Lecture 7 Flashcards
What is the purpose of inferential statistics?
Use samples to infer about populations (as long as the sample is representative)
What do inferential statistics calculate?
The probability that the resulsts could have occured by chance
What are the two criteria when choosing which inferential test to use?
Research question and type of data
If the data is nominal? Which inferential tests would you use?
Binomial/ chi-squared test
If the data is continuous data? Which inferential test would you use?
Depends on research question:
Relationship - correlation
Difference - t-test/ similar
If the data is parametric which inferential test would you use?
Methods for noramlly distributed data
If the data is non-parametric which inferential test would you use?
Methods for non-normally distributed data
Parametric data assumes:
- Data is interval/ratio
- Both conditions have similar variances
- data is normally distributed (proven by shapiro wilk)
When would you use non-parametric tests?
When the criteria for parametric tests arent met, however, it has less statistical power
What are the 2 types of t-tests?
Related t-test
Unrelated t-test
What is the related t-test used to test
the null hypothesis
How does the t-test work?
Calculates the difference score for each subject and test the null hypothesis. The null argues that the mean difference will be 0.
What is the central question behind the unrelated t-test?
Is there enough evidence to conclude that the 2 samples were drawn from different populations?
The unrelated t-test has different equations depending on?
if sample sizes are equal or unequal
How do you calculate the df?
n-1
