Sunday, February 13, 2005
In a study conducted one week before the 2004 presidential election, a representative sample answered a series of survey questions about John Kerry and George Bush, including their vote intention. ...NOTE: I have edited the original post to include an author comment from the blog responses. The authors of the study are looking for comments.
In the new Bush/Kerry study, participants always saw photographs of Bush and Kerry side by side. We had three conditions.
1) Self is morphed with Bush, unfamiliar face is morphed with Kerry.
2) Self is morphed with Kerry, unfamiliar face is morphed with Bush.
3) No morph for either Bush or Kerry.
Results showed that respondents prefered Bush more in 1 than 3, and more in 3 than 2. Vice versa for Kerry.
As a further control, the 'self' morphed photos for one respondent was always used as an 'other' photo for another respondent, ensuring that attractiveness, etc. of the photographs was controlled completely across conditions 1 or 2. ...
The photographs of the respondents were acquired long before the study began; consequently none of the 200 respondents realized their own photographs had been morphed with the candidates.
The results demonstrated that respondents were significantly more likely to vote for the candidate with whom their face had been morphed (for both Bush and Kerry). This effect was stronger for people who did not have strong party affiliations (i.e., independent voters) than for strong partisans.
In summary, one week before the presidential election, respondents' vote choice was swayed by a simple morphed photograph. In politics, as in life, birds of a feather flock together.