diff --git a/w02_evaluation/t04_benchmarking.pdf b/w02_evaluation/t04_benchmarking.pdf
index 254a5db..2573058 100644
Binary files a/w02_evaluation/t04_benchmarking.pdf and b/w02_evaluation/t04_benchmarking.pdf differ
diff --git a/w02_evaluation/t04_benchmarking.tex b/w02_evaluation/t04_benchmarking.tex
index ac974da..d0ca235 100644
--- a/w02_evaluation/t04_benchmarking.tex
+++ b/w02_evaluation/t04_benchmarking.tex
@@ -252,7 +252,7 @@
     Test statistic calculated as:
 
     $${\chi_F}^2 = \frac{SS_{Total}}{SS_{Error}} \sim \chi_{k-1}^2 \text{ for
-    large n ($>$15) and k ($>$5)}$$
+    large n ($>$15) and k ($>$4)}$$
 
     For smaller n and k, the $\chi^2$ approximation is imprecise and a look up of $\chi_F^2$ values that were approximated specifically for the Friedman test is suggested.
     \end{frame}