[2]+
i75*usualactivity[1]*depression[3]+
i76*usualactivity[2]*depression[1]+
i77*usualactivity[2]*depression[2]+
i78*usualactivity[2]*depression[3]+
i79*usualactivity[3]*depression[1]+
i80*usualactivity[3]*depression[2]+
i81*usualactivity[3]*depression[3]+
i82*pain[1]*depression[1]+
i83*pain[1]*depression[2]+
i84*pain[1]*depression[3]+
i85*pain[2]*depression[1]+
i86*pain[2]*depression[2]+
i87*pain[2]*depression[3]+
i88*pain[3]*depression[1]+
i89*pain[3]*depression[2]+
i90*pain[3]*depression[3]
/
U(alt2)=t1.dummy*mobility[1,2,3]+
t2.dummy*selfcare[1,2,3]+
t3.dummy*usualactivity[1,2,3]+
t4.dummy*pain[1,2,3]+
t5.dummy*depression[1,2,3]+
i1*mobility[1]*selfcare[1]+
i2*mobility[1]*selfcare[2]+
i3*mobility[1]*selfcare[3]+
i4*mobility[2]*selfcare[1]+
i5*mobility[2]*selfcare[2]+
i6*mobility[2]*selfcare[3]+
i7*mobility[3]*selfcare[1]+
i8*mobility[3]*selfcare[2]+
i9*mobility[3]*selfcare[3]+
i10*mobility[1]*usualactivity[1]+
i11*mobility[1]*usualactivity[2]+
i12*mobility[1]*usualactivity[3]+
i13*mobility[2]*usualactivity[1]+
i14*mobility[2]*usualactivity[2]+
i15*mobility[2]*usualactivity[3]+
i16*mobility[3]*usualactivity[1]+
i17*mobility[3]*usualactivity[2]+
i18*mobility[3]*usualactivity[3]+
i19*mobility[1]*pain[1]+
i20*mobility[1]*pain[2]+
i21*mobility[1]*pain[3]+
i22*mobility[2]*pain[1]+
i23*mobility[2]*pain[2]+
i24*mobility[2]*pain[3]+
i25*mobility[3]*pain[1]+
i26*mobility[3]*pain[2]+
i27*mobility[3]*pain[3]+
i28*mobility[1]*depression[1]+
i29*mobility[1]*depression[2]+
i30*mobility[1]*depression[3]+
i31*mobility[2]*depression[1]+
i32*mobility[2]*depression[2]+
i33*mobility[2]*depression[3]+
i34*mobility[3]*depression[1]+
i35*mobility[3]*depression[2]+
i36*mobility[3]*depression[3]+
i37*selfcare[1]*usualactivity[1]+
i38*selfcare[1]*usualactivity[2]+
i39*selfcare[1]*usualactivity[3]+
i40*selfcare[2]*usualactivity[1]+
i41*selfcare[2]*usualactivity[2]+
i42*selfcare[2]*usualactivity[3]+
i43*selfcare[3]*usualactivity[1]+
i44*selfcare[3]*usualactivity[2]+
i45*selfcare[3]*usualactivity[3]+
i46*selfcare[1]*pain[1]+
i47*selfcare[1]*pain[2]+
i48*selfcare[1]*pain[3]+
i49*selfcare[2]*pain[1]+
i50*selfcare[2]*pain[2]+
i51*selfcare[2]*pain[3]+
i52*selfcare[3]*pain[1]+
i53*selfcare[3]*pain[2]+
i54*selfcare[3]*pain[3]+
i55*selfcare[1]*depression[1]+
i56*selfcare[1]*depression[2]+
i57*selfcare[1]*depression[3]+
i58*selfcare[2]*depression[1]+
i59*selfcare[2]*depression[2]+
i60*selfcare[2]*depression[3]+
i61*selfcare[3]*depression[1]+
i62*selfcare[3]*depression[2]+
i63*selfcare[3]*depression[3]+
i64*usualactivity[1]*pain[1]+
i65*usualactivity[1]*pain[2]+
i66*usualactivity[1]*pain[3]+
i67*usualactivity[2]*pain[1]+
i68*usualactivity[2]*pain[2]+
i69*usualactivity[2]*pain[3]+
i70*usualactivity[3]*pain[1]+
i71*usualactivity[3]*pain[2]+
i72*usualactivity[3]*pain[3]+
i73*usualactivity[1]*depression[1]+
i74*usualactivity[1]*depression[2]+
i75*usualactivity[1]*depression[3]+
i76*usualactivity[2]*depression[1]+
i77*usualactivity[2]*depression[2]+
i78*usualactivity[2]*depression[3]+
i79*usualactivity[3]*depression[1]+
i80*usualactivity[3]*depression[2]+
i81*usualactivity[3]*depression[3]+
i82*pain[1]*depression[1]+
i83*pain[1]*depression[2]+
i84*pain[1]*depression[3]+
i85*pain[2]*depression[1]+
i86*pain[2]*depression[2]+
i87*pain[2]*depression[3]+
i88*pain[3]*depression[1]+
i89*pain[3]*depression[2]+
i90*pain[3]*depression[3]

$

See page 126 of the Ngene manual, you need to specify references to dummy attributes for interactions as i1*mobility.dummy[1] * selfcare.dummy[1], etc.

Thanks a lot! It helps.

Best wishes,
Anna

Hello all,

I would like to ask additional question: after I ran the syntax specified earlier, I receive the following error: No valid design has been found after 1000 evaluations. There may be a problem with the specification of the design. A common problem is that the choice probabilities are too extreme (close to 1 and 0), perhaps because some or all of the prior values are too large. Also, it is generally a good idea to start with a simple design (MNL, non-Bayesian), then add complexity. If you press stop, a design will be reported, which may assist in diagnosing the problem.
I have already run just the MNL, and I can't change the specification since my aim is to estimate all 2-way interactions, I also tried to eliminate priors to reduce complexity, but this didnn't help either. Could you suggest please what can be done to obtain a design for the interactions?
The syntax below is without priors:

Design
;alts=alt1, alt2
;rows=200
;eff=(mnl,d)
;model:
U(alt1)=b1+t1*mobility[1,2,3]+
t2*selfcare[1,2,3]+
t3*usualactivity[1,2,3]+
t4*pain[1,2,3]+
t5*depression[1,2,3]+
i1*mobility.dummy[1]*selfcare.dummy[1]+
i2*mobility.dummy[1]*selfcare.dummy[2]+
i3*mobility.dummy[1]*selfcare.dummy[3]+
i4*mobility.dummy[2]*selfcare.dummy[1]+
i5*mobility.dummy[2]*selfcare.dummy[2]+
i6*mobility.dummy[2]*selfcare.dummy[3]+
i7*mobility.dummy[3]*selfcare.dummy[1]+
i8*mobility.dummy[3]*selfcare.dummy[2]+
i9*mobility.dummy[3]*selfcare.dummy[3]+
i10*mobility.dummy[1]*usualactivity.dummy[1]+
i11*mobility.dummy[1]*usualactivity.dummy[2]+
i12*mobility.dummy[1]*usualactivity.dummy[3]+
i13*mobility.dummy[2]*usualactivity.dummy[1]+
i14*mobility.dummy[2]*usualactivity.dummy[2]+
i15*mobility.dummy[2]*usualactivity.dummy[3]+
i16*mobility.dummy[3]*usualactivity.dummy[1]+
i17*mobility.dummy[3]*usualactivity.dummy[2]+
i18*mobility.dummy[3]*usualactivity.dummy[3]+
i19*mobility.dummy[1]*pain.dummy[1]+
i20*mobility.dummy[1]*pain.dummy[2]+
i21*mobility.dummy[1]*pain.dummy[3]+
i22*mobility.dummy[2]*pain.dummy[1]+
i23*mobility.dummy[2]*pain.dummy[2]+
i24*mobility.dummy[2]*pain.dummy[3]+
i25*mobility.dummy[3]*pain.dummy[1]+
i26*mobility.dummy[3]*pain.dummy[2]+
i27*mobility.dummy[3]*pain.dummy[3]+
i28*mobility.dummy[1]*depression.dummy[1]+
i29*mobility.dummy[1]*depression.dummy[2]+
i30*mobility.dummy[1]*depression.dummy[3]+
i31*mobility.dummy[2]*depression.dummy[1]+
i32*mobility.dummy[2]*depression.dummy[2]+
i33*mobility.dummy[2]*depression.dummy[3]+
i34*mobility.dummy[3]*depression.dummy[1]+
i35*mobility.dummy[3]*depression.dummy[2]+
i36*mobility.dummy[3]*depression.dummy[3]+
i37*selfcare.dummy[1]*usualactivity.dummy[1]+
i38*selfcare.dummy[1]*usualactivity.dummy[2]+
i39*selfcare.dummy[1]*usualactivity.dummy[3]+
i40*selfcare.dummy[2]*usualactivity.dummy[1]+
i41*selfcare.dummy[2]*usualactivity.dummy[2]+
i42*selfcare.dummy[2]*usualactivity.dummy[3]+
i43*selfcare.dummy[3]*usualactivity.dummy[1]+
i44*selfcare.dummy[3]*usualactivity.dummy[2]+
i45*selfcare.dummy[3]*usualactivity.dummy[3]+
i46*selfcare.dummy[1]*pain.dummy[1]+
i47*selfcare.dummy[1]*pain.dummy[2]+
i48*selfcare.dummy[1]*pain.dummy[3]+
i49*selfcare.dummy[2]*pain.dummy[1]+
i50*selfcare.dummy[2]*pain.dummy[2]+
i51*selfcare.dummy[2]*pain.dummy[3]+
i52*selfcare.dummy[3]*pain.dummy[1]+
i53*selfcare.dummy[3]*pain.dummy[2]+
i54*selfcare.dummy[3]*pain.dummy[3]+
i55*selfcare.dummy[1]*depression.dummy[1]+
i56*selfcare.dummy[1]*depression.dummy[2]+
i57*selfcare.dummy[1]*depression.dummy[3]+
i58*selfcare.dummy[2]*depression.dummy[1]+
i59*selfcare.dummy[2]*depression.dummy[2]+
i60*selfcare.dummy[2]*depression.dummy[3]+
i61*selfcare.dummy[3]*depression.dummy[1]+
i62*selfcare.dummy[3]*depression.dummy[2]+
i63*selfcare.dummy[3]*depression.dummy[3]+
i64*usualactivity.dummy[1]*pain.dummy[1]+
i65*usualactivity.dummy[1]*pain.dummy[2]+
i66*usualactivity.dummy[1]*pain.dummy[3]+
i67*usualactivity.dummy[2]*pain.dummy[1]+
i68*usualactivity.dummy[2]*pain.dummy[2]+
i69*usualactivity.dummy[2]*pain.dummy[3]+
i70*usualactivity.dummy[3]*pain.dummy[1]+
i71*usualactivity.dummy[3]*pain.dummy[2]+
i72*usualactivity.dummy[3]*pain.dummy[3]+
i73*usualactivity.dummy[1]*depression.dummy[1]+
i74*usualactivity.dummy[1]*depression.dummy[2]+
i75*usualactivity.dummy[1]*depression.dummy[3]+
i76*usualactivity.dummy[2]*depression.dummy[1]+
i77*usualactivity.dummy[2]*depression.dummy[2]+
i78*usualactivity.dummy[2]*depression.dummy[3]+
i79*usualactivity.dummy[3]*depression.dummy[1]+
i80*usualactivity.dummy[3]*depression.dummy[2]+
i81*usualactivity.dummy[3]*depression.dummy[3]+
i82*pain.dummy[1]*depression.dummy[1]+
i83*pain.dummy[1]*depression.dummy[2]+
i84*pain.dummy[1]*depression.dummy[3]+
i85*pain.dummy[2]*depression.dummy[1]+
i86*pain.dummy[2]*depression.dummy[2]+
i87*pain.dummy[2]*depression.dummy[3]+
i88*pain.dummy[3]*depression.dummy[1]+
i89*pain.dummy[3]*depression.dummy[2]+
i90*pain.dummy[3]*depression.dummy[3]
/
U(alt2)=t1*mobility[1,2,3]+
t2*selfcare[1,2,3]+
t3*usualactivity[1,2,3]+
t4*pain[1,2,3]+
t5*depression[1,2,3]+
i1*mobility.dummy[1]*selfcare.dummy[1]+
i2*mobility.dummy[1]*selfcare.dummy[2]+
i3*mobility.dummy[1]*selfcare.dummy[3]+
i4*mobility.dummy[2]*selfcare.dummy[1]+
i5*mobility.dummy[2]*selfcare.dummy[2]+
i6*mobility.dummy[2]*selfcare.dummy[3]+
i7*mobility.dummy[3]*selfcare.dummy[1]+
i8*mobility.dummy[3]*selfcare.dummy[2]+
i9*mobility.dummy[3]*selfcare.dummy[3]+
i10*mobility.dummy[1]*usualactivity.dummy[1]+
i11*mobility.dummy[1]*usualactivity.dummy[2]+
i12*mobility.dummy[1]*usualactivity.dummy[3]+
i13*mobility.dummy[2]*usualactivity.dummy[1]+
i14*mobility.dummy[2]*usualactivity.dummy[2]+
i15*mobility.dummy[2]*usualactivity.dummy[3]+
i16*mobility.dummy[3]*usualactivity.dummy[1]+
i17*mobility.dummy[3]*usualactivity.dummy[2]+
i18*mobility.dummy[3]*usualactivity.dummy[3]+
i19*mobility.dummy[1]*pain.dummy[1]+
i20*mobility.dummy[1]*pain.dummy[2]+
i21*mobility.dummy[1]*pain.dummy[3]+
i22*mobility.dummy[2]*pain.dummy[1]+
i23*mobility.dummy[2]*pain.dummy[2]+
i24*mobility.dummy[2]*pain.dummy[3]+
i25*mobility.dummy[3]*pain.dummy[1]+
i26*mobility.dummy[3]*pain.dummy[2]+
i27*mobility.dummy[3]*pain.dummy[3]+
i28*mobility.dummy[1]*depression.dummy[1]+
i29*mobility.dummy[1]*depression.dummy[2]+
i30*mobility.dummy[1]*depression.dummy[3]+
i31*mobility.dummy[2]*depression.dummy[1]+
i32*mobility.dummy[2]*depression.dummy[2]+
i33*mobility.dummy[2]*depression.dummy[3]+
i34*mobility.dummy[3]*depression.dummy[1]+
i35*mobility.dummy[3]*depression.dummy[2]+
i36*mobility.dummy[3]*depression.dummy[3]+
i37*selfcare.dummy[1]*usualactivity.dummy[1]+
i38*selfcare.dummy[1]*usualactivity.dummy[2]+
i39*selfcare.dummy[1]*usualactivity.dummy[3]+
i40*selfcare.dummy[2]*usualactivity.dummy[1]+
i41*selfcare.dummy[2]*usualactivity.dummy[2]+
i42*selfcare.dummy[2]*usualactivity.dummy[3]+
i43*selfcare.dummy[3]*usualactivity.dummy[1]+
i44*selfcare.dummy[3]*usualactivity.dummy[2]+
i45*selfcare.dummy[3]*usualactivity.dummy[3]+
i46*selfcare.dummy[1]*pain.dummy[1]+
i47*selfcare.dummy[1]*pain.dummy[2]+
i48*selfcare.dummy[1]*pain.dummy[3]+
i49*selfcare.dummy[2]*pain.dummy[1]+
i50*selfcare.dummy[2]*pain.dummy[2]+
i51*selfcare.dummy[2]*pain.dummy[3]+
i52*selfcare.dummy[3]*pain.dummy[1]+
i53*selfcare.dummy[3]*pain.dummy[2]+
i54*selfcare.dummy[3]*pain.dummy[3]+
i55*selfcare.dummy[1]*depression.dummy[1]+
i56*selfcare.dummy[1]*depression.dummy[2]+
i57*selfcare.dummy[1]*depression.dummy[3]+
i58*selfcare.dummy[2]*depression.dummy[1]+
i59*selfcare.dummy[2]*depression.dummy[2]+
i60*selfcare.dummy[2]*depression.dummy[3]+
i61*selfcare.dummy[3]*depression.dummy[1]+
i62*selfcare.dummy[3]*depression.dummy[2]+
i63*selfcare.dummy[3]*depression.dummy[3]+
i64*usualactivity.dummy[1]*pain.dummy[1]+
i65*usualactivity.dummy[1]*pain.dummy[2]+
i66*usualactivity.dummy[1]*pain.dummy[3]+
i67*usualactivity.dummy[2]*pain.dummy[1]+
i68*usualactivity.dummy[2]*pain.dummy[2]+
i69*usualactivity.dummy[2]*pain.dummy[3]+
i70*usualactivity.dummy[3]*pain.dummy[1]+
i71*usualactivity.dummy[3]*pain.dummy[2]+
i72*usualactivity.dummy[3]*pain.dummy[3]+
i73*usualactivity.dummy[1]*depression.dummy[1]+
i74*usualactivity.dummy[1]*depression.dummy[2]+
i75*usualactivity.dummy[1]*depression.dummy[3]+
i76*usualactivity.dummy[2]*depression.dummy[1]+
i77*usualactivity.dummy[2]*depression.dummy[2]+
i78*usualactivity.dummy[2]*depression.dummy[3]+
i79*usualactivity.dummy[3]*depression.dummy[1]+
i80*usualactivity.dummy[3]*depression.dummy[2]+
i81*usualactivity.dummy[3]*depression.dummy[3]+
i82*pain.dummy[1]*depression.dummy[1]+
i83*pain.dummy[1]*depression.dummy[2]+
i84*pain.dummy[1]*depression.dummy[3]+
i85*pain.dummy[2]*depression.dummy[1]+
i86*pain.dummy[2]*depression.dummy[2]+
i87*pain.dummy[2]*depression.dummy[3]+
i88*pain.dummy[3]*depression.dummy[1]+
i89*pain.dummy[3]*depression.dummy[2]+
i90*pain.dummy[3]*depression.dummy[3]
$

Thanks a lot in advance,
Anna

Dear Anna,

First of all, please make sure that you correctly specify the dummy coefficients, so you should write t1.dummy[0|0], not just t1.

Second, it is not possible to estimate all interactions with dummy coded variables. For example, note that dummy codes for the base level are always equal to zero, so an interaction with zero will always be zero, and therefore such an interaction cannot be estimated (especially if it appears in both alternatives). This is not a limitation of Ngene, but rather an identification problem for logit models. So you need to think very carefully which interactions you can include (and estimate) and which ones you cannot include. Below is an example of a model that you can estimate, and Ngene will generate a D-error. If Ngene does not generate D-errors (i.e. they are Undefined), then the model can very likely not be estimated.

Code: Select all

Design
;alts=alt1, alt2
;rows=200
;eff=(mnl,d)
;model:
U(alt1)=b1+t1.dummy[0|0]*mobility[1,2,3]+
t2.dummy[0|0]*selfcare[1,2,3]+
t3.dummy[0|0]*usualactivity[1,2,3]+
t4.dummy[0|0]*pain[1,2,3]+
t5.dummy[0|0]*depression[1,2,3]+
i1*mobility.dummy[1]*selfcare.dummy[1]+
i2*mobility.dummy[1]*selfcare.dummy[2]+
i4*mobility.dummy[2]*selfcare.dummy[1]+
i5*mobility.dummy[2]*selfcare.dummy[2]  /

U(alt2)=t1.dummy*mobility[1,2,3]+
t2.dummy*selfcare[1,2,3]+
t3.dummy*usualactivity[1,2,3]+
t4.dummy*pain[1,2,3]+
t5.dummy*depression[1,2,3]+
i1*mobility.dummy[1]*selfcare.dummy[1]+
i2*mobility.dummy[1]*selfcare.dummy[2]+
i3*mobility.dummy[1]*selfcare.dummy[3]+
i4*mobility.dummy[2]*selfcare.dummy[1]+
i5*mobility.dummy[2]*selfcare.dummy[2]+
i8*mobility.dummy[3]*selfcare.dummy[2]
$


Dear Michael,

Thanks for your answer. If to leave only levels 2 and 3 for the interactions, the design can be found.
Now I would like to ask additional question: Is that possible to fix two attributes in each task in such a way, that two attributes are fixed on the same level and only the other three attributes are different? For example, in the 1st task: attribute 1 is fixed at level 1 for both alternatives ,and attribute 4 is fixed at level 3 in both alternatives, all other attributes are different in the alternatives. In the second task: attribute 2 is fixed at level 2 for both alternatives ,and attribute 4 is fixed at level 1 in both alternatives, etc. This should be held for all 200 choice tasks.

Thanks in advance,
Anna

This sounds like what so-called "partial profiles" do, in which certain attributes are kept fixed within choice tasks. Ngene cannot do this unfortunately.

Michiel