blog/2023/07/25/conjoint-bayesian-frequentist-guide/

[giscus[bot]]

blog/2023/07/25/conjoint-bayesian-frequentist-guide/

Learn how to use R, {brms}, and {marginaleffects} to analyze conjoint data and find causal and descriptive quantities of interest, both frequentistly and Bayesianly

https://www.andrewheiss.com/blog/2023/07/25/conjoint-bayesian-frequentist-guide/?s=08

[joscani]

Nice post. Thanks for sharing

[yangl-github]

Nice post and thanks for sharing the source code. I saw you use format: html in your yaml header, instead of output: html_document. Can you please advise the difference between the two options? Some reference reading will be good too. Thanks.

[andrewheiss]

That's because I'm using Quarto instead of R Markdown, which uses different syntax for output formats.

[AlekFisher]

This is an awesome post, thank you very much! Have you ever run this on partial profile conjoint, where only a subset of attributes are seen on each screen? I use sawtooth for my conjoint analysis but would love to see if I can do most of my analysis in R.

[Ales-G]

This is really helpful! thank you very much for writing this post!
I have one question, do you know what is the correct way to formally test if differences in marginal means are statitstically significant? so for example, in this contest, how can I test whether the marginal mean for None group is statistically different from the one for Catholic?

candidate_mms_auto <- cregg::mm(
  candidate,
  selected ~ atmilitary + atreligion + ated +
    atprof + atinc + atrace + atage + atmale,
  id = ~ resID
)
candidate_mms_auto %>% as_tibble()
## # A tibble: 40 × 10
##    outcome  statistic feature          level                  estimate std.error     z         p lower upper
##    <chr>    <chr>     <fct>            <fct>                     <dbl>     <dbl> <dbl>     <dbl> <dbl> <dbl>
##  1 selected mm        Military Service Did Not Serve             0.458   0.00907  50.5 0         0.440 0.475
##  2 selected mm        Military Service Served                    0.543   0.00914  59.4 0         0.525 0.560
##  3 selected mm        Religion         None                      0.556   0.0200   27.8 1.88e-170 0.517 0.595
##  4 selected mm        Religion         Jewish                    0.520   0.0186   28.0 2.77e-172 0.484 0.557
##  5 selected mm        Religion         Catholic                  0.526   0.0177   29.7 2.82e-193 0.491 0.560
##  6 selected mm        Religion         Mainline protestant       0.543   0.0194   28.0 3.93e-172 0.505 0.581
##  7 selected mm        Religion         Evangelical protestant    0.437   0.0200   21.9 4.11e-106 0.398 0.476
##  8 selected mm        Religion         Mormon                    0.417   0.0194   21.5 7.40e-103 0.379 0.455
##  9 selected mm        College          No BA                     0.340   0.0186   18.3 1.12e- 74 0.303 0.376
## 10 selected mm        College          Baptist college           0.482   0.0200   24.1 3.89e-128 0.443 0.521
## # ℹ 30 more rows

Thank you in advance for your help on this