Kaukab Jhumra Smith, a correspondent with Youth Today, was curious why data about candidate support referenced in two different posts on the Youth Engagement Poll seemed not to be consistent. This question allows us to clarify a bit how we reported support for Obama and Romney. The two posts Kaukab referenced were:

Romney Trails Among Young Adults — posted July 31, 2012

68% of Currently-Undecided Youth Voted for Obama in 2008, 28% for McCain; Romney Supporters Less Likely to be Unemployed, while 86% of Obama supporters “admiring” or “satisfied” with him — posted August 22, 2012

The two posts were actually referencing two different variable configurations. As one will note from the toplines, a question asked:

“If the election for President were being held today, and you had to make a choice, what is the likelihood that you will be voting for these candidates?”

We then provided data for Barack Obama, Mitt Romney and Someone Else. Actually, the results were derived from three separate questions asked of respondents. So, for instance, respondents were asked, “If the election for President were being held today, and you had to make a choice, what is the likelihood that you will be voting for President Barack Obama?” 33.8% said “would definitely vote for”, 21.1% said “might vote for”, etc. Respondents then were asked this question replacing President Barack Obama with Governor Mitt Romney and were finally asked a third question replacing the option with “Someone Else.”  Given this configuration of the question, it is very likely that some respondents would overlap in their answers, and indeed some respondents said they “would definitely” vote for Barack Obama and Mitt Romney.

So the post referenced here – should have said:

When asked if they would vote for Barack Obama if the election were held today, 34% of respondents said they would “definitely” vote for the President. When asked the same question about Mitt Romney, only 16.1% of respondents said they “definitely” would vote for the ex-Governor.

As for this post –We actually calculated a new variable to determine who was an Obama supporter, who was a Romney supporter, and who truly was undecided. We took the above referenced question about voting in the presidential election today and said, if someone “definitely” or “might” vote for Obama AND “definitely” would NOT vote for Romney, they were an “Obama” supporter.   And similarly, if they “definitely” or “might” vote for Romney AND “definitely” NOT vote for Obama they were a Romney supporter. We did the same thing for alternative / someone else supporters and everyone left over we determined were truly “undecided”.

So in the text in this post read:

38.4% of young Americans 18-29 years of age indicated they would “definitely” or “might” vote for President Obama if the election were held at that time. This was a thirteen percentage point lead over those who indicated they would “definitely” or “might” vote for Governor Romney (25.4%).

It has been corrected to more accurately read:

38.4% of young Americans 18-29 years of age indicated they would “definitely” or “might” vote for President Obama and “definitely” would not vote for Romney if the election were held at that time. This was a thirteen percentage point lead over those who indicated they would “definitely” or “might” vote for Governor Romney (25.4%) and “definitely” would not vote for Obama.

This entry was posted on Wednesday, August 29th, 2012 at 9:58 am and is filed under CIRCLE Blog, Data Sets. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.