diff --git a/DESCRIPTION b/DESCRIPTION
index 1659aa0..45a5367 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: miceFast
 Title: Fast Imputations Using 'Rcpp' and 'Armadillo'
-Version: 0.9.1
+Version: 0.9.1.9000
 Authors@R: person("Maciej", "Nasinski", email = "nasinski.maciej@gmail.com", role = c("aut", "cre"))
 Description: 
   Fast imputations under the object-oriented programming paradigm. 	
diff --git a/NEWS.md b/NEWS.md
index ada4fc3..76e49ca 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,3 +1,7 @@
+# miceFast 0.9.1.9000
+
+* Update the Little test description in the missing data and imputation vignette.
+
 # miceFast 0.9.1
 
 ## Bug fixes
diff --git a/vignettes/missing-data-and-imputation.Rmd b/vignettes/missing-data-and-imputation.Rmd
index 6939721..80b0479 100644
--- a/vignettes/missing-data-and-imputation.Rmd
+++ b/vignettes/missing-data-and-imputation.Rmd
@@ -76,11 +76,12 @@ $Y_{mis}$ are the missing values.
 to the measurements.
 
 **Testing for MCAR:**
-MCAR is the only mechanism that can be tested from the observed data.
+
 Little's (1988) test compares the means of observed values across
-different missingness patterns; a significant result rejects MCAR.
-However, failure to reject does not prove MCAR. The test has limited
-power, especially with small samples or few patterns. It is a necessary
+different missingness patterns. The failure to reject does not prove MCAR. 
+The Little's test is not a satisfactory way to determine if a missingness is MCAR. 
+For example, in the case of omitted variables. We always have to understand the missingness process.
+Moreover, the test has limited power, especially with small samples or few patterns. It is a necessary
 check, not a sufficient one.
 
 **Implications:**