Stats Metacog 4 Flashcards
What is the consequence of making a Type I error?
Concluding the treatment has an effect when it truly does not
T or F: a statistically significant treatment effect does not always indicate a large treatment effect
True
T or F: if the power for a hypothesis test is .80, then the probability of committing a Type II error = .20
True
What do you first use to calculate power?
Delta
What is the equation for delta?
δ = 𝑑 × (square root of n/2)
How do you interpret the delta?
Using the delta table
How does the size of the sample effect delta?
As the sample size increases, so does the power
What does a power table do?
Can be used in place of a delta table. You work backwards to figure out how large of an effect or sample size you would need for a given sig level
When can you calculate power?
Before completing a study or after as a post hoc
What is correlation?
Correlation is the measure of how related two things are to one another
What are other names for Pearson’s correlation coefficient?
Pearson’s r, Pearson’s p, Bivariate correlation, r, or p
What does Pearson’s r measure?
measures the linear relationship between two continuous variables (X and Y). It also describes the strength of the relationship
What is the range of r?
- Ranges from -1 to +1
What does a + r represent?
A positive relationship between X and Y
What does a - r represent?
A negative relationship between X and Y
What does a scatterplot do?
It allows us to plot both X and Y at the same time
T or F: Scatterplots can be in different scales
True
What are we able to do if there is a stronger relationship between X and Y?
We are better able to guess or predict a value knowing the other
In what situations does r not work well?
- Nonlinear data
- When variation in one variable increases as the other variable changes values
- Skewed data
- When the data has outliers
What are the steps to calculate Pearson’s r?
- Make a new XY variable
- calculate the mean of X, Y, and XY
- Calculate the sum of X, Y, and XY
- Calculate the SS for X and Y
- Solve for Rxy (using the formula)
What will a strong relationship look like?
Data points that are close together without a lot of variation
What can correlations help us do?
Allows us to predict where a data point will lie to a certain extent
