U.S. FERTILITY PATTERNS


Module 1: Using CHIPendale

What follows is a sequence that will be used in clas today. This practice exercise is important because you will work in small groups using this software to answer similar types of questions. If you are interested in learning more about the software and data visit the website "www.psc.lsa.umich.edu/SSDAN".

IMPORTANT NOTE: If you are using a MAC you need to restart the machine and hold down the SHIFT key as the machine restarts. The purpose is to disable the extensions. If you do not do this the software will not work! !

To access CHIP put the disk in the drive.

Double click on the Student Chip ICON FILE, OPEN, FREYCEN(DIR), OPEN, CENTREND(DIR), OPEN, BORN5090, OPEN

COMMAND9 ALL MARGINALS
COMMAND, CROSSTABS
CHILD, OK, YEAR, OK
OPTIONS, PERCENT DOWN
OPTIONS, PERCENT ACROSS

OPTIONS, EXIT
FILE, CLEAR OR QUIT

CONTROL for additional variables:

If we want to further investigate fertility patterns, we can assess fertility varies for particular subgroups. We can observe the effect that a third variable has on the process. Often we want to make comparisons within specific groups. We call this "controlling" the original table for a variable, such as education. In effect, we are examining the fertility differences under the controlled conditions, all women have the same education.

To do this simply select the OPTIONS menu and select CONTROL. From there, you select the variable you want to control, click the OK button. Then select NO MORE, and click the OK button one more time. Then in the OPTIONS menu select PERCENT DOWN or PERCENT ACROSS depending upon your table of interest. You will have as many tables appear as categories exist in your control variable. By clicking the mouse you can scroll through your new tables.

OPTIONS, CONTROL
Select variable, OK
Select No More, OK
OPTIONS, PERCENT DOWN

MODIFY a variable:

In the example there are 8 categories for the CHILD variable. It is possible to modify any of the variables included in the data file. You can combine categories of a variable or omit categories you don't want.

For example, we may want to combine the categories of one, two, three or more children into one group. To do this, go to the MODIFY menu (main menubar) and select the COMBINE option. Next, select the variable to be mod)fied. In this case you will select CHILD and then click the OK button. Then you will have to select the categories to be mod)fied. In this example select the One, Two, Three, Four, Five, Six or more, hitting the OK button after each one. Then select the NO MORE, and hit the OK button one more time. You will then be asked to give your new category a label. I have chosen SOME for the combined categories name. You can use the ALL MARGINALS options, located under the COMMAND menu, to check the results of your modifica_ions.
LAB EXAMPLES

BORN5090

VARIABLES:
Education: LTHS=less than high school
HSGRAD=high school graduate
SOMECOLL= some college
COLLGRAD=college graduate

Child:

QUESTIONS:

NoneEM=No children ever-married
NoneNM=No children never-married


1. Did women in 1990 have fewer children than in 1950? What percentage of women in 1950 had six or more children? What percentage of women in 1990 had six or more children?

2. Are women in 1950 more likely to have no children than women in 1990? What percentage of women in 1950 had no children? What percentage of women in 1990 had no children?

3. In 1990 did women with lower education levels have more children than women who are college graduates? Control for education! What percentage of women in 1990 with less than high school degree had 6 or more children? What percentage of women in 1990 who were college graduates had 6 or more children?

***************************************************************************** *************

FILE: KIlDEMP-9

VARIABLES:
Employment: NIFL=not in the labor force
Unempd=unempioyed
EmpFull=employed full-time
EmpPart=employed part-time

Kid: KidsOt=Children over age 6
Kids<6=Preschool age children
None=No children

QUESTIONS:

1. Are women with no children more likely

to be employed full-time or part-time? CROSSTAB
KID and EMP.
What percent of women with no children are employed full-time and what percent are employed
part-time?

2. Are women who are employed full-time more likely to be childless than women who are employed part-time? CROSSTAB KID and EMP. What percent of women employed full-time have no children and what percent of women employed part-time have no children?

3. Are women employed full-time with no high school degree more likely to have children than women employed full-time with a college degree? CROSSTAB KID and EMP, CONTROL for EDUC. What percent of women with less than a high school degree who are employed full-time have children and what percent of women with a college degree who are employed full-time have children?