fix typo in visualizing_uncertainty

visualizing_uncertainty.Rmd

@@ -1223,7 +1223,7 @@ ggplot(df, aes(y = y)) +
   theme(axis.text.y = element_text(vjust = 0.5))
 ```
 
-To summarize, a Bayesian credible interval makes a statement about the true parameter value and a frequentist confidence interval makes a statement about the null hypothesis. In practice, however, Bayesian and frequentist estimates are often quite similar (Figure \@ref(fig:bayes-vs-ols)). Once conceptual advantage of the Bayesian approach is that it emphasizes thinking about the magnitude of an effect, whereas the frequentist thinking emphasizes a binary perspective of an effect either existing or not. 
+To summarize, a Bayesian credible interval makes a statement about the true parameter value and a frequentist confidence interval makes a statement about the null hypothesis. In practice, however, Bayesian and frequentist estimates are often quite similar (Figure \@ref(fig:bayes-vs-ols)). One conceptual advantage of the Bayesian approach is that it emphasizes thinking about the magnitude of an effect, whereas the frequentist thinking emphasizes a binary perspective of an effect either existing or not. 
 
 (ref:bayes-vs-ols) Comparison of frequentist confidence intervals and Bayesian credible intervals for mean chocolate ratings. We see that both approaches yield similar but not exactly identical results. In particular, the Bayesian estimates display a small amount of shrinkage, which is an adjustment of the most extreme parameter estimates towards the overall mean. (Note how the Bayesian estimate for Switzerland is slightly moved to the left and the Bayesian estimate for Peru is slightly moved to the right relative to the respective frequentist estimates.) The frequentist estimates and confidence intervals shown here are identical to the results for 95% confidence shown in Figure \@ref(fig:mean-chocolate-ratings).