Health risks among North Carolina adults : 2001 : a report from the Behavioral Risk Factor Surveillance System  Page 8 
Previous  8 of 77  Next 

small (250x250 max)
medium (500x500 max)
Large
Extra Large
large ( > 500x500)
Full Resolution
All (PDF)

This page
All

4 The demographic characteristics shown in the main tables of this report are generally the characteristics of the person responding to the survey. However, income reflects the reported annual income of the household from all household members and sources. Unless otherwise specified, respondents who answered that they did not know or who refused to answer were not included in the calculation of the percentages. Therefore, the sample sizes used to calculate the estimates in this report vary. The main data tables in this report show the 95 percent confidence interval associated with each percentage ( labeled “ C. I.”). Since the results are based on a relatively small random sample of the total population of adults in North Carolina, the results will be subject to some degree of sampling error. The 95 percent confidence interval shows the range within which we would expect the true value for the entire population to fall 95 percent of the time. For smaller sample sizes ( for example, for a particular age or race group), the confidence intervals will be wider. The confidence intervals shown in this report may not be exactly the same as those that CDC calculates for the same measures. We use a method that may result in non symmetrical confidence intervals, which is more appropriate when the prevalence is close to 0 or 100 percent. Given the complex nature of the BRFSS sample ( i. e., it is not a simple random sample), the SUDAAN software was used to calculate the confidence intervals for the estimates. This software takes the complex sampling design into account when computing the errors of the estimates. In general, any percentage with a numerator of less than 50 will have a relatively large degree of sampling error and should be considered cautiously. Tests of the statistical significance of a difference between two percentages ( for example, between the percentages for two age groups) can be performed after calculating the standard error of the difference. The data user should contact BRFSS staff of the Center for assistance with this calculation. Though not exactly technically correct, a rough approximation of the statistical significance of a difference between two percentages can be derived by comparing the confidence intervals shown in the data tables of this report. If the confidence intervals of the two percentages being compared do not overlap, then it is likely that the difference between the two percentages is statistically significant at the 95 percent confidence level. Stated another way, one can be 95 percent certain that a difference that large would not be observed just due to random variation in the two percentages. Nationwide data for CDC core and optional questions were used to generate maps and trend tables ( if data were available 3 or more years). A clustering program was used to create four groups for the national maps and three groups for North Carolina maps. North Carolina was ranked against all participating states and territories. Organization of the Data All of the data tables in this report are in the same format. The left hand column shows the demographic group, including the ten over sampled counties and three regions, for which the data are displayed. The second column shows the total number of respondents in each category. The next columns show three items for each survey question: the number responding in the specified way to the question, the weighted percentage of respondents with the specified response, and the 95 percent confidence interval of the percentage. Note that the weighted percentage cannot be calculated directly from the unweighted numerator and denominator that are shown in the table. In general, the overall percentage or mean is shown. In the state maps of the United States, the median state value is also shown, consistent with the way CDC often reports the BRFSS data. This is the value where half the states are above and half are below.
Object Description
Description
Title  Health risks among North Carolina adults : 2001 : a report from the Behavioral Risk Factor Surveillance System  Page 8 
Pres Local File PathM  \Preservation_content\StatePubs\pubs_borndigital\images_master\ 
Full Text  4 The demographic characteristics shown in the main tables of this report are generally the characteristics of the person responding to the survey. However, income reflects the reported annual income of the household from all household members and sources. Unless otherwise specified, respondents who answered that they did not know or who refused to answer were not included in the calculation of the percentages. Therefore, the sample sizes used to calculate the estimates in this report vary. The main data tables in this report show the 95 percent confidence interval associated with each percentage ( labeled “ C. I.”). Since the results are based on a relatively small random sample of the total population of adults in North Carolina, the results will be subject to some degree of sampling error. The 95 percent confidence interval shows the range within which we would expect the true value for the entire population to fall 95 percent of the time. For smaller sample sizes ( for example, for a particular age or race group), the confidence intervals will be wider. The confidence intervals shown in this report may not be exactly the same as those that CDC calculates for the same measures. We use a method that may result in non symmetrical confidence intervals, which is more appropriate when the prevalence is close to 0 or 100 percent. Given the complex nature of the BRFSS sample ( i. e., it is not a simple random sample), the SUDAAN software was used to calculate the confidence intervals for the estimates. This software takes the complex sampling design into account when computing the errors of the estimates. In general, any percentage with a numerator of less than 50 will have a relatively large degree of sampling error and should be considered cautiously. Tests of the statistical significance of a difference between two percentages ( for example, between the percentages for two age groups) can be performed after calculating the standard error of the difference. The data user should contact BRFSS staff of the Center for assistance with this calculation. Though not exactly technically correct, a rough approximation of the statistical significance of a difference between two percentages can be derived by comparing the confidence intervals shown in the data tables of this report. If the confidence intervals of the two percentages being compared do not overlap, then it is likely that the difference between the two percentages is statistically significant at the 95 percent confidence level. Stated another way, one can be 95 percent certain that a difference that large would not be observed just due to random variation in the two percentages. Nationwide data for CDC core and optional questions were used to generate maps and trend tables ( if data were available 3 or more years). A clustering program was used to create four groups for the national maps and three groups for North Carolina maps. North Carolina was ranked against all participating states and territories. Organization of the Data All of the data tables in this report are in the same format. The left hand column shows the demographic group, including the ten over sampled counties and three regions, for which the data are displayed. The second column shows the total number of respondents in each category. The next columns show three items for each survey question: the number responding in the specified way to the question, the weighted percentage of respondents with the specified response, and the 95 percent confidence interval of the percentage. Note that the weighted percentage cannot be calculated directly from the unweighted numerator and denominator that are shown in the table. In general, the overall percentage or mean is shown. In the state maps of the United States, the median state value is also shown, consistent with the way CDC often reports the BRFSS data. This is the value where half the states are above and half are below. 