cruk.cam.ac.uk
This app was developed by the Bioinformatics Core of Cancer Research Uk Cambridge Institute to accompany a training course. On the course webpage you will find lecture notes from the course and practical exercises that use this app Introduction to Statistical Analysis




View source Code for app

Data Import Parameters

Are your samples paired?

If your two groups are dependent, you should choose a paired test. If your groups are independant, leave this box un-ticked

Direction of comparison

If A vs B is selected, the difference will be the first column minus the second. Selecting B vs A will compute the second column minus the first You can choose to transform the data prior to statistical testing 

            

          

        

      

      

        

          

            

              
Display Parameters

              

                

                  
                

              

              You can use the algorithm in R to guess how many bins to use in the histogram
              

                

                  
                

              

              Otherwise, you can choose your own number of bins
              

                
                
              

              

                

                  
                

              

            

          

          

            The boxplot and histogram of the data are shown below
            

            
Basic Summary

            

            

            

          

        

      

      

        

          

            

              If you have selected a paired analysis, you will be able to assess the distribution of the differences here
              
Display Parameters

              

                

                  
                

              

              You can use the algorithm in R to guess how many bins to use in the histogram
              

                

                  
                

              

              Otherwise, you can choose your own number of bins
              

                
                
              

            

          

          

            

            

          

        

      

      

        

          

            

              Use the histograms and boxplot to judge whether you need to use a parametric, or non-parametric test
              

                

                  
                

              

              
Variances

              Inspect the histograms and boxplots, or use the result of the F-test to judge whether the variances of each group are approximately the same
              

                

                  
                

              

              Note that changing this option will have no effect for a non-parametric test
              

              
              

                
                

                  

                    
                  

                  

                    
                  

                  

                    
                  

                

              

            

          

          

            
Two-sample test

            

            If you have chosen a Parametric test, the comparison of the calculated test-statistic to the reference distribution will be shown here
            

            
F test for differences in variance

            F test to compare the variances of two samples from normal populations
            WARNING: Please use the result of this test with caution. Sometimes you can better judge differences in variance by inspecting the data distribution
            

          

        

      

      

        

          

            

              
Report Parameters

              

                
                
              

              

                
                
              

              

                
                
              

            

          

          

            
R Script

            You will be able to re-run this analysis in R by downloading the R code below
            We recommend RStudio to run the R code and compile a pdf or HTML report that will show the results of your analysis along with the code used
            
            

            RStudio
            

            The input file that you are analysing must be in your R working directory in order for the script to run
            In order to compile the report in RStudio, you will need to install the tidyverse package
            install.packages('tidyverse'))
            

            Download R Script
            

            Download R Markdown file