Question

7. You have five quantitative variables and two categorical variables in a dataset. You would like to examine the pairwise relationship between all variables in the dataset.

A) What kind of plots or tables are you going to use and why?

B) How many unique correlation coefficients r are you expecting to calculate? Show your reasoning/calculations. (harder question)

Answer 1

we have 5 Qauntitative variable and 2 catagarical variable, we want to check pair wise relation

A)we use scatter plot :

we can check relation between two variables using  scatter plot..

and relation between atrribute we can drow contengency table.( for chi-square)

B) correlation:

Correlation is usually defined as a measure of the linear relationship between two quantitative variables (e.g., height and weight)

we have 5 quantitative variable suppose ( A,B,C,D,E). we can draw correlation matrix 5*5,each variable pair with each variable,

number of coor is

using nCx= n! /(n-x)! x! formula n= total number, x number of pair.

5C2= 5! / (5-2)!2!

=5! / 3! 2!

=5*4*3*2*1/ 3*2*2

=5*4/2

=20/2

= 10

in 5 variable we can calculated 10 different corr. we use combination because corr(A,B) = corr(B,A) boths meanings are same.

correlation coefficient is a statistical measure of the strength of the relationship between the relative movements of two variables,correlation lies between -1 to 1, if correlation value near by 1 , then we can cunclude that the relation between two variable is strong positive relation, and if correlation value near by -1 , then we can cunclude that the relation between two variable is strong negative relation,if correlation value is zero then we cunclude both variable not co- related to each other..