Description
Problem 1. The file Utility.dat contains a monthly record of
telephone, electricity, and fuel costs for several years. Naming the
5 columns as Month ,Year, Telephone, Electricity and Fuel,
respectively.
(a) Import data set into SAS using the INFILE statement. Add the
title ‘Descriptive Statistics for Utilities’ to the dataset and print out
the first ten observation of the dataset with the title.
(b) Report only the three descriptive statistics mean, s.d. and t-test
for being zero for variable Telephone and Fuel. Use PROC
Univariate. Please provide only relevant information from your
SAS output, rather than copying everything from your output
window without screening.
Problem 2. The file China.dat contains export and import
information (in dollars) by year. Naming four columns as Years
Total, Exports and Imports.
(a)Import data set into SAS using the INFILE statement. Create a
new variable Trade Balance, where Trade_Balance = Exports –
Imports. Print the first ten observation of the whole dataset,
including the new variable.
(b) Show these three histograms for Imports, Exports, and
Trade_Balance, using PROC UNIVARIATE.
(c) Histogram, by definition, is a type of bar plot that group whole
data into bins. In the histogram of Exports from (b), there are 6
bins by default, and each bin with the same size of roughly 8 units,
in terms of range of Exports. You may check the fact from (b).
Now we try to produce a finer description of the data, in another
word, more bins are need in the histogram of Export. Please draw a
new histogram of Export, each bin with size of 4 units. (hint:
histogram … midpoints=…by.)
Problem 3. The file Handinj.dat contains the costs (in Irish
pounds) and lost work days due to hand injuries for workers in
Dublin, Ireland. Naming the four columns as ID,Type, Days and
Cost, respectively;
(a) Import the dataset using the INFILE statement.
(b) Create a scatter plot of Days and Cost using PROC PLOT, with
Days on y-axis and Cost on x-axis. Then please label each point in
the scatter plot by values of Type, i.e. given an observation of
Days and Cost, a point in the 2-dimentional scatter plot is fixed.
Then assign ‘w’ to the point if the observation of Type is ‘work’
and ’s’ if ‘sport’. Based on the scatter plot, is it possible to
seperate these two classes(Type = ‘work’ and Type = ‘sport’) by a
linear function on the plane. If true, draw the line on the plot,
otherwise, venture a guess what function could separate them?
Here, a line seperates two class of points is defined as one class of
points sitting on one side of a line and the other class on the other
side.
Problme 4. The file Debate.dat contains the survey responses from
debate students at various schools. Naming these variables as
SurveyNumber, School, Gender, Comparison, Argumentation,
Research, Reasoning, and Speaking.
(a) Write a SAS program to read the above data set into SAS using
the INFILE statement.
(b) Generate a two way table (with PROC FREQ) for Gender and
Research when School=7. And compute the chi-square statistic
to test independence of Gender and Research. What is your
conclusion?
