I will be extremely unhappy with the results of this election, no matter who wins.
But one thing I will be extremely happy about is that the election will finally be over. Finally, after what seems like years.
Short of a 2000-type brouhaha, that is.
I’ve noticed recently that things have heated up, both on the internet and in the non-cyber-world, during this past week especially. Everyone seems testy and on edge. My liberal friends in particular are terrified that Hillary might lose, because even if they don’t like her (and many don’t like her), they feel Trump is a very dangerous man. And on this blog and others, sniping in the comments section between people on the right has increased even more than before.
There’s a general weariness that is almost palpable. I feel so weary of discussing the pros and cons of Trump vs. Hillary that I could scream.
And as I wrote that last sentence, it occurred to me that “the cons and cons” would be a much better phrase to use.
So forgive me—or thank me, as you wish—if I don’t post my near-daily Trump/Hillary discussion today. The next week promises to be a lulu, and that’s about all I’ll say right now on that score.
I also find myself spending more time lately in the comments section here, responding to various people. Often the research I do for those comments requires as much time as the research I do for a main post, and yet far fewer people see a comment than see a post.
So now I’ll take the opportunity to offer the gist of one of those previous comments in order to give you more information I found on the topic of “oversampling” (as discussed previously vis a vis the Podesta emails). Hope this will clarify things even further.
“Oversampling” is a technique that is sometimes used in certain types of polls and research in general to study small groups. Here is an explanation of what it actually means.
Oversampling is the practice of selecting respondents so that some groups make up a larger share of the survey sample than they do in the population. Oversampling small groups can be difficult and costly, but it allows polls to shed light on groups that would otherwise be too small to report on.
This might sound like it would make the survey unrepresentative, but pollsters correct this through weighting. With weighting, groups that were oversampled are brought back in line with their actual share of the population ”“ removing the potential for bias…
. When we are interested in learning about groups that make up only a small share of the population, the usual approach can leave us with too few people in each group to produce reliable estimates. When we want to look closely at small groups, we have to design the sample differently so that we have enough respondents in each group to analyze. We do this by giving members of the small group a higher chance of being selected than everybody else.
A good example is a Pew Research Center survey from June of this year, in which we wanted to focus in depth on the U.S. Hispanic population. In the previous survey from March, there were 291 Hispanic respondents out of 2,254 total respondents, or 13% of the sample before weighting. This is pretty close to the true Hispanic share of the population (15%), but we wanted to have more than 291 people responding so we could do a more in-depth analysis. In order to have a larger sample of Hispanics in June, we surveyed 543 Hispanics out of 2,245 total respondents, or 24% of the unweighted sample. This gave us a much larger sample to analyze, and made the estimates for Hispanics more precise.
If we just stopped here, estimates for the total population would overrepresent Hispanics. Instead, we weight them back down so that when we look at the whole sample, the share of Hispanics falls back in line with their actual share of the population. This way, we still have more precise estimates when looking at Hispanics specifically, but we also have the correct distribution when looking at the sample as a whole.
That’s an explanation from Pew Research of what the technique is and how it’s used in polling.
So it does not mean biased sampling or skewed polls in an attempt to fool anyone. Also, these polls Podesta was talking about were not polls that were released to the public. They were internal polls that had the purpose of studying certain small subgroups in certain areas, so that the campaign could learn more about them.
It’s possible to think, because that article by Pew that I just quoted was written after the Podesta emails came out, that it was just some sort of ex-post-facto made-up excuse. So I refer you to articles about oversampling in small populations that were written significantly before this election cycle, and therefore are not responses to Podesta and aren’t trying to make any political points at all.
Here’s an article about using oversampling in small groups, written quite a while ago (the comments there date from 2015, but according to my Google search the article was written in 2012).
Here’s some information about oversampling in small populations that was written in 2010.
Here’s an example of some type of study from 2011 that used oversampling (it’s not about politics).
There are other examples of research that uses oversampling as a technique to study groups that are a small fraction of a population. This information is easy to obtain, and has been in the public domain for a long time.
Once you are aware of what the technique of oversampling of minority populations is, you understand the Podesta emails better. His emails fit into the framework of what is being described here, and do not appear to have anything to do with faking data or deceiving anyone.
As I’ve stated before, that doesn’t mean that there aren’t attempts to deceive, either with polling or in other ways. It merely indicates that there’s no “there” there in these particular charges about Podesta and the use of oversampling.