Revision (1) Flashcards
What is at the heart of all tests except for tests of correlation?
Finding out whether a change in an independent variable causes a change in a dependent variable.
Which test would be most appropriate to run if you have a continuous independent variable?
A test of regression.
Which test would be most appropriate to run if you have a categorical independent variable involving two levels?
A t-test
Which test would be most appropriate to run if you have a categorical independent variable involving more than two levels?
An ANOVA test.
What must always be the nature of the dependent variable involved in order to run tests of regression, t-tests, and ANOVA tests?
The dependent variable must always be continuous to run these tests.
When are t-tests conducted?
When you only have two sets of scores to compare and you wish to ask whether they appear to be different.
What is the formula for the t-statistic generated after having run a t-test?
t = difference between group means/ overall variability of data (signal/ noise)
If two datasets, one with a greater difference between group means than the other, and both with equal overall data variability, were to be compared, which dataset would be more likely to result in a larger t statistic?
The dataset with a greater difference between group means.
If two datasets, one with lesser overall data variability than the other, and both with an equal difference between group means, were to be compared, which dataset would be more likely to result in a larger t statistic?
The dataset with lesser overall data variability.
Is a larger t statistic likely to be associated with a larger or smaller p value?
A smaller p value, seeing that a larger t statistic signifies a larger effect size (and a larger effect size represents a lower likelihood of the null hypothesis being true).
To what does the term ‘covariance’ refer?
To when a participant deviates from the mean on one variable, as well as another, in either a similar or opposite way.
