Open science and meta science

Question 1

What can be the cause of type 1 and type 2 errors?

Answer 1

• Randomly unlucky
• Systematic, consistent errors due to some human factors e.g. publication bias of QRPS

Question 2

What is meta- science

Answer 2

Using scientific methodology to study the scientific process itself (science about the science)

Question 3

Why is meta science useful?

Answer 3

• to identify strengths and weaknesses of the process and explore potential solutions
• Became more used since the replication crisis
  Can be due to the publication bias (novelty + significance) and the over emphasis on statistical testing

Question 4

What are the 4 types of replication?

Answer 4

Reproducible

Robust

Replicable

Generalisable

Question 5

What makes a study reproducible?

Answer 5

A result is reproducible when the same analysis steps performed on the same dataset consistently produces the same answer

Question 6

What makes a study robust?

Answer 6

• when the same dataset is subjected to different analysis

Question 7

What makes a study replicable?

Answer 7

• the same analysis is performed on different datasets

Question 8

What makes a study generalisable?

Answer 8

• New data set and new analysis

Question 9

What does open science do?

Answer 9

makes it easier to do science that other people can replicate and generalise

E.g. pre- registration, open access to publication, open data, open evaluation etc