## Data on the Web

Look at the various data sets available at http://www.gapminder.org/data/.

1. Based on the names of the data sets available, construct several hypotheses that you would like to test using this sort of country-by-country data. Later in the course, you'll learn how to assemble such data in a form you can use for your analysis.
2. Look at one of the data sets (by pressing the little “View” magnifying-glass icon). What are the cases? What are the variables?

## Body Shape Data

Read the the simplified version of the NHANES dataset using

data(nhanes)


You can use help(nhanes) to see t Look at the names of the variables and make sure you understand what is the meaning of each of them.

You're going to be making scatter plots using mScatter(). To generate the graphics in this document, remember to cut-and-paste the command output of mScatter() into a fenced R command in this document.

#### Create a Sample

It takes several seconds to generate a graph using this number of data points. To speed things up, take a random sample of 2000 people and develop your graphs with that.

small = sample(nhanes, 2000)


Then, when you know exactly what you want, you can translate your commands to use the whole data set, if appropriate.

#### QUESTION:

Describe the relationship between height and weight. Is there reason to think that it's different for the two sexes?

#### QUESTION

Describe the relationship between weight and BMI. Is it different from the two sexes? Where do the people with diabetes show up?

#### QUESTION

Is there a relationship between BMI and age? Where do the people with diabetes show up?

#### QUESTION

Is cholesterol level a good predictor of the development of diabetes?

Calculate body-mass index according to the formula $$\frac{m}{h^2}$$ and see how it corresponds to the body mass index in the data. (Hint: You can plot out one versus the other.)