19.6 Airbnb median model

tidy(airbnb_model_2, effects="fixed")

  term                  estimate std.error
  <chr>                    <dbl>     <dbl>
1 (Intercept)             1.92     0.305  
2 walk_score              0.0166   0.00342
3 bedrooms                0.265    0.0139 
4 rating                  0.221    0.0282 
5 room_typePrivate room  -0.538    0.0224 
6 room_typeShared room   -1.06     0.0583 

This leads to a median model:

\[ median(log(\text{price})) = (1.9 + 0.017 \text{ walk_score}) + 0.27 \text{ bedrooms} + \\ 0.22\text{ rating} - 0.54 \text{ private_room} - 1.1 \text{ shared_room} \]

• For example, the median log price increases by about 17% for every 10 points increase in walkability.

tidy(airbnb_model_2, effects = "ran_pars")

  term                        group        estimate
  <chr>                       <chr>           <dbl>
1 sd_(Intercept).neighborhood neighborhood    0.202
2 sd_Observation.Residual     Residual        0.366

• These are estimates of the group level variation and within group variation

• Book notes that the ‘unexplained’ neighborhood to neighborhood variation is less now that we have included the group-level predictor (makes sense!)