57 percent or sloppy polling?

Posted in Uncategorized by Ryan Locke on April 6, 2009

Politico reported today that 57 percent of poll respondents support a military response to eliminate North Korea’s nuclear capabilities.  Does this really reflect what Americans think?

Probably not.  Look at the actual Rasmussen poll here.  This is the actual question:

3* If North Korea launches a long-range missile, should the United States take military action to eliminate North Korea’s ability to launch missiles?

57% Yes
15% No
28% Not sure

First, does North Korea really have a long-range missile capability?  The Wall Street Journal says no:

The three-stage rocket failed during its second stage and plunged into the Pacific Ocean about 1,300 kilometers east of Japan, according to analysts in the U.S., Japan and South Korea. On Monday, Russia, a country more friendly to North Korea than the other three, also announced that the missile appeared to fail before reaching its third stage, according to news reports.

However, this poll was conducted two days before the rocket failed.  Back then, it was kinda scary that North Korea was about to launch a missile; now, we realize that they can’t launch a long-range missle.  Would this poll question come up differently if it was polled two days after the rocket failed?  Maybe.

Second, here’s the question that immediately preceded the military action question:

2* How concerned are you about the possible threat of North Korea using nuclear weapons against the United States?

39% Very concerned
34% Somewhat concerned
20% Not very concerned
5% Not at all concerned
2% Not sure

Right after answering a question about North Korea using a nuke against the US (which experts believe could only happen via airplane, not missile), the respondents answered a question about North Korea’s missile capability.  This is called “priming” — where an early stimulus influences response to a later stimulus.  The Pew Research Center explains this phenomenon in their FAQ, explaining why they ask presidential approval first in their surveys:

For example, if the survey first asks about the economy and then asks about presidential approval, the respondent may still be thinking about the economy when answering the latter question. While economic conditions may be important in assessing the president’s overall performance, so are many other issues. If the respondent is only thinking about the economy because we brought up the issue, his or her response about the president may be biased by what we call a context effect: in this case we would be priming the respondent to consider the economy in an assessment of the President.

This poll isn’t an accurate reflection of what Americans think–or if it actually is, we can’t be sure because of sloppy polling methodology.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: