I tell my statistics students that as n increased the t-distribution approaches the standard normal distribution. But if you look at my plushies, this is not the case.
The green standard normal plushie is considerably shorter than the t-distribution plushie. I don't agree with this.


I know what the problem is. I used R to create my patterns. Both pdfs go from -4 to 4 on the x-axis. But, when I copied over the images to a image editing software, to clean them up and size them, I didn't keep the heights the same, I should have.
So, even though both are correct, my t-distribution is taller than my z-distribution by about 2 inches. I just need to trim that pattern back. Or make the z-distribution taller. I just don't know which one to do. But I do believe I should do something. These two distributions should look more like each other (when df=200 for t) than they currently do.
0 comments:
Post a Comment