Table 2. Mixed-effects regression models estimating the associations of depressive symptoms and burnout with medical career interest (N=286)
Variables Model 1 Model 2 Model 3 Model 4
Depressive Symptoms -.07* -.06*    
  [.03] [.03]    
Burnout     -.50*** -.50***
      [.13] [.13]
Gender1        
     Female   -.60*   -.50*
    [.25]   [.24]
Class Year 2        
     Freshman   -.47   -.57*
    [.26]   [.25]
Race/Ethnicity3        
     Black   1.26**   1.14**
    [.43]   [.42]
     Asian-American   .03   .14
    [.41]   [.40]
     Hispanic   .06   -.01
    [.52]   [.51]
  American Indian / Alaska   .63   .46
    Native / Non- Hispanic       Multiracial   [.69]   [.71]
Academic Achievement        
    High school grade point   1.09*   1.00*
    Average   [.47]   [.48]
Socioeconomic status        
    Median income (log)   .16   .08
    [.32]   [.31]
  First generation college   .01   -.14
     Student4   [.34]   [.32]
Time5 -.65*** -.65*** -.54*** -.54***
  [.15] [.15] [.15] [.14]
  BIC 2550.11 2588.92 2528.91 2566.81

Notes: Unstandardized beta coefficients presented with standard errors in brackets. All models include post-stratification weights for gender and inverse probability weights to adjust for non-response at Time 2. Model constant and random-effects not shown. 1 Male is referent group; 2 Sophomore is referent group; 3 White is referent category for race/ethnicity; 4 Non-first-generation student is referent group; 5 Time is coded as 0=September 2015, 1=April 2016.

*p<.05; **p<.01; ***p<.001 (two-sided tests).

Int J Med Educ. 2018; 9:302-308; doi: 10.5116/ijme.5be5.8131