Tweet
Don't Bet on Obama's Likeability Rating
American Thinker
Frank Burke
10/10/2012
Excerpt:
"Despite declining approval numbers, pundits are opining that Obama will be re-elected because of his personal likeability rating. Lacking a background in market research, they are taking at face value a question with much more subtle implications.
In legitimate polling, three major factors include the sample, the questions, and the methodology (manner in which the poll is administered). To provide valid results, the sample must be truly representative of the target audience. For example, a purported "national sample" that skews heavily toward Democratic or Republican voters does not meet this criterion. Also, a sample drawn from the general population for the purposes of predicting the outcome of a national election will not be as effective as one drawn from a sample of likely voters. Another consideration is the size of the sample. If properly constructed, it follows that the larger the sample size, the more valid the results will be.
The second element, the questionnaire, must be carefully composed so as to avoid presenting the questions in a manner likely to prejudice the response. For instance, the question, "Despite the dangers involved, do you favor the early release of drugs under consideration by the FDA in critical cases?" is much more a leading question than "In critical cases, do you favor the early release of drugs under consideration by the FDA?" Even with the best intentions, experienced researchers can pose questions in such a way that they are prejudicial, misleading, or poorly understood. Because of this, ethical researchers will pre-test questionnaires on a control group to ensure that the questions are effective.
A third critical element involves the methodology. Faulty methodology can doom a poll to failure before it is even administered. The 1936 presidential poll conducted by The Literary Digest -- a then-well-regarded magazine -- predicted a landslide victory for Republican Alf Landon, when in fact the election went to Franklin D. Roosevelt. On re-examination, it was discovered that the poll was conducted by telephone, and in 1936, most telephones were located in upscale households that tended to favor Republicans. Similarly, today, mail questionnaires are not considered effective, as upscale, professional people tend to disregard them as "junk mail." E-mail questionnaires are often similarly regarded. Even telephone polling has become difficult in an age where many individuals are abandoning land-line phones, which are traceable by zip code, in favor of cell phones that might carry an area code from three moves ago.
Nonetheless, the professional pollsters rightly favor telephone questionnaires as the fastest, most efficient, and generally most reliable means of obtaining data in a timely manner. As any professional pollster will admit, however, this is not without its problems. A one-on-one conversation, even with an anonymous poll-taker, results in a relationship factor. Even if the questioner's tone, emphasis, and verbal nuance are as neutral as possible, other factors enter into the equation. Depending on the group and the subject, these can involve a desire to please the questioner, to impress the questioner, or to avoid saying "the wrong thing." This is especially true when it comes to questions of a personal nature. For instance, surveys taken among adolescent boys regarding their drinking habits are likely to overstate alcohol use. A 16-year-old wishing to appear more mature or macho is liable to claim more frequent indulgence and greater consumption. Likewise, when young people are questioned regarding their sexual practices, they have been known to exaggerate.
Based on my experience, this need to meet expectations and say the right thing explains the conundrum that is the Obama Likeability Factor. Questions about the economy, the direction of the country, national debt, and other general topics are fairly neutral in nature, and the respondent's answers may reflect a political position but not an emotional one. When asked, however, whether he or she finds the president likeable, the survey crosses the line into a highly personal area. Given the extent to which this administration has played the race card, it is possible that many respondents provide a positive answer for fear of being thought racist.
......................................
View the complete article at:
http://www.americanthinker.com/2012/...ty_rating.html
American Thinker
Frank Burke
10/10/2012
Excerpt:
"Despite declining approval numbers, pundits are opining that Obama will be re-elected because of his personal likeability rating. Lacking a background in market research, they are taking at face value a question with much more subtle implications.
In legitimate polling, three major factors include the sample, the questions, and the methodology (manner in which the poll is administered). To provide valid results, the sample must be truly representative of the target audience. For example, a purported "national sample" that skews heavily toward Democratic or Republican voters does not meet this criterion. Also, a sample drawn from the general population for the purposes of predicting the outcome of a national election will not be as effective as one drawn from a sample of likely voters. Another consideration is the size of the sample. If properly constructed, it follows that the larger the sample size, the more valid the results will be.
The second element, the questionnaire, must be carefully composed so as to avoid presenting the questions in a manner likely to prejudice the response. For instance, the question, "Despite the dangers involved, do you favor the early release of drugs under consideration by the FDA in critical cases?" is much more a leading question than "In critical cases, do you favor the early release of drugs under consideration by the FDA?" Even with the best intentions, experienced researchers can pose questions in such a way that they are prejudicial, misleading, or poorly understood. Because of this, ethical researchers will pre-test questionnaires on a control group to ensure that the questions are effective.
A third critical element involves the methodology. Faulty methodology can doom a poll to failure before it is even administered. The 1936 presidential poll conducted by The Literary Digest -- a then-well-regarded magazine -- predicted a landslide victory for Republican Alf Landon, when in fact the election went to Franklin D. Roosevelt. On re-examination, it was discovered that the poll was conducted by telephone, and in 1936, most telephones were located in upscale households that tended to favor Republicans. Similarly, today, mail questionnaires are not considered effective, as upscale, professional people tend to disregard them as "junk mail." E-mail questionnaires are often similarly regarded. Even telephone polling has become difficult in an age where many individuals are abandoning land-line phones, which are traceable by zip code, in favor of cell phones that might carry an area code from three moves ago.
Nonetheless, the professional pollsters rightly favor telephone questionnaires as the fastest, most efficient, and generally most reliable means of obtaining data in a timely manner. As any professional pollster will admit, however, this is not without its problems. A one-on-one conversation, even with an anonymous poll-taker, results in a relationship factor. Even if the questioner's tone, emphasis, and verbal nuance are as neutral as possible, other factors enter into the equation. Depending on the group and the subject, these can involve a desire to please the questioner, to impress the questioner, or to avoid saying "the wrong thing." This is especially true when it comes to questions of a personal nature. For instance, surveys taken among adolescent boys regarding their drinking habits are likely to overstate alcohol use. A 16-year-old wishing to appear more mature or macho is liable to claim more frequent indulgence and greater consumption. Likewise, when young people are questioned regarding their sexual practices, they have been known to exaggerate.
Based on my experience, this need to meet expectations and say the right thing explains the conundrum that is the Obama Likeability Factor. Questions about the economy, the direction of the country, national debt, and other general topics are fairly neutral in nature, and the respondent's answers may reflect a political position but not an emotional one. When asked, however, whether he or she finds the president likeable, the survey crosses the line into a highly personal area. Given the extent to which this administration has played the race card, it is possible that many respondents provide a positive answer for fear of being thought racist.
......................................
View the complete article at:
http://www.americanthinker.com/2012/...ty_rating.html