Payment card data suggests that the drop in consumption during the second wave of Covid-19 is much smaller than in Spring and is “localized” only in some sectors (more goods, fewer services). Less uncertainty about the future income and jobs, progress on a vaccine front, readiness to sell and buy online and other factors contribute to a milder economic effects.
What statistical concepts should every person know in order to get a better understanding of social science research, such as economics, sociology, psychology, etc.?
As a psychologist, I see a lot of value in understanding the following:
- population vs. sample vs. data
- independent vs. dependent variable
- causation vs. correlation vs. regression
- moderation vs. mediation
- gaussian vs. bayesian statistics
Hope it's not to much. Thanks!
Asked by Povilas Godliauskas
Thanks for the question! This is a broad one. Let me talk about one point for now.
Binary outcome VS multi-option(or even continuous) outcomes.
We(people) tend to generalise everything to binary - yes/no outcomes too much.
1. Profit or Loss
2. Success or Failure
3. Happy or Sad
4. Political right or left
However, world is more complex than that. Thus we should think more about multi-option outcomes. Maybe this way we could better understand social phenomena and each other.
Want to generalise a bit on Marius insight.
Did society always struggled to keep people in charge accountable for their promises? Or is it just a recent trend?
It became hard to even remember promises due to increase in information overflow from web/TV. Not even talking about making someone accountable.
What do you think?