lecture 1- Ai Flashcards
What is the rationale behind null-hypothesis significance testing (NHST)?
Researcher formulates a null hypothesis (no effect) and an alternative hypothesis (there is an effect) after collecting data from a sample
What does NHST allow researchers to do with their hypotheses?
Reject the null hypothesis if data provide sufficient evidence against it, or reject the alternative hypothesis if insufficient evidence is found
True or False: A non-significant result in NHST indicates that the null hypothesis is true.
False
What is one major problem with NHST regarding the null hypothesis?
The null hypothesis is a hypothetical construct assuming the difference between groups is exactly 0, which rarely exists in reality
What does a non-significant result indicate?
The effect is not large enough to be detected with the given sample size
Why is it problematic to interpret a non-significant result as ‘no difference’?
It could be due to the null hypothesis being true or a failure to gather sufficient evidence
What does practical significance mean?
Statistical significance does not necessarily imply practical significance; small effects can be statistically significant but not important in practice
Provide an example of a practically irrelevant statistical finding.
IQ difference of 0.8 points between genders is statistically significant but practically irrelevant
What is the problem with all-or-nothing thinking in NHST?
Small differences in p-values can lead to opposite conclusions about significance despite indicating similar effects
What is the common alpha level used in psychology for significance testing?
α = 0.05
What are the proposed alternatives to NHST?
Effect size
What does effect size estimate?
The size of group differences or the effect of treatment
List three uses of effect size.
- Measure of how large an effect is
- Estimating sample size needed for sufficient statistical power
- Combining data across studies (meta-analysis)
What are the types of effect sizes?
- Group difference indices (e.g., Cohen’s d)
- Strength of association (e.g., eta squared, R squared)
- Risk estimates (e.g., relative risk)
What is Cohen’s d used for?
To measure group differences
What is the guideline for classifying effect size using Cohen’s d?
- d between 0.2 and 0.49 = small
- d between 0.5 and 0.79 = medium
- d of 0.8 and higher = large
How does Glass’ delta differ from Cohen’s d?
Glass’ delta uses the standard deviation from the control group rather than the pooled standard deviation from both groups
Fill in the blank: A measure of effect size for paired samples t-test is called _______.
[Cohen’s d for paired samples t-test]
What is Hedge’s g?
A corrected effect size that outperforms Cohen’s d when groups have different sample sizes and are below 20
What is the significance of the alpha level in NHST?
It is arbitrary and can lead to many published papers with p-values just below 0.05
