A study was performed in 1994 (see the References section for more information) to further explore the issue of condom use, sexual activity, and exposure to HIV. Some believe that school-based condom availability programs can help prevent the spread of AIDS among school-aged children and young adults.
To assess the program, students from 12 New York City schools and 10 Chicago schools were surveyed. Both school systems are similar in many aspects and they are both ethnically diverse and possess a high drop-out rate. The important difference for the purpose of this study is that the Chicago schools did not have a condom availability program; the New York City schools did. Presumably, differences found in student behavior may be linked to the presence or absence of the condom availability program.
It is important to note that studies of this kind place an extra burden on the researchers: They must locate two populations that differ only in the effect they are studying. Often this is difficult, yet it is ethically impossible to design an experiment on humans when studying life versus death situations. Researchers cannot give a treatment to one group, withhold treatment from another, and then compare the results. They must instead find naturally occurring groups that differ in the effect under study.
The question you will examine in this project is whether sexually active students in New York City are more likely to use a condom than are sexually active students in Chicago. In the following chart, you can see the percent of sexually active students who reported condom use, for boys and girls, in each school system.
The data for the study is as follows:
| New York City | Chicago | |||
|---|---|---|---|---|
| Boys | Girls | Boys | Girls | |
| Total Number Surveyed | 2,495 | 2,769 | 1,972 | 2,292 |
| Percent Sexually Active | 64.4% | 55.3% | 66.9% | 54.3% |
| Number of Sexually Active Responding | 1,200 | 1,172 | 783 | 818 |
| Percent Report Using Condoms | 66.5% | 55% | 61.2% | 50% |
With this data, you can compare the proportions between the two schools. There are four questions that might be asked of this data:
To begin, examine the third question from the preceding list: this question involves testing the null hypothesis that p1=p2. Calculate z as given in the text:
and the values are as follows:
This value of z falls in the rejection region for an alpha value of 0.05. The data show a significant difference for the behavior of boys in the New York and Chicago school systems.
In the following Analysis section, you can use the interactive tools to answer the remaining three questions posed here.