3 ms·

Each month, the university conducts a minimum of 500 phone interviews across the continental U.S. Phone interviews are laughably inaccurate in the present day

Each month, the university conducts a minimum of 500 phone interviews across the continental U.S. Phone interviews are laughably inaccurate in the present day and a count of 500 is totally inadequate to capture what the entire US thinks about the economy. The left keeps trying to tell people that the economy is fine and that everyone is actually very happy about sky-high rents, unending layoffs, and astronomical grocery prices as if we can't see what's going on with our own bank accounts. This strategy of "don't believe your own eyes or ears" is straight out of 1984. Even worse, they use phone area codes to correspond respondents to area[0]. This survey is stupid. [0]https://data.sca.isr.umich.edu/fetchdoc.php?docid=57449 https://data.sca.isr.umich.edu/fetchdoc.php?docid=57449

bryanlarsen 3y ago
The issues are mostly systemic, not random. The numbers from one survey aren't particularly meaningful, differences between surveys are meaningful.
glimshe 3y ago
Don't forget medicine shortages
belltaco 3y ago
> unending layoffs Sounds like a frog in the well mentality, big tech is a very small part of the economy, even if you add companies like CitiGroup it's still tiny compared to everythign else where hiring is still strong. A good chunk of the laid off folks have savings from high salaries so they will keep on spending like normal so even those don't affect the larger economy too much.
- bryanlarsen 3y ago
  Yah, I still see help wanted signs in every restaurant window. OTOH, we're hiring and we're getting some fabulous resumes. win-win from my perspective, but if I was the one applying rather than interviewing, I know I'd have a different perspective.
mrala 3y ago
With random sampling, a sample size of approximately 400 provides a ±5% margin of error for a population of 331,000,000. A ±3% margin can be reached with about 1,000 respondents. Here are a few links that explain why this works: https://www.pewresearch.org/short-reads/2017/05/12/methods-101-random-sampling/ https://www.pewresearch.org/short-reads/2017/05/12/methods-1... https://www.surveymonkey.com/curiosity/how-many-people-do-i-need-to-take-my-survey/ https://www.surveymonkey.com/curiosity/how-many-people-do-i-... https://www.cloudresearch.com/resources/guides/statistical-significance/determine-sample-size/ https://www.cloudresearch.com/resources/guides/statistical-s...

bryanlarsen 3y ago

The issues are mostly systemic, not random. The numbers from one survey aren't particularly meaningful, differences between surveys are meaningful.

glimshe 3y ago

Don't forget medicine shortages

belltaco 3y ago

> unending layoffs Sounds like a frog in the well mentality, big tech is a very small part of the economy, even if you add companies like CitiGroup it's still tiny compared to everythign else where hiring is still strong. A good chunk of the laid off folks have savings from high salaries so they will keep on spending like normal so even those don't affect the larger economy too much.

bryanlarsen 3y ago
Yah, I still see help wanted signs in every restaurant window. OTOH, we're hiring and we're getting some fabulous resumes. win-win from my perspective, but if I was the one applying rather than interviewing, I know I'd have a different perspective.

mrala 3y ago

With random sampling, a sample size of approximately 400 provides a ±5% margin of error for a population of 331,000,000. A ±3% margin can be reached with about 1,000 respondents. Here are a few links that explain why this works: https://www.pewresearch.org/short-reads/2017/05/12/methods-101-random-sampling/ https://www.pewresearch.org/short-reads/2017/05/12/methods-1... https://www.surveymonkey.com/curiosity/how-many-people-do-i-need-to-take-my-survey/ https://www.surveymonkey.com/curiosity/how-many-people-do-i-... https://www.cloudresearch.com/resources/guides/statistical-significance/determine-sample-size/ https://www.cloudresearch.com/resources/guides/statistical-s...