People lie to researchers and pollsters. For what possible reason? The best way I can explain it is to offer an example. Psychology students set up a booth in a somewhat secluded area and stopped passing pedestrians. The students asked them assorted questions which they really weren’t interested in, then asked the key question, “If you happened along a ten dollar bill laying on the ground, and no one was around, what would you do with it?” The exact percentages escape me now, it is not that important to the point, but a significant portion of the respondents indicated they would attempt to find the person who lost it, or search out a lost and found and turn the money over.
The students continued the polling for a few days then removed their booth. The next day they waited until no one was around and dropped a ten dollar bill in the courtyard and hid out of sight, but still able to see what happened. Without fail, the lucky person looked around and placed the bill in their pocket or purse and continued on their way. They repeated the process, over and over, and observed the same phenomenon – not one person searched out a lost and found or stopped anyone to inquire if they had lost the bill.
This is the difference between “self report” ( what people say they will do ) and “observed behavior” ( what the poll takers actually observed happening), a key principle in psychology. The respondents to the questioning apparently wanted the question taker to think that they were nice, and honest, and all those nice warm and fuzzy attributes .
In this current election, we may possibly be seeing an example of this in the polling data. The issue of race has arisen many times in the news and more than a little finger-pointing has occurred. It has been reported that blacks may be voting as highly as 95% in favor of Barack Obama. To perhaps cover up for this fact the press has gone out of its way to track down and find incidences of racist views among white voters. This reporting has placed every white voter in the position of, no matter how carefully they research the opinions of the candidates, or their party affiliation, feeling open to the charge of racism by preferring the other candidate. Unfair? Of course it is .
The telephone rings and it is a pollster. Despite claims of privacy everyone realizes the pollsters know who you are, where you live, and can probably make an educated guess as to your race. What does the average person do? Self report the truth, or give the pollster the nice, warm and fuzzy answer?
The most famous example of this happening was the 1982 Governor’s contest in California. Tom Bradley, an African-American clearly ahead in the polls, ultimately lost the election.
UPDATE: Additional to the original post:
Self-report and poll data is so notorious for inaccuracy that many psychological tests, predominately of the pre-employment type, commonly add what is referred to as “lie” questions. In the above I mentioned the ten dollar bill found lying on the ground. The researchers knew from experience that while a significant portion of the sample would report that they would turn the money in, no one ever did. So they knew from observational experiments that a person who answered affirmatively to the question of returning the lost money was telling a lie. Many psychological tests have numerous “lie” questions like that example added to detect the potential employee who is presenting an overly unrealistic ‘honest’ image of himself. The professor in my statistical psychology class was involved in the creation of a number of these tests and warned the class of their presence.
This short little article in ‘Psychology Today’ gives an example:
Tags: Election 2008