So here we are with our first four definitions. I'm going to tell you about descriptive statistics, inferential statistics, population, and sample. So what is descriptive statistics? Now we've looked at this paper before, the one by Donald and colleagues, where they looked at HIV encephalopathy. Now, they included a certain number of children in their trial, and there was actually 87 of them. But they also looked at the gender, the age, the percentage who started the antiretroviral therapy before a certain age. Were there any delays in their developmental milestones? So what did they do in essence? They just described the sample set. Now, descriptive statistics does that. It provides a summary for us, a summary of everyone who was involved in that study. And human beings really love summaries, don't we? Because look at this chart of 20 numbers on screen, 20 values. It's very difficult for a human being just to look at those 20. And imagine even more. And to make some kind of sense of all of those numbers. But we can describe them. I can tell you that there's a sample size. There are 20 of them. Usually we'll use the term n, n equals 20. The average is 35.7. So I can summarize all those values. I can describe all of those values. And that is descriptive statistics. Now let's move on to what the difference is to inferential statistics. Now I'm going to use that same article by Donald and colleagues. And the last paragraph of the results section, remember we said they compared they formed some groups and they compared some values between those. Now this is more than a mere summary, I think you can agree. They went through a certain app, they investigated a sample, they made groups and they compared those samples and they gave us the result. But so what? What does that mean? Well we want to use their results and we want to infer it on a larger population. We want to sit in a clinic, a different clinic and use their results of their sample into our population. Now of course, we can't investigate the whole population, not everyone on the face of the Earth. So we take small little samples and we analyzed them and we inferred those results onto a larger population. Now remember also that's why it's not precise, purely because we only dealing with a small sample, and not the whole population. So what is this population? Is that everyone? Well really it is, it's all 7 billion of us, if we're talking about human beings in the face of the Earth. But we've gotta bring a bit of context to it. Maybe the whole population is everyone on Earth with a certain disease. Or an extreme example, you could create a new kind of bacteria and only ones that are alive, those are the ones in your petri dish, that's the whole population. And you can take a few of them and that will be your sample. But there's definitely this difference between sample and a population. A population is everyone and we certainly can never include everyone in a study. What I want to tell you about though is the calculations of data points of a whole population. Imagine, we could take all 7 billion people and take some value, say for instance their age or their white cell count. We could calculate the mean, say the mean of their ages, that mean we would call a parameter. So if we do a calculation on the whole population, that result we call a parameter. Now for a sample. A sample definitely doesn't include the whole population. We're just going to take a few people into our study. It includes some members of the population. Now the accuracy of statistical inference really lies in how well our sample generalizes to the larger population, how well it represent the larger population. And the larger our sample size is, the better it represents our population. That's why you always in statistics we want to make our group sizes as large as possible. Now the calculation remember we had a parameter. If we take a sample and do a calculation say on their mean ages we don't call it the parameter any more. I'm going to use the term a statistic. So any calculation on some data values of a sample we'll call a statistic. And we want to use our statistic to infer to a parameter. In the next section, we're going to move on to the data types and data definition.