Not surprisingly, when you ask about the President’s Job Approval in different ways, you get different results.

On Tuesday, November 24, 2009, Rasmussen Reports asked three separate samples of 800 Likely Voters a question about the President’s Job Approval. In all three cases, it was the first question asked on the survey. Survey, sampling, and weighting techniques were identical for all three. See toplines and question wording.

When we asked the question in the standard Rasmussen Reports format, we found that 47% of voters approved while 52% disapproved. Those figures include 28% who Strongly Approve and 41% who Strongly Disapprove.

However, we got different results when we simply asked if people Approved or Disapproved. When we took away the options of Strongly Approve/Disapprove and Somewhat Approve/Disapprove, the results were 50% approve and 46% disapprove. It’s important to note that the difference could be just statistical noise in a survey with a +/- 4 percentage point margin of sampling error.

(Want a free daily e-mail update? If it's in the news, it's in our polls). Rasmussen Reports updates are also available on Twitter or Facebook.

However, the difference is consistent with years of observations that Rasmussen Reports polling consistently shows a higher level of disapproval for the President than other polls. That may reflect the fact that there may be some people willing to offer a “somewhat disapprove” rating rather than say they “disapprove.”

Some of the gap on the negative side is probably the result of using an automated polling system (see methodology). Automated polls almost always record a higher negative rating for all politicians—regardless of party or ideology--than operator-assisted polls. One theory is that automated polls pick up a more honest reaction on the negative side because people may be a bit reluctant to tell an operator that they don’t like someone (especially someone they’ve never met).

A third way of asking how the President is doing is to ask whether he is doing an Excellent, Good, Fair, or Poor Job. Using this scale, 27% said he’s doing an Excellent job, 11% said good, 14% fair, and 46% poor. This is consistent with years of observations that the President’s Excellent and Good total will be lower than any reported “Approval” figure. The reason appears to be that some people who rate the President’s performance as “Fair” would also be willing to say they “approve.” It’s also likely that some who say “fair” would either disapprove, slightly disapprove, or say they’re not sure.

With three different questions yielding three different results, it’s important to focus on what can be learned from Job Approval surveys.

First, the most important thing is the trend. Whether the President’s Job Approval is 47% or 50% matters less than whether it’s up or down significantly over some period of time.

The only time the absolute number seems to matter is when an incumbent President is running for reelection. Typically, in that case, the President’s Job Approval among Likely Voters will be close to his vote total on Election Day. In Election 2004, the Rasmussen Reports Job Approval rating for President George W. Bush was just about identical to his share of the vote.

Second, when comparing polls, it’s always most instructive to look for the common ground. One example from the current data is that 41% Strongly Disapprove using the first measure, 46% Disapprove using the second, and 46% rate the President’s performance as poor.

Third, it’s important to be mindful of the differences when comparing numbers from different polling firms. One difference obviously can be question wording. There are many others. Rasmussen Reports typically reports its results from a sample of likely voters while other firms often show results from all adults. Generally speaking, a survey of all adults will yield a number 3-5 points higher for President Obama than a poll of likely voters. The reason for this is that some segments of the population that think highly of the President are less likely to vote (i.e.—voters under 30).

Finally, if question wording can impact something as seemingly straightforward as the President’s Job Approval rating, caution is advised when comparing polls on other topics. One recent example was found in the health care debate where a pundit compared two polls and suggested they showed very different results. However, a closer look revealed that the problem was with the analysis, not the polls. Both the Rasmussen Reports poll and the Kaiser Foundation poll reflected similar attitudes on health care reform overall and on the legislation currently before Congress.

Please sign up for the Rasmussen Reports daily e-mail update (it’s free) or follow us on Twitter or Facebook. Let us keep you up to date with the latest public opinion news.

See question wording and toplines.