 We have already covered descriptive statistics. Today, we will start inferential statistics. Inferential statistics basically allows you to take a sample from the population, study it, do all the descriptives on that and then make a generalization or draw inference about the bigger population. As I have already emphasized in my lectures, as a researcher and as a scientist, we are never interested in the smaller sample that we have extracted. Rather, we are more interested in the bigger population overall picture. In inferential statistics, we draw samples from a population and study the sample in a good way. Then, we draw inference from the samples and draw conclusions about the population and we generalize those results toward the bigger population. For example, if I want to study screen time is increasing in the children and what are the effects of that in youth, in adolescence. For example, I want to study the relationship between aggression and the screen time. I will draw a very representative sample using a required sampling technique and then after extracting the sample, I will see how much screen time I will study through questionnaires, whatever. And then mainly, whatever sample I have taken of 100 students or 200 students, but overall, I will be more interested in generalizing whether there really is any relationship with the aggression of screen time and inferential statistics allows me to do all this. So, primarily in inferential statistics, there are two areas. Number one, estimating population parameters and number two, testing hypothesis. So, estimating population parameter means that the statistics I have calculated, for example, average screen time in young adolescents and average aggression in young adolescents, I have to estimate that overall, in bigger populations, in adolescents, within Lahore city or maybe within Pakistan, what is that? So, I will estimate population parameter from the sample statistics. And the second area is to test my hypothesis. Hypothesis testing means that I will make an assumption or a research hypothesis about my sample and I will basically test that hypothesis using different techniques and different tests. In that, in testing hypothesis, we will study T-test, Z-test, analysis of variance, which we also call F-test. We will also do model testing, regression, correlation. We test all these things using different tests to see if our research hypothesis is true or false. So, we will make an assumption, we will make null hypothesis, then we will make research hypothesis. We will test those hypothesis using different test statistics like FT, Z, regression correlation or we will draw inferences once we test the hypothesis. We will draw conclusions, we will draw inferences and we will generalize the results toward that population which we call parent population sample. So, all these topics we will cover in inferential statistics. One more time, we will talk about inferential statistics whose researcher or a statistician or a scientist have to take into account is the sampling error. What is sampling error? Sampling error means that you drew a sample from a large population and you calculated all the statistics on it. And then, actually, the parameters of the population and the characteristics of the population, we actually are estimating those with the sample statistics which mostly are not equivalent. They are close, they can be a little distant but they don't exactly have the same values that our population has that truly exist. So, the difference between sample statistic and population parameter, that difference is known as a sampling error. And the less the sampling error is, our inferences and conclusions will be more reliable. And the more errors we have, the more chances that our type 1, type 2 errors and different kinds of errors increase. So, there are many techniques to handle a sampling error which we will talk about. But we will mainly cover testing hypothesis and population parameters estimation in inferential statistics.