a4

> help_glm_h <- lm(happy_o ~ 1 + drugcond + 
                 emocond + drugcond:emocond,
               data = h_dat) # Model from observed data
               
> help_syn_glm_h <- lm.synds(happy_o ~ 1 + drugcond + 
                           emocond + drugcond:emocond,
                         data = h_dat_s) # Model from synthesized data


> compare(help_syn_glm_h, h_dat) # A comparison of the models

Call used to fit models to the data:
lm.synds(formula = happy_o ~ 1 + drugcond + emocond + drugcond:emocond, 
    data = h_dat_s)

Differences between results based on synthetic and observed data:
                 Std. coef diff p value CI overlap
(Intercept)          -0.7845429   0.433  0.7998578
drugcond              0.8837740   0.377  0.7745433
emocond               1.1781803   0.239  0.6994383
drugcond:emocond     -1.2289564   0.219  0.6864850

Measures for one synthesis and 4 coefficients
Mean confidence interval overlap:  0.7400811
Mean absolute std. coef diff:  1.018863
Lack-of-fit: 1.758649; p-value 0.78 for test that synthesis model is compatible 
with a chi-squared test with 4 degrees of freedom