Homework 1
Due Wednesday, 9 October 1996

For all of the problems below, the use of statistical programs such as Stata is permitted. All of the data sets in the book are contained on the diskette (in DOS form) that comes with the book. These files can be used with the "infile" command of Stata to read the data into a data set. This Stata data set can be manipulated, saved, and in later visits, can be used directly ("Use" command.)
  1. Selvin, Problem 1.
    For this problem, the following output from Stata may be helpful. [For those who are learning to use Stata, note that the "infile" command would actually all be on one typed line, but is split here for readability.] 
           . infile obs gest sex
         bwt cigs ht wt0 wt1 using ":Stat 227 Aut95:SDS:TBL1-4.DAT"
       (48 observations read)
       . sort sex
       . by sex: summarize bwt

       -> sex=        0  
       Variable |     Obs        Mean   Std. Dev.       Min        Max
       ---------+-----------------------------------------------------
            bwt |      29    3728.103   573.5438       2450       5220  

       -> sex=        1  
       Variable |     Obs        Mean   Std. Dev.       Min        Max
       ---------+-----------------------------------------------------
            bwt |      19    3379.211   411.2186       2900       4270
  2. Selvin, Problems 3, 4.
    Hint: See the contents of box 1.2.

  3. Selvin, Problem 6

  4. Selvin, Problem 7
    Comment: What does the Stata command Correlate X Y, cov do?

  5. Suppose that you are a researcher who is studying whether mother's weight gain during pregnancy can predict the health (or sickness) of the newborn children during the first six months after birth. As part of one of your studies, you have obtained the results given in Selvin, Problem 8. To present your findings at a national meeting, you must submit an abstract of your results, which the program committee will use to determine which researchers will be given speaking time at the meeting.

    A typical abstract in the biological sciences has five sections: background, methods, results, discussion, and references. Abstracts are typically highly restricted in the amount of space they can occupy, often being limited to approximately 250 words.

    For this assignment, write the methods, results, and discussion portion of the abstract, based on the data given in the book. Since your co-authors will have to have some space for the background and references, you have been limited to 150 words for your part of this abstract.

    Although a picture is said to be worth a thousand words, if you wish to include a diagram or picture in your abstract, you may do so. Assume that each picture uses up 25 words worth of space.

    [Your methods section should describe a situation in which the data were collected, and should cite your statistical methods. The discussion should deal only with issues of (a) the conclusions you draw based on the data, (b) any limitations that might apply in interpreting the conclusions, and (c) how strongly the data support the conclusions.]
Last updated 28-Sep-96 by RT
For questions, send mail to r-thisted@uchicago.edu