 Hello and welcome to the session. In this session we shall discuss variations in the predictions when multiple samples of same size are taken. If someone asks why do we collect a sample then the obvious answer would be we collect a sample to draw some conclusion or to make some prediction. Till now we have seen it with one sample but what happens when we take multiple samples of same size from the same population then we see that the predictions from the two samples can vary. Now we shall determine the variations in the estimate or prediction. Let us take an example. In our school a survey is to be conducted to predict the winner of a school election. Suppose two randomly chosen samples are there we shall name it as sample 1 and sample 2. There are 25 students in each of the sample they are asked to predict the winner of the school elections with the help of the given data. That is using the given data we need to predict whether candidate A will be selected or candidate B will be selected. In both the samples that is sample 1 and sample 2 we can see that maximum students voted for candidate A as we can see in sample 1 16 students voted for candidate A and 9 students voted for candidate B. Similarly in sample 2 there are 14 students who voted for candidate A and 11 students voted for candidate B means that chances for candidate A to win the elections are more than candidate B. So we can predict that candidate A will win the school elections also. Both the samples are randomly selected are representatives of the population so we can say our conclusion is valid. Now if we compare both the samples we see that we were not being favourable to any candidate so we reached the same conclusion from both the samples. Although there is some variation in the data of both the samples that is in sample 1 16 candidates favoured candidate A and in sample 2 14 candidates favoured candidate A. So there is the difference of 2 votes. Now suppose if we take a random sample of 25 students from school and if we take another sample of 25 students who are classmates of candidate B then there will be a lot of variation in both the samples. Then it is obvious that students of second sample will favour candidate B. This means in selecting second sample we were given some preference to candidate B so our conclusion will not be valid. So we can say that if we draw multiple random samples from same population then there may be some variation in the data collected for the multiple samples. Since all the samples are random not biased so the conclusion drawn will be more or less same for all the samples drawn from same population. This completes our session. Hope you enjoyed this session.