ECON 452 (A & B) Winter 2001 C. Ferrall / A. Gregory

IN-CLASS LAB WEEK II.1


TO BEGIN


  1. Set Up.
    1. Start Stata (and if Qprofile doesn't start automatically, type Qprofile). Do not fill out the form yet.
    2. Start a second browser window (Ctrl-N) Netscape and go to Project II.1
    3. Start a third browser window (Ctrl-N) and go to Bowlus and Seitz paper
    4. Log on to your qlink account. One log on for each student using the computer.
    5. You should have 5 or more windows open now. Use Alt-Tab to go through open windows.

  2. Registration.
    1. Do step 2 of Proj II. Each person fills out the DLI form if they haven't done so already.
      RESOURCE REQUIRED: data
    2. Each student check their qlink e-mail that the DLI password has arrived.
    3. Go to the Stata window and fill out the Qprofile form. Select
    4. In Stata, enter Qregstr.
    5. Fill in the form. The QED Archive password will be displayed. Remember it.
    6. Repeat Qregstr for each person.
    7. In Stata, type BREAK to finish the session. Choose "OK" to transfer the log file to edith.
    8. In Netscape, go to the 452 log file directory: http://edith.econ.queensu.ca/statausr/logfiles/Econ452/. Find your log file.
    9. Start a new Stata Qprofile session, fill out the form again entering the archive passwords and the password of the day.

  3. The QED Data Archive.
    1. In Netscape, go to the QED Data Archive
    2. Find the Violence Against Women Survey, 1993.
    3. Find the data set file number for this survey
    4. Browse the archive to find the following info about the VAWS:
      1. the sample mean of the variable a17
      2. the description (variable label) of variable v37c
      3. what it means when variable dvlvgrgr=7 (value label for 7)
    5. Copy the answers and enter them into their Stata log as comments (command lines that begin with *).
    6. Example, type in Stata:
      1. *The sample mean of a17 = -0.2334234
      2. *v37c is "favorite breakfast cereal"
      3. *dvlvgrgr=7 means "raised Catholic"
    7. Find the variables in the VAWS that would allow you to select observations based on age, school attendance, marital status, and number of marriages (see Section 2 of the Bowlus/Seitz paper.)

  4. Using Qextract
    1. In Stata, enter Qextract. Enter the file number of VAWS and select "Expand Memory to fit ..."
    2. Choose to "Keep" the QEDid, perwght, prov, and the variables you identified in the last section. Then go on to step 3, and extract how.
    3. Use the keep, drop, and count commands to replicate the numbers 5260,87, and 432 in the paper.

  5. Wrapping up
    1. Enter BREAK and make sure you log file is transferred to edith.
    2. Exit all your open programs, especially your qlink connections.