diff --git a/17-linear_mixed_effects_models1.Rmd b/17-linear_mixed_effects_models1.Rmd
index dc6b6c5..074a344 100644
--- a/17-linear_mixed_effects_models1.Rmd
+++ b/17-linear_mixed_effects_models1.Rmd
@@ -273,11 +273,11 @@ fit.augmented = lmer(formula = value ~ 1 + condition + (1 | participant),
data = df.original)
# compare models
-# note: the lmer model has to be supplied first
+# note: the lmer model has to be entered as the first argument
anova(fit.augmented, fit.compact)
```
-Yes, the likelihood of the data given the linear mixed effects model is significantly higher compared to its likelihood given the linear model.
+Yes, the linear mixed effects model explains the data better than the linear model.
## Additional resources
diff --git a/18-linear_mixed_effects_models2.Rmd b/18-linear_mixed_effects_models2.Rmd
index e107c13..c5336bc 100644
--- a/18-linear_mixed_effects_models2.Rmd
+++ b/18-linear_mixed_effects_models2.Rmd
@@ -289,8 +289,10 @@ df.complete_pooling = fit.complete_pooling %>%
df.no_pooling = df.sleep %>%
group_by(subject) %>%
nest(data = c(days, reaction)) %>%
- mutate(fit = map(data, ~ lm(reaction ~ days, data = .)),
- augment = map(fit, augment)) %>%
+ mutate(fit = map(.x = data,
+ .f = ~ lm(reaction ~ days, data = .x)),
+ augment = map(.x = fit,
+ .f = ~ augment(.x))) %>%
unnest(c(augment)) %>%
ungroup() %>%
clean_names() %>%
@@ -400,8 +402,10 @@ df.partial_pooling = fit.random_intercept_slope %>%
df.plot = df.sleep %>%
group_by(subject) %>%
nest(data = c(days, reaction)) %>%
- mutate(fit = map(data, ~ lm(reaction ~ days, data = .)),
- tidy = map(fit, tidy)) %>%
+ mutate(fit = map(.x = data,
+ .f = ~ lm(reaction ~ days, data = .x)),
+ tidy = map(.x = fit,
+ .f = ~ tidy(.x))) %>%
unnest(c(tidy)) %>%
select(subject, term, estimate) %>%
pivot_wider(names_from = term,
@@ -517,7 +521,7 @@ Let's fit a model to this data now and take a look at the summary output:
```{r}
# fit model
fit.mixed = lmer(formula = value ~ 1 + condition + (1 | participant),
- data = df.mixed)
+ data = df.mixed)
summary(fit.mixed)
```
@@ -687,7 +691,8 @@ n_observations = 10
slope = -10
sd_error = 0.4
sd_participant = 5
-intercept = rnorm(n_participants, sd = sd_participant) %>% sort()
+intercept = rnorm(n_participants, sd = sd_participant) %>%
+ sort()
df.simpson = tibble(x = runif(n_participants * n_observations, min = 0, max = 1)) %>%
arrange(x) %>%
@@ -753,7 +758,7 @@ Let's fit a linear mixed effects model with random intercepts:
```{r}
fit.lmer = lmer(formula = y ~ 1 + x + (1 | participant),
- data = df.simpson)
+ data = df.simpson)
fit.lmer %>%
summary()
@@ -762,7 +767,6 @@ fit.lmer %>%
As we can see, the fixed effect for `x` is now negative!
```{r}
-
fit.lmer %>%
augment() %>%
clean_names() %>%
diff --git a/docs/404.html b/docs/404.html
index 837738f..b965585 100644
--- a/docs/404.html
+++ b/docs/404.html
@@ -23,7 +23,7 @@
-
+
@@ -51,12 +51,6 @@
-
-
-
-
-
-
-
@@ -929,20 +901,6 @@ Page not found
});
-
-