Establishing Associations Flashcards
What is an association?
An association is wherebycertain values of one variable tends to correspond with certain values of another variable.
Why do correlations not always equal causation?
Correlations do not equal causation as there may be issues of reverse causality or spurious association.
What are the three components of association?
- Nature/direction: increase or decrease! Between which variables?
- Strength of the relationship: what is the strength of an increase/decrease
- Statistical significance of the relationship: how likely is it that an association will generalise to the population?
How can you establish the nature of a linear association and the strength of the association?
The nature of a linear association can be seen through the sign of the slope coefficient (positive vs negative), while the strength can be seen through the size of the slope coefficient.
What is Pearson’s r? What values does it range between and what do these values mean?
Pearson’s r summarises the association between two quantitative variables in a numerical form. It ranges between -1 and 1, with 1 as the perfect positive correlation and -1 as the perfect negative correlation.
What is a t-test? What does it allow us to derive?
A t-test is a statistic which collects all evidence against a null hypothesis. It compares how much variation in Y can be explained by X and how accurately we can measure the population slope.
It allows us to derive the p-value.
What do statistical control techniques aim to achieve?
Statistical control techniques aim to measure and account for potential confounding variables. It allows us to rule out these playing a role in producing our results.
