Poll Administration

In class today we discussed administering polls. In my poll design for class I decided to use self-administered polls through the mail. This is the cheapest way of administering a poll. It also provides participants with the opportunity to take their time, read through the poll before answering the questions (which decreases the negative aspects of question order), and they can answer the questions in their own privacy which encourages and supports confidentiality. Despite these positive areas in self-administered polling, there are some drawbacks. The number one issue is response rate. The response rate for self-administered polls received through the mail is quite low which drastically affects the poll and can even prevent the poll from being publicly announced. In an attempt to receive a higher response rate with this poll, I am designing the poll to be administered to more people than is needed for a representative sample. By doing this, the chance of receiving enough completed polls of a representative sample increases. Another drawback in a self-administered poll in regards to my poll design is that there may not be an up-to-date list of all of the U.S. population, which is where I want my representative sample to come from. However, I feel that this problem is not as big of an issue as it would be with telephone polling. This is because I think there is more of a chance to reach people today through the mail versus through a landline since so many people today rely solely on cellular phones. I am not saying that telephone polling does not work, when in fact it does. It is just that since the results of my poll are not rushed seeing as though the issue of embryonic stem cell research is not salient at the moment, a mailed self-administered poll may work better.

Issues with Truthful Respondents in Polls

Yesterday in class, we discussed two important concepts that relate to "nonattitudes" in public opinion polling: social desirability response bias and insincerity. Social desirability response bias refers to the idea that people actually have an opinion on the question they are being asked, but they lie about their answer in giving one that is socially desirable. Insincerity is similar in which people have an opinion, yet for various reasons, they report one that is different from their own whether it is socially desirable or not. This can definitely cause issues in polling by providing inaccurate results to the public. I have been looking at various polls on my issue and it seems that they all provide similar results.

One thing that may show a social desirability response bias that I found is when a certain organization is administering a poll. For instance, it seems that those who are Catholics are opposed to embryonic stem cell research. A poll was sponsored by the U.S. Conference of Catholic Bishops and it showed that more people were in opposition to federal funding of embryonic stem cell research than those who favor it. Based on other polls administered by credible polling organizations such as Gallup and CBS news, the results are opposite. While the website I found does not show the poll, it does show that perhaps people were told who was sponsoring the poll. The point here is that if people know a Catholic group is administering the poll, it is possible that a social desirability response bias and/or insincerities may result. The poll itself claims it was representative and what changed the results was the option of human stem cell research versus embryonic stem cell research. However, the Gallup poll that I posted about earlier does in fact ask the question using the term "embryonic" and not just stem cells. This is the only information regarding different polling results that I have found so far.

As for my poll design, I will focus on a national representative sample, as did other credible organizations. I will not worry about social desirability response bias or insincerities other than the issue that President Obama supports embryonic stem cell research. In order to try to minimize any opinions based on supporting the president, I will make sure that my questions will not provide any information that shows political support one way or the other. If anyone is interested in reading the article about the Catholic sponsored poll, the link is below:

http://www.lifenews.com/bio434.html

Gallup poll:

http://www.gallup.com/poll/116485/Majority-Americans-Likely-Support-Stem-Cell-Decision.aspx

Are There "Nonattitudes" Reflected in Embryonic Stem Cell Research?

Yesterday my public opinions class discussed "nonattitudes" in public opinion polls. This concept means that people may not have an opinion about an issue for various reasons but respond in polls as if they do. While designing my poll, I need to consider whether I believe "nonattitudes" may be a problem and if so, what to do about it. It doesn't seem like "nonattitudes" would be an issue in polls regarding embryonic stem cell research mainly since opinions have been stable throughout the years and although this issue isn't really debated in the news at the moment, it is still a popular controversy. One issue I think may be a problem is that some people don't really understand what exactly stem cell research is and why there is such a debate. I have spoken to a few people on the issue and some admitted that they really don't know what the research entails yet they would still give an opinion if asked whether or not they support it. This seems to be the only problem I may have to think about regarding "nonattitudes" as I design my poll. A screening question I can use would be whether or not you have an opinion about the topic and maybe ask why they are in favor or in opposition and then continue with the survey if they have an opinion. I haven't really come across a poll on this issue that seems to reflect "nonattitudes." If I decide I need to ask this question to weed out those who don't really have an opinion it is something that will definitely be worth while. I will also add the option of "don't know" or "no opinion" in the questions that need that as an option because I wouldn't want anyone to feel forced to make a decision regarding a question they really don't have an opinion on which would result in results that are not accurate. As I continue to brainstorm and write out my poll design I will post more information regarding the project.

Sampling Methods

As we begin to construct our poll designs, my class discussed sampling methods that should and shouldn't be used. For instance, convenient sampling is not scientific and results are likely to be skewed. As I am contemplating on which method I will use in my poll design I wanted to provide a link for readers that shows an example of how to get the results poll designers are looking for. The clip shows how to get a "balanced" sample using questions that will lead to the ultimate question in which the poll is looking to report on:

http://www.youtube.com/watch?v=2yhN1IDLQjo

Poll Displaying Demographics

Hello everyone. I found a poll that shows demographic factors on the issue of embryonic stem cell research. Although the poll is from 2005, it is from a credible source. I thought it was worth posting in order to help readers gain an understanding of what groups of people oppose the research compared to the national population. While 58% approve of the research according to this poll (which is the same results as the Gallop Organization from a previous post), this poll shows which groups of Americans are in opposition. The groups are conservatives, white Evangelicalists, and those who attend services weekly. All 3 groups have a higher disapproval rate than approval by 45%-53%. The poll consisted of 938 participants and was a random sample from the U.S. population:

http://www.cbsnews.com/stories/2005/05/24/opinion/polls/main697546.shtml

"Pseudopolls"

Hello everyone! In class we discussed the issue of "pseudopolls" which are fake or unscientific polls. These types of polls lead to bias results which pave the way to public skepticism. Those who participate are not part of a random sample and they do not represent public opinion. These are polls that can be taken online or even by text messaging the answer to a question to the people who are reporting on the poll and the results. The key here is that people are choosing to participate and chances are they are responding to a poll posted by a group who holds their views.
I found such a poll on embryonic stem cell research. The poll is posted on a website entitled "WorldNetDaily." There a few obvious issues with this poll. Not only is it an online poll in which anyone can vote (and you must sign up for an account in order to vote), but there are 22 answers to choose from for 1 question ("What's your view on embryonic stem cell research?") in which participants can cast their vote. The website does however only allow participants to vote 1 time but it does not let people know that this is an unscientific poll. Also, once a person has voted, s/he can view the results of the poll. I casted a vote and viewed the results which displays the most bias results I have yet to see in any poll relating to this issue. This is obviously a website for those in oppostition to embryonic stem cell research because about 99% of the votes are answers that in some form represent opposition to the issue at hand.
To compare these results with the poll I posted earlier from the Gallop Organization, which is a credible polling organization, less than 50% were opposed to embryonic stem cell research. This just proves the point that this poll is a "pseudopoll." Also, if you are bored one day, you can access polls from their website and start voting on many issues. It is quite humorous so if you feel like laughing, go to the website and scroll down below the article to begin voting, but remember you must create an account in order to cast your vote:

http://www.wnd.com/index.php?fa=PAGE.view&pageId=86957