df/bad science Flashcards
What’s R2?
Strength of relationship
How well the model (regression line) represents the data.
X100 to get the result
Explains how much variability there is.
What are parametric tests?
They make assumptions.
What is the central limit theorem?
safely assume that the sampling distribution of the mean will be normal in most cases
shape of the sampling distribution will approach normality as the sample size (N) increases.
Assumes data is normally distributed regardless of the type of data we are sampling
How can we evaluate the parametric assumptions of a statistical model?
Extract and inspect the models residuals
What are residuals mathematically speaking?
The difference between observed data and model predictions.
How to work out independent samples?
Difference between groups \ variance within groups.
How to work out paired samples
Score change/ variance of that change
What consists of bad science?
Keep running the same experiment to make a type 1 error occur
Measuring so many things so might be significant by chance e’g fMRI, EEG, lots of conditions,
Could conduct different types of analyses on same data
What is the r squared?
Coefficient of determination.
Variance of one variable can be explained by the other variable
A value of 0 indicated it cannot be explained by the predictor variable at all.
Value of 1 indicates response variable can be perfectly explained without error by predictor
How do you work out total degrees of freedom?
Number of observations - 1
How do you work out residuals degrees of freedom?
Total df- regression df
How is the mean squares calculated?
Regression Ss divided by regression df
What does the pair wise t test tell you?
Which conditions there is a significant difference in the number of intrusions
What does the command rnorm do?
Takes random samples that is normally distributed
