Flashcards
extra chapter: Neyman, Pearson and hypothesis testing (Dienes, 2008)
Cards in this deck (9)
- 1difference subjective and objective probabilitysubjective probability: degree of conviction in a belief
objective probability: not in the mind but in the world; not your own belief, but based on observations. the probabilities exist independently of our states of knowledge - 2what is a collective or reference class?an infinite set of observations of a certain phenomenon
- 3what is heavily influenced by sample size?power (e.g. 20 per group can make the power 0.67, 44 per group can make the power 0.95)
- 4what's the consequence of having low power?getting a null result tells you nothing in itself, because the chance of getting a null result is very high when you have low power.
for example: if there really was an effect and the power was only 50%, out of 100 studies (that should get a significant result) only 50 would be labelled significant - 5meaning of stopping rulesthe conditions under which you'll stop collecting data for a study
- 6what's the standard Neyman-Pearson stopping rule?to use power calculations in advance of running the study to determine how many participants should be run
- 7what are 'sequential stopping rules'?in sequential testing you can look at your data several times and decide to stop collecting data at any of those times, because e.g. it's unethical to continue when there's already strong evidence that one treatment is more effective than another
- 8what is 'fisherian inference'?according to Fisher, the smaller the p-value, the stronger the evidence provided by the sample against the null hypothesis.
Neyman and Pearson disagreed, saying that you either reject or accept the null. - 9why does a more significant result not necessarily mean a more important result/larger effect size?because a large mean difference can be insignificant (e.g. small sample size), and a small difference significant (very large sample size)
Information
- Description
- ssr exam 2 Flashcards on extra chapter: Neyman, Pearson and hypothesis testing (Dienes, 2008), created by Sabine - on 18/10/2020.
- Created by
- Sabine -
- Views
- 0
- Rating