 Hello and welcome to the session. In this session we will discuss correlation versus causation. First of all we shall discuss correlation. We know that correlation describes the relationship between two variables or we can say that correlation describes how variables change with respect to each other that relationship can be weak, moderate, strong, negative or positive. For example we can say marks in our subject increase with increase in class attendance. By looking at the scatter plot for our bivariate data we can tell how the two variables x and y are related to each other and we can tell if there is strong or weak correlation between the two variables but the question arises does the correlation tell us what causes this relationship between the variables x and y the correlation between two variables does not necessarily mean that one caused the other or that they are actually related in real life correlation between two variables means that there is some sort of mathematical relationship between the two this means that when we plot the values we can see a pattern and make prediction about what the missing values might be what we don't know is whether there is an actual relationship between two variables and we certainly don't prove whether one caused the other or if there is some other factor at work thus we define causation as when change in one variable causes the change in other variable for example smoking causes cancer the study have revealed that the people who smoke have higher risk of cancer so there is high positive correlation between the variables smoking and cancer so we can have correlation with causation and correlation without causation first we discuss correlation with causation the variables x and y may be correlated when x causes y to change or vice versa if we consider the following statement working hours has a positive correlation with earnings if we analyze this statement we can clearly see that there is a relationship between working hours and earnings if a person works for less number of hours then he will earn less and if a person works for more number of hours then he will earn more thus with increase in working hours the earnings also increase and with decrease in working hours the earnings also decrease so here change in number of working hours causes earnings to change so there is a positive correlation so there exists a correlation with causation here the independent variable that is number of working hours changes first followed by a change in the dependent variable that is earnings there is no third variable that causes this relationship now we shall discuss correlation without causation let us consider the following statement that is street food consumption has a positive correlation with crime if we think logically we can see that there is no causation in this case that is street food consumption has nothing to do with crime here both variables increase at same time but some third variable controls them it means because of some third variable there is a correlation between these two variables that third variable may be climate which causes the other two variables to change thus the two variables are influenced by a third variable so there is correlation without causation thus after above discussion we conclude that it is wrong to assume that statistical correlation implies causation but the reverse may be true that is causation may lead to statistical correlation this completes our session hope you enjoyed this session