diff --git a/vignettes/hdmax2_multivariate.Rmd b/vignettes/hdmax2_multivariate.Rmd
index 7e10311..9160754 100644
--- a/vignettes/hdmax2_multivariate.Rmd
+++ b/vignettes/hdmax2_multivariate.Rmd
@@ -86,7 +86,7 @@ plot((pc$sdev^2/sum(pc$sdev^2))[1:10],
      xlab = 'Principal Component',
      ylab = "Explained variance")
 
-K=5 #pca conclusion : it is better to select too many factors that too few
+K=4 #pca conclusion : it is better to select too many factors that too few
 ```
 
 
@@ -196,27 +196,6 @@ hdmax2_step2 = hdmax2::estimate_effect(object = hdmax2_step1,
 ```
 
 
-
-This step use `mediation::mediate` function to obtain several effects estimation:
-
-- ACME Average Causal Mediation Effect
-
-- PM Proportion Mediate
-
-- TE total effect
-
-- ADE Average Direct Effect
-
-This step also compute Overall effects : 
-
-- OIE (Indirect effect)
-
-- ODE (Direct Effect)
-
-- OTE (Total Effect)
-
-And regression effects size.
-
 ### Vizualisation of results
 
 We propose a set of plots which including:
diff --git a/vignettes/hdmax2_univariate.Rmd b/vignettes/hdmax2_univariate.Rmd
index 2ffb1e2..5646162 100644
--- a/vignettes/hdmax2_univariate.Rmd
+++ b/vignettes/hdmax2_univariate.Rmd
@@ -90,7 +90,7 @@ plot((pc$sdev^2/sum(pc$sdev^2))[1:10],
      xlab = 'Principal Component',
      ylab = "Explained variance")
 
-K=8 #pca conclusion : it is better to select too many factors that too few
+K=4 #pca conclusion : it is better to select too many factors that too few
 ```