Hey, hi everyone, Phil from StatisticsMentor.com. I don't know about you, but I'm more comfortable dealing with histograms than QQ plots. So we know with QQ plots, we can use it to assess whether our variable follows a given distribution. Typically, we'll look at the normal QQ plot.
I was searching around the internet, I couldn't really find any places where we put all the shapes together. a normal QQ plot, so I've done it here for you. And this is quite a nice picture. This video is in two parts. First, we'll have a quick look at the shapes of the normal QQ plots corresponding to different shape distributions.
And then I'm going to show you how I get these nice pictures. Okay, so the first part of the video. Look, for a symmetric distribution, this is the histogram, this is associated normal QQ plot. This is actually a normal distribution.
You can see that all the dots fall on a line. Since this is applied, we're supposing we applied statisticians, we're not really going to go into these axes. For symmetric distribution with fat tails compared to a normal Q-Q plot, we see that it's only the ends of the ends here, the bottom left and the top right, which kind of move away from that line. The idea is that if it's a normal distribution, that the dots fall roughly on that line, okay, like here.
For a positive skew, another common distribution that we see, we see this upward curve shape. And for a negative skew, we see this, it's like a log kind of curve, isn't it? So it's sweeping up from below like that and beneath the line. So in terms of reading the normal Q-Q plots, This is what we need to know.
Compare the histograms to the associated shape for the normal QQ plot. Right, next I'm gonna show you how I get these two sets of graphs on two separate windows. If you're interested, keep watching. Right, it's in two steps. First I have to simulate the data from various distributions, then I...
Draw the histogram and the normal QQ plot. The command to generate from distribution is very easy in this package R. We use the R followed by the name of the distribution command. So here RNorm generates random variables from the normal distribution.