Measuring "Rareness"
The distribution you created, although beautiful, by itself really tells us nothing. We need to use it to evaluate the actual observed study result (14 of 16). We could report the proportion of times we observed 14 heads in the 1000 repetitions, but as we have larger sample sizes (number of coin tosses), any individual outcome (number of heads) will be increasingly unlikely. So to give us a more "standard" number, that we can compare across studies, we will look at the proportion of results "as or more" extreme. In this study, "more extreme" means even more of the 16 tosses landed heads.
Use the applet to again create a null distribution with at least 1000 repetitions of sets of 16 coin tosses.
- Enter 14 in the As extreme as box.
- Press the Count button.
In your lab report, include a screen capture showing the proportion of repetitions with 14 or more heads in 16 tosses.
(j) Write a sentence which interprets/explains the probability you just estimated as a long-run relative frequency (see Lab 0). This sentence should state the model assumptions, and explain what your probability measures, without using the words “probability”, “chance”, “odds”, “likelihood, or any other synonym for probability.
 |
The previous question is pretty important, use the Week 1 Discussion Board if you are not sure. |
|---|
