http://fivethirtyeight.com/features/pollsters-say-they-follow-ethical-standards-but-they-arent-so-sure-about-their-peers/
With elections coming up in November, we like to have a sneak peek of the results, which is done through polls. There are multiple poll companies throughout the United States, and the staff of FiveThirtyEight decided to take a poll of one pollster from 24 different companies. Basically they asked the pollsters questions about their ethical habits (like if they copied other companies' data, let campaigns and partisan bias determine weighting, or if they thought recalling a poll was alright). To summarize the results, all of the pollsters that answered said they do follow ethical standards when polling. They even gave examples of how they make sure to publish accurate data, like going back into the field to double check poor data. On the other hand, many of the pollsters believed that other polling companies did copy data, and were susceptible to partisan motives. Obviously the pollsters might be biased when answering these questions because who would want to admit to cheating? However, we have to consider that the poll was anonymous (they published who was in the study, but not what each person individually answered).
I think that overall the data from FiveThirtyEight's poll is probably accurate. Those who answered the questions most likely told the truth, and it is more likely that those who chose not to respond to certain questions were the ones guilty of unethical polling. I think overall most polling companies do their best to collect accurate data, although this doesn't mean they all practice ethical polling standards. Besides, the pollsters are paid to collect accurate information. Their job is doing the polls right. I liked how the author of the article, Carl Bialik, presented the data from the pollsters. The article included individual answers to questions and many possible reasons for bias in answers. FiveThirtyEight obviously understands polls, since they base their political predictions on them. Do you trust the information polling companies are publishing, or do you think there is some unethical polling going on?
Unit 3
While they may be getting accurate information, who is to say they are actually getting ALL of the information? They could be leaving out key facts that completely change someones interpretation.
ReplyDeleteHehe...someone has some trust issues. Kinda like me...oh, wait...darn.
ReplyDeleteWhat an interesting study. I agree with Allie; polling companies might only be publishing half-truths. However, it would seem that individuals would be much more likely to practice ethical habits than the polling companies as a whole, so maybe the individuals interviewed were giving an honest assessment. I dunno.
ReplyDeleteWe talked about this in statistics today!:) Of course everyone believes that they are acting ethically-if someone believed that they themselves were unethical, they would stop doing what they are doing, Did the study define "ethical" or did they leave it up to personal moral standards? Because that could cause some serious validity issues.
ReplyDeleteGood point, Frankie. Morals and values vary so much from person to person (and company to company...). Perhaps this study is simply not specific enough when it comes down to what "ethical behavior" actually looks like.
DeleteThis is really interesting. I think that it's up to people as individuals to decide what is ethical, BUT then to make sure they act in such a way that it doesn't violate the ethics of those they associate with. In other words... there is no way to monitor, label, or be specific in determining the definition of 'ethical'.
DeleteOf course everyone thinks they are doing it right, but don't trust the other guys.
ReplyDeletePolling is not and has never been an exact science. I image most polls are flawed. What's interesting though is when the same poll is conducted in the same manner over time. Then the polls are flawed in the same way, so the statistical changes actually mean something.