polling-cartoonArticles about political polls should include numbers:

Poll of 1,040 People Finds Americans Anxious Over Future, Obama’s Performance

There are over 318,890,000 people in America; does a sample size of 1,040 people truly represent how everyone feels? If you knew how small the sample size was, would you read the article?

Instead, we see this headline on January 28, 2014: Poll Finds Americans Anxious Over Future, Obama’s Performance. Consider this statement printed inside this article in a font-size smaller than 8pt:

The Wall Street Journal/NBC News poll was based on nationwide telephone interviews of 800 adults, including a sample of 240 who use only a cellphone. It was conducted from Jan 22-25, 2014 by the polling organizations of Bill McInturff at Public Opinion Strategies and Fred Yang at Heart Research Associates. Individuals were selected proportionate to the nation’s population in accordance with a probability sample design that gives all landline telephone numbers, listed and unlisted, an equal chance to be included. Adults age 18 or over were selected by a systematic procedure to provide a balance of respondents by sex. The cellphone sample was drawn from a list of cellphone users nationally, and respondents were screened to ensure that their cellphone is their only phone. The data’s margin of error is plus or minus 3.46 percentage points. Sample tolerances for subgroups are larger.

The 2014 mid-term election year promises to produce many Polls in the hopes of changing your behavior and persuading you to vote one way or the other. Be sure to read the fine print of a published poll before believing the hype.