diff --git a/README.md b/README.md
index ef5de37..ff40531 100644
--- a/README.md
+++ b/README.md
@@ -14,7 +14,7 @@ Is it possible to recreate and classify topics assigned to news articles by usin
 
 * Are articles of a certain topic classified more accurately than others?
 
-### Results
+## Results
  
 * The model was run on a linear SVM, with topics generated through LDA.
 * The average accuracy on both the test set and the entire dataset is 79%, with a recall of 78%, with precision ranging from 68% to 93% between the (unevenly distributed) classes.