Estimating baysian priors from a pilot study

[TamsinD]

Thank-you so much for your help with estimating the priors. Your explanations have been really clear and helped me understand it better.
My modified syntax for Ngene is now as follows, so will use this for the survey distribution. Sample size is 242.
Thanks again.

design
;alts = opt1, opt2
? efficient design
;eff = (mnl, d, median)
;rows = 36
;bdraws = sobol(1000)

;model:
U(opt1) = b1.dummy[(n,0.485,0.129)|(n,0.199,0.116)|(n,-0.006,0.111)] * x1[0,1,2,3]
+ b2[(n,-0.013,0.006)] * x2[30,40,50]
+ b3[(n,-0.041,0.0014)] * x3[10,20]
+ b4[(n,-0.013,0.006)] * x4[2,5,10]
+ b5[(n,-0.095,0.030)] * x5[2,3,4,5,6]
+ b6.dummy[(n,0.171,0.044)] * x7[0,1]
+ asc1[(n,-0.201,0.546)]
/
U(opt2) = b7[(n,-0.022,0.011)] * x2_2[30,35,40]
+ b8[(n,-0.039,0.047)] * x5_2[1,2,3,4,5]
+ b9[(n,-0.070,0.009)] * x6_2[0,5,10,15]
+ b6 * x7
$