Iran or North Korea More of a Threat? Part II

Two weeks ago, I discussed the findings from a recent Gallup survey that asked about the threat of Iran and North Korea to the United States. In half the sample (Form A), Gallup asked first about Iran and then North Korea; and in the other half of the sample (Form B), Gallup asked about North Korea and then Iran. The results show that the order in which these countries were mentioned affected people's responses.

When Iran was mentioned first, fewer people (16%) said it was an "immediate threat" than when Iran was mentioned after North Korea (30%). That was also true of North Korea -- when it was mentioned first, fewer people said it was an immediate threat (16%) than when it was mentioned after Iran (23%).

               

When each country is mentioned first, we classify that as a "non-comparative" context. That means people base their responses on whatever information they already have in their minds when we ask the question. They are not explicitly thinking about either country in comparison with the other.

But when each country is mentioned after the other, we call that a "comparative" context, because now the respondent has been reminded of the possible danger of the other country. With that in mind, the respondent will almost inevitably, either consciously or subconsciously, compare the potential threat of the two countries.

There are two noteworthy differences in the results of the comparative context and non-comparative context. First, more people see each country as a threat after another threatening country has been mentioned than when the country is mentioned first. For Iran, the difference is +14 percentage points (from 16% to 30%), while for North Korea, the difference is +7 percentage points (from 16% to 23%).

What does this mean? It suggests that the more people think about the issue, the more concerned they become about the possible threat of the country. If I am asked just about one country, I may not be especially concerned, but after you remind me about another -- and perhaps another -- suddenly I realize there is a lot more danger out there than I had originally thought.

This "priming" of respondents, reminding them of other dangers and then asking them about yet another one, is inevitable in polling whenever we are interested in asking about more than one item. Yet, the very inclusion of multiple items means that we can inadvertently induce the respondents to express greater concern than they would have if they had not been "primed" in this way.

Are the "primed" responses (in the comparative context) more valid, or less valid, than the "unprimed" responses (in the non-comparative context)? Answer: They are both valid -- but they give different insights into public opinion.

The second noteworthy difference between the two contexts is that in the comparative context, more people cite Iran as an immediate threat (30%) than North Korea (23%). In the non-comparative context, each country was viewed equally (16% said each country was an immediate threat).

How do we interpret these results? They suggest that when people explicitly compare Iran and North Korea (after both countries have been mentioned), more people see Iran than North Korea as an immediate threat. This view seems consistent with recent news reports, which have focused much more on Iran's growing nuclear capability than on North Korea's.

Bottom line: The fact that we get different results depending on which country is mentioned first does not invalidate polling, but rather shows that opinion on these matters is complex. This split rotation of Iran and North Korea provides information about two underlying dynamics of public opinion. One is that as people become "primed" about an issue, their concern about each country's potential threat increases. This also suggests that as public discussion and media attention on this issue increase, the general public becomes more concerned. The second underlying dynamic is that when people see the two countries in a comparative context, they are more concerned about Iran than North Korea, even though an initial reaction (but not explicitly comparative) shows equal concern about the two countries.

These results are also a reminder that when polling organizations report different results about the same issue, the differences are more likely to reflect the complexities of public opinion than the comparative accuracy of different polls.