Question wording effects

A lot of the conclusions in our book come from surveys, so it’s good to be reminded that polls are only as good as their questions. Noah sent in this example from today’s NBC/WSJ poll:

20a. As you think about the presidential race and the direction in which the next president will take the country, who do you think would be the riskier choice for president–John McCain or Barack Obama? * +
John McCain ……………………………….. 35% [215]
Barack Obama ……………………………. 55
Both (VOL) ……………………………………. 3
Neither (VOL) ……………………………….. 1
Not sure ………………………………………… 6
* Asked of one-half the respondents (FORM A)
+ Results shown reflect responses among registered voters..

20b. As you think about the presidential race and the direction in
which the next president will take the country, who do you think would
be the safer choice for president –John McCain or Barack Obama? ** +
John McCain ……………………………….. 46% [216]
Barack Obama ……………………………. 41
Both (VOL)……………………………………… 1
Neither (VOL) ………………………………… 5
Not sure …………………………………………. 7
** Asked of one-half the respondents (FORM B)
+ Results shown reflect responses among registered voters.

People appear to be much more likely to say that Obama is “riskier” than that McCain is “safer.” As Noah points out, this seems related to all sorts of loss-aversion, uncertainty-aversion kinds of findings from psychology.

1 thought on “Question wording effects

  1. Pingback: Survey question wording effects « PoliSci@UST

Comments are closed.