When Respondents Lie

In class today, we talked about "social desirability response bias" (SDRB), or innacurate polling results stemming from respondents giving answers that make them seem "socially desirable" but are not necessarily true.

Here's a link to the Wall Street Journal article we discussed. Feel free to play around at the Project Implicit tests, which are available here.

Regarding the "Bradley Effect," you might take a look at some of the following resources, which are just a drop in the vast ocean of coverage this phenomenon received during and just after the 2008 election season:

• "A Brief History of the Bradley Effect" from Time
• "Will the Bradley Effect Be Obama's Downfall?" from ABC News
• "Ghost of Bradley Effect" from NPR's On the Media
• "The Disappearing Bradley Effect" from the Princeton Election Consortium blog
• "Goodbye, Bradley Effect" from the Boston Globe
• "Obama on the Bradley Effect" from CBS News:

Sample Size Calculator

(source)

As we discussed in class yesterday, sampling is the process of selecting a relatively small group of individuals whose opinions will be polled in order to draw conclusions about the opinions of the whole target population.

You can use this sample size calculator to determine how large your sample should be based on the size of your target population and the statistical confidence that is to be associated with your results. To find population sizes within the U.S., the Census Bureau is a good place to start.