Phil12_S11_Correlation&amp;causation(5-12-2011)

Phil12_S11_Correlation&causation(5-12-2011) -...

This preview shows pages 1–7. Sign up to view the full content.

Correlation and Causation Phil 12: Logic and Decision Making Spring 2011 UC San Diego 5/12/2011 Thursday, May 12, 2011

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Review: 2 types of correlational studies 1. When same items have values on two score variables, correlate the scores on one with the scores on the other - Measure degree of correlation in terms of Pearson coefFcient r - Predict value on one variable from that on the other using the regression line: y=ax+b 2. When one nominal variable divides a population into two or more sub-populations, compare the two populations on another (score) variable in terms of their central tendencies - If the means are different, predict the value on the score variable depending on the value of the nominal variable Thursday, May 12, 2011
Review In both types of correlational studies, one commonly makes inferences from a sample to an actual (total) population - Does what is found in the sample apply to the actual population? - Addressed in terms of statistical signifcance Is the result in the sample one that would be unlikely to happen by chance if there weren’t a correlation or a difference in the actual population? The p-value speciFes the likelihood of the result in the sample happening by chance (in drawing the sample) - p < .05 indicates there is less than 5% chance of the result happening by chance Thursday, May 12, 2011

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Review In testing a correlational claim, one tests the null-hypothesis - There is no correlation - There is no difference in the means The strategy is to try to reject the null hypothesis in terms of the results in the sample - If result for the sample is statistically signiFcant (at a chosen level), one infers that the null hypothesis is false - If the result for the sample is not statistically signiFcant (at the chosen level), one cannot reject the null hypothesis Whatever correlation or difference between group means there might be, they will not have been detected Thursday, May 12, 2011
Suppose that a drug company tests whether there is any difference between the rate of strokes in patients taking drug A versus drug B. It gives one group of people drug A while another group gets drug B; then measures the incidence of strokes in each group. The people in group A averaged 3 fewer strokes than group B. The p value was 0.04. The company concludes that people taking drug A have fewer strokes. But in reality, people taking drug A and drug B have the same rate of strokes. Clicker question - setup Thursday, May 12, 2011

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Suppose that a drug company tests whether there is any difference between the rate of strokes in patients taking drug A versus drug B. It gives one group of people drug A while another group gets drug B; then measures the incidence of strokes in each group. The people in group A averaged 3 fewer strokes than group B. The p value was 0.04. The company concludes that people taking drug A have fewer strokes. But in reality, people taking drug A and drug B have the same rate of strokes. Clicker question
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 06/07/2011 for the course PHIL 101 taught by Professor Brown during the Spring '08 term at San Diego.

Page1 / 36

Phil12_S11_Correlation&causation(5-12-2011) -...

This preview shows document pages 1 - 7. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online