Statistical Analysis

By Christine Aclan, Research Assistant

What can we tell by looking at particular indicators in the survey? Do any of the indicators have a large significance in relation to a city’s overall grade? We ran several statistical analyses to determine whether there was any relation between individual indicators and a city’s final grade in each category and overall rank. All of our indicators were weighted evenly in determining the final grade, but we wanted to know which indicators would show the largest correlation to the overall grade.

We performed a correlation study to determine the strength of the relationship between the final grades in each category (Population, Health, Education, and Community) and each independent variable in the category. Figures 2, 3, 4, and 5 show the results. Correlation values close to + 1.0 or – 1.0 suggest a strong relationship. For example, the percent of a city’s population with a high school diploma shows a high correlation to a city’s education grade (0.656). This means that a city with a high percent of high-school educated citizens scored high overall in education, while a city with a low percent of high-school graduates scored low overall. Correlation values of less than 0.40 indicate a weak relationship and are essentially insignificant. Reading proficiency data shows an inconsequential correlation to a city’s education grade at 0.305. We included indicators with both a high and low correlation in the charts for you to see the full range of the analysis.

We then used the correlation values to determine the coefficient of determination (the square value of the correlation coefficient), which shows the percent that a particular indicator helped to determine the overall score. You can see the results in Charts B, C, D, and E.

We also conducted a correlation study for all of the indicators in relation to the final overall grade. The results are in Figure 1. We used those numbers to get the coefficient of determination, shown in Chart A. We found that the percent births to teens, recycling policies, and the percent of the population with a high-school diploma had the highest correlation, and the greatest determination towards the overall grade. We left the percent increase in urbanized area and Population Education indicators off the chart, as those indicators have a negative correlation.

As in years past, the percent births to teens had the highest correlation and coefficient of determination percentage in relation to the overall grade. This means that one can look at the birth to teens indicator and make a reasonable guess as to the city’s overall grade – if a city has a low rate of births to teens, they probably got a good grade overall, while a city with a high teen birthrate most likely scored poorly.

Births to teens serves as a signpost – if a city has its teen birthrate under control, it has probably made positive steps in many other areas as well. Our data show that a city with a low teen birthrate probably also has good schools and good health care. If a city has a high teen birthrate though, it probably means that the city also has a poor health care system and a poor educational system.

The data show a relationship between to births to teens and other pressing social issues -- issues that contribute to the kid-friendliness of a city. We don't necessarily know which came first -- high teen birthrates, poor schools, high poverty, or a lack of adequate health care. But we do know that all of these issues are connected, and fixing one problem area can lead to an improvement in other areas.

Since teen pregnancy has a high correlation to a city's overall grade, then reducing teen pregnancy might be a good place to begin in making our cities more kid-friendly. To reduce teen pregnancy we must improve access to reproductive health care for teens and improve the quality of sexuality education in the schools. Population Connection's activists have been working hard on these issues for years -- it's time for everyone who is concerned about building a more kid-friendly country to join us in this mission.