ABSTRACT
All statistics have errors and even the best studies have unexpected glitches. Scientific reports and government surveys always have a methodology or design section that explains how the study was conducted. Some news outlets have a “poll of the day” on their website or broadcast, where viewers can click, text, or call to volunteer their opinions about an issue. News outlets state that these polls are unscientific and do not represent public opinion, but sometimes other people quote the statistics without mentioning their unreliability. Probability samples that have nonresponse can sometimes yield statistics that are far from the population values. The statistics published in the article accurately reflect the percentages for the 4,156 persons who responded to the survey. The response rate, which is the percentage of persons selected for the survey who provided answers, provides some information about the quality of a survey, although it is not always a reliable measure of bias.
