For this part we
will re-analyze the data from the paper "Tele-connecting
local consumption to global land use (Links to an external
site.)."

The data is:

Statistics
Final Project Data.xlsx

(You can also find the same file on the article page.)

This data is a little hard to read, and it may be important to
spend some time looking through it, and figuring out what it's saying.
(Skimming the paper may be helpful too.) We'll be mostly working with the
"Cropland", "Forestland", "Grazingland" and
"Products aggregation".

A: T-tests

This part will use the data from the first tab. Randomly select
17 countries with current GDP/capita greater than $20,000; and 35 with
GDP/capita less than $20,000. Use a t-test to determine whether there is a
difference between the average % share of land use that is foreign for each of
these two types of country. Use the following steps:

A) What countries did you select for each type, and how did you
select them?

B) This test will only be valid if the test statistic is roughly
normally distributed for the countries with GDP/capita greater than $20,000.
Make a histogram of the test statistic for the 17 countries you selected, and
determine whether the test is valid. (Even if it isn't, I'd like you to do the
following questions as an exercise.)

C) State the null and alternative hypotheses.

D) Perform the hypothesis test, record the p-value.

E) Make a decision.

F) Summarize the results in context.

B: Bootstrap

A) Bootstrap a confidence interval of the difference between the
% share of land use that is foreign for countries with GDP/capita above $20,000
and those with GDP/capita below $20,000.

Include your code, or a picture of your sampler.

B) Interpret the results in context.

C: Randomization

Randomly select 12 countries from EU + UK, Switzerland, and
Norway, and 14 sub-Saharan African countries.

For all of these countries, calculate the per capita
land-footprint . Use a randomization test to determine whether geography in
part determines the per
capita land-footprint.

A) Report the null and alternative hypotheses.

B) Paste in histograms representing the two samples

C) Describe the procedure you will use to test the hypothesis.

D) What statistic will you collect (or in R, what statistic will
you record.)

C) Paste your code, or your sampler model.

D) Paste a histogram of the results of the simulation. (That is,
of all the sample statistics.)

E) Find a p-value.

D) Make a decision.

E) Interpret the results in context.

D: Chi-squared.

Choose six sub-Saharan African countries. For these countries,
are cropland footprint, forestland footpring, and grazingland footprint
independent of nation.

A) Paste your contingency table and report the degrees of
freedom your test will use.

B) Report the null and alternative hypotheses.

C) Carry out the test: Find the test statistic and a
p-value.

D) Make a decision.

E) Interpret the results in context.

Get Higher Grades Now

Tutors Online