The issue: Pollsters often use a polling technique called “sampling”, a process that involves selecting a few respondents from a larger population. This theoretically should yield a fairly accurate representation of the opinions of the larger group.
Major Strength: Some people are in favor of sampling. Some are opposed to it. And while it is clear that Asher approves of sampling, he fairly presents the pros and cons of both sides of the dispute. In doing so he is able to introduce, defend and then disprove the “anti-sampling” mentality, thus strengthening his argument. By demonstrating knowledge of the opposite opinion, it heightens Asher’s credibility.
Major Weakness: Some studies hold up over time, but those centered on technology tend to become outdated pretty quickly. Looking at some of the dates that Asher uses to back his “telephone sampling” section, I’m seeing numbers like 1987 and 1996; and we know that technologically we are in a completely different place in 2014. In 2013 the Wall Street Journal reported that 71% of Americans still had a landline. While that number is relatively high, pollsters that sample are going to have a tough time painting an accurate picture when 3 out of every 10 people are “mobility unlisted.” With that in mind, I found it to be a bit of a stretch to buy into this section.
Underlying Assumption: As technology evolves, polling strategies, including sampling, must evolve too. Americans are skeptical to sampling, and they probably always will be. Sampling isn’t prefect, but that’s why the results are presented with a sampling error. And it’s the best system there is right now.
Provocative Questions: 1) Telephone sampling is still a very effective way to poll, but how much lower will that 71% landline rate drop in the next few years? 2) When someone figures out the best alternative to sampling, can that person call Nielsen? I’m tired of my favorite TV shows getting canceled.