You are on page 1of 2

1)Variance and mean characterize a sample or a variable in a sample?

2)What is a residual?

Select one or more:

1) The deviation of an observed value form is a theorical value.


2) Extreme values remove from the analysis.
3) Statistics describing a good model
4) Statistics describing a bad model
5) A well to access if your model explains well the data or if it lacks some inputs variables.

3) What is a normal value of a variable? Why 100% coverage intervals are rarely, if ever used?

4)Which is larger, the variance of the residuals, the variance of the models values, or the variance of
the actual values?

5) Comment the graphic below highlight its problems and make an improved version of it

6)The following questions need to be answered using the GaltonFamilies dataset.

First question: What R expression gives the percentile 85 of childrens heights?

7) Provide a 50% coverage interval of fathers heights?

8)And of mothers heights?

9)How many girls are larger than their mothers?

10)And boys smaller than their fathers?


11)Create a variable, say hdiff, with the difference of heights between children and mothers

12)And the median?

13) Its standard deviation?

14)Calculate a 50% coverage interval for hdiff

15)Make a boxplot(box and whisker) of hdiff distribution

16) Calculate and represent graphically the distribution of the number of children by family. Whats
the mean and the median?

17)Use the hills dataset(from the R package MASS) describe graphically (at least one plot
expected)and numerically the data. Can you see some pattern on them?

You might also like