Contemporary Economic Challenges B: 9 & 10 - Should We Trust Empirical Results in Economics Journals? Flashcards
- Understood the concept of p-value as a measure of statistical significance, and what we might learn from looking at patterns of p-values in an empirical literature. - Discussed the problem and implications of p-hacking, why it arises, and what could be done about it. - Discussed the problem and implications of publication bias, why it arises, and what could be done about it.
1
Q
Answer simply, how far should we trust
empirical results published in Economics journals?
What’s the silver lining?
A
Not very far, most papers are wrong. The silver lining is that it makes reading studies exciting because this field of research is much more contested than one might think
2
Q
Give 2 examples of what sort of questions might be asked on this topic
A
- There is increasing concern about the credibility of published research
findings in empirical economics. Explain why this is the case, and what
can be done about it. - “Research practices in the social sciences mean that the evidence upon
which citizens and policymakers must base policy is often far from
credible”. Explain in detail the reasons why that might be the case, and
evidence in support of the conjecture. What can be done about this?
3
Q
Introduce the main basis of this topic
A
- Research credibility and the trustworthiness of empirical results in
economics and other social sciences has come under intense scrutiny
in recent years. - Question: If you read a result X causes Y, written up in an academic
journal, or reported in the media, should you ‘believe’ it? Or should
you be sceptical? - Experimental Psychology in the past 20 years have been having HUGE problems with repeatability. Economics has been facing the same issue but not to the same extent
