 उस्ब नहीं, वैद बाट बीगी वाध में देखा है, लेकिन गेगरेश्चन एकोईजन जब हम बनाते है, तो वो हम में अप यप प्डिक्क्छन, इप ग्रेश्चन एकोईजन के साथ हम एकस की लेएगट बैएगट बनाते है. of the regression it is customary to calculate or compute the standard error of estimate. So standard error of estimate basically gives us accuracy of the prediction. Standard error of estimate gives a measure of the standard distance between the actual y and the predicted y. If it is already data, then why should we do it? Because you definitely have to have data before if you want to calculate, you first take out the correlation and then you predict it with the help of it. But the biggest benefit of standard error of estimate is that it tells you not only the accuracy of prediction or regression but also tells you how many percent of your variance is explained by or caused by the x-variable. So standard error of estimate is very much like standard deviation. Standard error of estimate, if this is the actual y predicted by the distance and then we square it, then it is very much like standard deviation because in standard deviation, we see the distance of the x-score from every mean. How much away is the x-score, each x-score from the mean and then we square it and sum it. So standard error is also very much like standard deviation. To calculate standard error of estimate, we need the sum of squared deviations i.e. y minus predicted y and then its square and then we need the sum. Just like we have written. So sum of squares is commonly called SS or which we call residual. So whenever the term residual is, it means the error between actual and predicted scores. So because it is based on the remaining distance between the actual y-scores and the predicted y-values, we call it sum of squared or sum of deviation and we call it residual. Standard error of estimate obtained, how we have to calculate it, we have to calculate the sum of squared deviations first and then we have to divide it on degrees of freedom. And the degrees of freedom, whether it is correlation or regression since we are talking about two variables, we will divide it from n minus 2. So simply the formula for calculating the standard error of estimate is y minus predicted y, distance of each y from the predicted and then squared and then sum it and then divided by degrees of freedom and minus 2 under root kisaal. Since the standard error of estimate provides a measure of how accurately regression equation predicts the y value, in this case standard distance between actual data points and the regression line is measured by the standard error of distance. Since it is telling us accuracy, the greater the error, the less the accuracy of prediction will be. And the smaller the error, the less the accuracy of prediction will be. So squaring the correlation provides a measure and coefficient of determination we have talked about that when we square the correlation, we get a coefficient of determination r squared which tells us what proportion of the variability in the y variable is predicted by x variable or is explained by or is accountable by y variable. Because r squared may use the predicted portion of the variability in the y score, we can use the expression 1 minus r squared to measure the unpredicted portion. So if our r squared is telling us how predicted variance is, r squared means that in y, how many percent of the change we can predict from the x variable. So how much variance is predicted by x variable in the y, that means we can also take out unpredicted variance. So I have also given an example of smoking and health and we are predicting health from smoking and our r is 0.5 and our r squared is 0.25 so this means that 25 percent variability in the health can be predicted by smoking and if we have to find out unpredictable, then we will simply do 1 minus r squared so how many percent variance that is unexplained or unpredictable we can also calculate. So standard estimate can also be puted by this formula. We can also take out the sum of squared deviation but luckily you don't have to calculate manually you can use SPSS to calculate. So just remember that in our regression what is the logic, what is the reason and how we step forward from correlation towards the real goal of science and psychology making prediction and then we can do it through regression.