Biased vs. Unbiased surveys (biostatistics)

95% confidence interval of .10 to .21 is calculated based on the sample of the 200 women who agreed to participate in the survey. We do not have any information about the 100 women who did not agree to participate; they could have different characteristics or risk factors compared to those who did participate. Non-response bias could affect the results, and we cannot assume that the proportion of those with a recent STD among the non-participants would fall within the same confidence interval established by the participants. More information about the non-participants would be required to make any statement about their STD proportion. Problem: A survey is performed to estimate the proportion of 18-year old females who have had a recent sexually transmitted disease (STD) defined as an STD in the past year. In a random sample of 300 women, 200 have agreed to participate. Based on these 200 women, a 95% confidence interval for the proportion who had a recently sexually transmitted disease was .10 to .21. Which of the following is true about the proportion who had a recent STD among the 100 who did not agree to participate in the survey: A) The proportion will definitely be in the interval .10 to .21. B) The proportion will definitely not be in the interval .10 to .21. C) Proportion will be in the interval .10 to .21 with 95% confidence. D) No general statement can be made without additional information.
Back to Top