230 Economics Fall 2005
Don Negri Smullin 307
Homework #2
Due: Thursday, September 15th
I. MINITAB ASSIGNMENT
NOTE: Solutions for all computer assignments must be turned in on "notebook-sized" paper so they can be read and graded easily. Include ONLY the portions of your printout which pertain to the problem and be certain to answer all of the questions. This assignment is intended to help you feel comfortable using MINITAB and interpreting MINITAB output. All the commands you need can be found in the "COMPUTER ASSIGNMENTS IN MINITAB" handout which was distributed in class and available on the class Website. This homework requires the phone log from homework #1. In homework #1 you experienced the tedium of calculating statistics (such as the standard deviation) by hand in the hopes that this would foster a better understanding of these measures. In particular, you calculated the mean, median, and standard deviation of phone calls made on July 9th by a particular broker at Dewey, Churnham, and Burnham. Now we are going to reproduce these results using minitab. Again, I have duplicated this data for you after converting the data to fractions of minutes.
CALL 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 DURATION .85 2.48 2.28 .35 .48 .31 .58 1.21 15.68 1.02 .35 .45 .31 .25 2.12 .58 .45 .38 1.71 1.05
hw assignment 2.doc
DO THE FOLLOWING: BE CERTAIN TO USE THE Save Window as...COMMAND ON THE File menu SO THAT YOU WILL BE ABLE TO HAND IN A COPY OF YOUR WORK. (Note: You may also want to save the data in your spreadsheet. To do that you must use the Save Worksheet as ...command from the file menu. ) 1. Enter the data on length of call into the data spreadsheet. (Alternatively use the SET or READ commands in the session window). 2. Put names in the column headings or use the NAME command to give the variable an appropriate name. 3. Use the PRINT command to verify the naming of the variable and to check the accuracy of your input. 4. Use commands from the "Computer Assignments in Minitab" handout to calculate the total length, the mean length, the median length, and the standard deviation of the length of phone calls made on July 9th. 5. Now use the DESCRIBE command to generate all the same information. 6. Generate the histogram for length of phone calls made on July 9th. a) Use this histogram to comment on the relative merits of the mean and median for summarizing information on the "average" length of phone calls made on July 9th.
It is often necessary to transform data in some way when you are doing computer work. First, type the data for July 8th on "call duration" and "call type" into the spreadsheet-data window - use the second and third columns. (Alternatively, use the READ command to input information on "call duration" and "call type" for July 8th. If you use the READ command, keep in mind that you have already created one variable by using the SET command to input data on phone call length in C1. Therefore, you will want to create a second and third variable by using the READ INTO C2 C3 command.) Remember, you will need to "transform" call duration to fractions of minutes. More importantly, transform qualitative information on call type into quantitative information. To do this let "call type" equal 1 for local calls (that is, input a 1 for all local calls) and let "call type" equal 0 for all other types of calls. variable A with values of only zero and one is commonly called a "dummy variable." 7. Use the PRINT C1-C3 command to print your data. a) From this printout identify two fifteen second local calls made on July 8th.
8. Use the DESCRIBE command to generate information on all three variables. a) From this information calculate the total length of phone calls made on July 8th.
hw assignment 2.doc
b) c)
Interpret the mean of C3. What does this statistic tell us about phone calls made on July 8th? Carefully explain your reasoning. Calculate the mean length of local calls and the mean length of long distance calls made on July 8th. This can be done using the MINITAB subcommand below. (Note that any time you use subcommands you must be careful to remember the semicolon and the period.) MTB> DESCRIBE C2; SUBC> BY C3.
LABOR MARKET DATA FROM THE CURRENT POPULATION SURVEY We have seen that even for extremely small samples, such as this one on phone activity, computers are wonderful tools for "crunching numbers" (though it is my job to make certain that you understand what all this crunching means). However, the tedium of in-putting data remains. Thankfully, once a data set is created and saved it can easily be retrieved for future statistical analysis. The best of all possible worlds is if someone else inputs and saves the data for you. This has been done. You will be using the RETRIEVE command throughout the semester to retrieve and work with data sets that have been saved into my files on the network system. One of the data sets we will be working with is 9281 observations on 25 variables drawn from the June 2005 Current Population Survey. You can access this data set by typing: MTB> RETRIEVE 'L:\dnegri\CPS05' NOTE: Using the RETRIEVE command in this way will retrieve these 25 variables into columns 1-25, thus "erasing" your phone-call data in columns 1-3. You should always be very careful about entering and retrieving your data. Erasing the information in C1-C3 is not a concern here, but this is a rare exception to the rule....