Skip to main content

Table 2 Results comparison between the different estimation methods

From: The specialty choice of medical students in China: a stated preference experiment

Factors Total counts Individual proportion Conditional logit Shares of preferences K-means clustering
Most attractive Least attractive BW scores Standardised BW Most attractive Least attractive BW scores Standardised BW SD Cluster 1 Cluster 2
Internal medicine 364 44 320 0.337 1.916 0.232 1.684 0.337 0.397 0.955 0.182 2.083 1.159
General surgery 481 65 416 0.438 2.532 0.342 2.190 0.438 0.469 1.209 0.235 1.398 3.232
Ob & Gyn 277 153 124 0.131 1.458 0.805 0.653 0.131 0.516 0.396 0.104 1.472 −0.427
Paediatrics 251 173 78 0.082 1.321 0.911 0.411 0.082 0.527 0.303 0.095 1.907 −1.561
Ophthalmology 174 113 61 0.064 0.916 0.595 0.321 0.064 0.354 0.246 0.090 0.194 0.488
Otolaryngology 112 208 −96 −0.101 0.590 1.095 −0.505 −0.101 0.423 −0.219 0.056 −0.898 0.012
Oncology 151 197 −46 −0.048 0.795 1.037 −0.242 −0.048 0.422 −0.079 0.065 −1.352 1.220
Geriatric medicine 21 266 −245 −0.258 0.111 1.400 −1.290 −0.258 0.315 −0.644 0.037 −0.963 −1.720
Dermatology 70 313 −243 −0.256 0.368 1.647 −1.279 −0.256 0.421 −0.624 0.038 −1.611 −0.841
Psychiatry 53 403 −350 −0.368 0.279 2.121 −1.842 −0.368 0.420 −0.922 0.028 −2.250 −1.305
Rehabilitation 136 155 −19 −0.020 0.716 0.816 −0.100 −0.020 0.399 0.000 0.070 0.019 −0.256
sum - - - - - - - - - - 1.000 - -
  1. Note: the number in italic style is the baseline level for conditional logit