Snap judgments about candidates are the best way to pick winners, study suggests

Candidate A

Candidate B

The study participants viewed these same two clips (among many others), and answered these same three questions. Their answers were distributed as shown below.

Question 1
One of these candidates is a Democrat, and one is a Republican. Which one do you think is the Democrat?

Study participants' responses:

• Candidate A: 54%
• Candidate B: 46%
• Correct Answer: Candidate B

Question 2
Who would you vote for in an election in your home state?

Study participants' responses:

• Candidate A: 61%
• Candidate B: 39%

Question 3
Who do you think actually won this election for governor?

Study participants' responses:

• Candidate A: 61%
• Candidate B: 39%
• Correct Answer: Candidate A

Note that even though most participants incorrectly believed that Candidate B was the Republican, most participants nevertheless correctly predicted that Candidate A won. This was representative of the study as a whole, in that respondents generally did no better than chance in guessing which candidate belonged to which party, but did significantly better than chance in ascertaining who won the election. This set of clips was unususal in that exactly the same number of people who would have voted for the actual winner themselves chose the correct candidate as the winner; in the larger study, those numbers were correlated but usually not identical.

These particular clips were drawn from the 2002 race for Governor in South Carolina, in which Republican challenger Marshall Clement Sanford, Jr. (Candidate A) defeated Democratic incumbent James Hodges (Candidate B), receiving 53% of the vote to Hodges' 47%.

Return to press release