 Other two techniques which are very often used with the ordinal data are called point-bicereal and phi-coefficient. So point-bicereal correlation is when one variable is like nominal scale, like dichotomous data and the other one is continuous data. There are many examples. For example, we want to see the scores and then male and female, what is the relationship between that if male are scoring more or the female are scoring more. So we will change the dichotomous data to zero-one ranking and then we will calculate the correlation by simply using the same method as the Pearson product moment correlation. So a variable with only two values is called a dichotomous variable or a binomial variable. For example, male-female, for example, college graduates versus non-college graduates. For example, older than 30 years or lesser than 30 years, younger than 30 years. So if you have one dichotomous variable in that the one continuous variable, you simply will go for the point-bicereal correlation. To compute the point-bicereal correlation, the dichotomous variable is first converted to numerical value by assigning value of zero to one category and the one to the other category and then regular Pearson product moment correlation is used. So we talk about our square value. Whenever we are using correlation, we usually determine the amount of variance explained by one variable in the other variable by calculating our square, which is also called a coefficient of determination. So similarly for all correlations, we used our square to find out or to predict how much variance is explained in the y-variable with the help of x-variable. So squaring the value of the point-bicereal correlation produces our square, which is exactly the value of our square we obtain measuring effect size of independent sample t-test. The sum of the bucking, maybe our square, we have calculated, our square simply tells you effect size amount of variance, which can be explained by the independent variable into the dependent variable. As with most correlations, the strength of the correlation is best described by the value of our square. Our square is called the coefficient of determination, which measures how much of the variability in one variable is predicted or determined or explained by the association of the second variable. So simply, when we calculate correlation, then we are interested in this that if our one independent variable is explained by x-variable in the y-variable, how much of the percent change is explained. For example, if my smoking or health and I find out the correlation in the smoking or health, and first, we have talked about the correlation value that ranges from 0 to 1. And if there is a negative correlation between them, the more we smoke, the more health it deteriorates. First, the correlation coefficient is minus 0.5. So now, when we look at the coefficient of determination, we will take the square of 0.5. So that will be 0.25. This means that because of smoking, the difference in health, the health when deterioration is coming, that is 25% or 0.25. Or this means that 25% variability is explained in the dependent variable, which was health with the help of just one independent variable which was smoking. So we use any correlation method in it. R-square, that is, coefficient of determination, we find out what it is. But sometimes, in correlation, your IV-DV does not happen. Simply, you want to find out the relationship of two variables. You are not necessarily specified that this is IV and this is DV. Sometimes, you definitely do it. Sometimes, you don't do it because correlation means that it just states that two variables are moving in the same or the opposite direction. You don't infer any causation or causality. Fourth and last technique, which we use a lot in correlation, that is 5 coefficient. What happens in 5 coefficient is that both of our variables acts or y are our dichotomous. Muscle male or female are smoking. So our gender and smoking are two variables. Gender is our dichotomous. And our smoking, smoker or non-smoker are also our dichotomous. So both are dichotomous variables. We want to find out their correlation. We will assign them numbers like point-by-serial. So we will assign one to zero and one to one. For example, male to zero, female to one. Smoker to zero and non-smoker to one. And then we will find out the relationship. 5 coefficient basically is a test of association. It tells us that both are in the same direction or there is no relationship or association between them. And when we convert them to zero, one, and the other variable to zero, one, then we apply the same formula, Pearson product correlation to find out the relationship between the two variables. 5 coefficient can be used to assess the relationship between two dichotomous variables. The more common statistical procedure is the chi-square statistic to calculate it. So we don't have to see the significance of it. We mainly use chi-square test to find out the relationship between or to find out the association between two dichotomous variables. But the technique or procedure behind it is our Pearson product moment correlation. So this was some overview about other correlational methods. You see that if our scale is an ordinal variable, it is very important to us. I say that if you really want to be a pro in the start, then mainly you have to understand the scale of the measurement because we have to go into your research and use what design we have to use, what analysis we have to use, how we have to find out the correlation. It is very important that you understand what is the level of measurement of our variables. So nominal-ordinal interval ratio, you must always remember, that will help you not only in deciding which statistical test to be used, but also for your research design, for your sampling to move ahead. It is important because this is the basic to decide how you're going to do your research.