Searching around on Google Scholar is something that I’m pretty adept at by now. This happens to be the 116th Substack post and almost all of them have involved some type of research. Trying to get a set of the best articles on mail polling methodology was actually really challenging [1]. I know that modern polling has moved away from mail polling in general. It’s expensive. The response rate has gotten worse over time. Probably the worst part about it is that compared to phone or internet based polling it is very slow. My interest here is to really understand polling methodologies. Getting to a high level of understanding about these methodologies will help me evaluate that current literature against automated sentiment analysis. It’s the combination of those two literatures that really has my attention at the moment.
Google Scholar was actually recommending a related search, “design method mail and internet surveys” [2]. I’ll admit that the results on that search were a little more interesting, but none of it was really targeted at what I wanted to learn about. I ended up broadening my searches to include content related to mail surveys. That yielded a few highly cited sources. One of the problems with this area of research compared to AI or ML is that all the research is not easily accessible for download.
Brown, M. L. (1965). Use of a postcard query in mail surveys. The Public Opinion Quarterly, 29(4), 635-637.
Sudman, S., Sirken, M. G., & Cowan, C. D. (1988). Sampling rare and elusive populations. Science, 240(4855), 991-996.
Dillman, D. A. (1991). The design and administration of mail surveys. Annual review of sociology, 17(1), 225-249.
Scott, C. (1961). Research on mail surveys. Journal of the Royal Statistical Society: Series A (General), 124(2), 143-195.
Kanuk, L., & Berenson, C. (1975). Mail surveys and response rates: A literature review. Journal of marketing research, 12(4), 440-453.
Maybe you were wondering what ChatGPT model 3.5 would come up with today. Using the prompt, “write a substack post called, ‘A literature study of mail polling methodology’”, a post was created.
A literature study of mail polling methodology
In an era where online and mobile polling have become more popular, it can be easy to overlook the traditional method of mail polling. However, mail polling remains a valuable tool for collecting data, particularly for certain populations. In this literature study, we will explore the latest research on mail polling methodology.
Firstly, studies have shown that mail polling can be particularly effective for reaching older populations, who may be less likely to use online or mobile polling methods. A study by the Pew Research Center found that mail polling had higher response rates among adults aged 65 and older than online or phone polling.
Secondly, mail polling can also be useful for collecting data on sensitive or controversial topics. A study by the National Opinion Research Center (NORC) found that mail polling resulted in higher rates of disclosure on topics such as drug use and sexual behavior, possibly because respondents feel more comfortable disclosing sensitive information in writing rather than in person or over the phone.
Thirdly, mail polling can be useful for collecting data from hard-to-reach populations, such as those without internet or phone access. A study by the American Association for Public Opinion Research (AAPOR) found that mail polling resulted in higher response rates among rural populations than online or phone polling.
Fourthly, researchers have explored the use of incentives to increase response rates in mail polling. Studies have shown that offering incentives such as cash or gift cards can lead to higher response rates, although the effectiveness of incentives may vary depending on the population being surveyed.
Finally, researchers have also explored the use of mixed-mode polling, which combines mail polling with other methods such as online or phone polling. This approach can help reach a wider range of respondents and increase response rates.
In conclusion, mail polling remains a valuable tool for collecting data, particularly for certain populations. While online and mobile polling methods have become more popular, mail polling should not be overlooked, particularly for collecting data on sensitive topics or hard-to-reach populations. The use of incentives and mixed-mode polling can also help increase response rates and improve the accuracy of results.
Links and thoughts:
Top 5 Tweets of the week:
It’s possible that these tweets will not show up as embeds and will just be links. This is due to a change in how Twitter is handling things and not a Substack issue.
https://twitter.com/sama/status/1644878154870427649
https://twitter.com/jdevalk/status/1644439739645194279
https://twitter.com/MKBHD/status/1644562346843467776
https://twitter.com/AndrewYNg/status/1644418518056861696
https://twitter.com/CaseyNewton/status/1644358645084459013
Footnotes:
[1] https://scholar.google.com/scholar?hl=en&as_sdt=0%2C6&q=mail+polling+methodology&oq=mail+polling+met
What’s next for The Lindahl Letter?
Week 117: A literature study of non-mail polling methodology
Week 118: A paper on political debt as a concept vs. technical debt
Week 119: All that bad data abounds
Week 120: That one with an obligatory AI trend’s post
Week 121: Considering an independent study applied AI syllabus
If you enjoyed this content, then please take a moment and share it with a friend. If you are new to The Lindahl Letter, then please consider subscribing. New editions arrive every Friday. Thank you and enjoy the year ahead.
A literature study of mail polling methodology