Chapt 6 - Correlation are two things related or associated...

Info iconThis preview shows pages 1–5. Sign up to view the full content.

View Full Document Right Arrow Icon
Correlation are two things related or associated? beliefs of association common to all of us (could be true or not) moon phase (full) and behavior in humans (crazy) astrology—date/time of birth – persons personality red hair – temperament immediately soothing baby when crying – bad sex/gender – emotionality Statistic to measure the strength and the nature (direction) of relationship between two variables degree to which they go, or vary, together A statistical index that indicates degree to which two variables are related to each other in a linear fashion can range from +1.00 perfect positive relation as scores on one go up scores on other go up amount of studying and grades to .00 no linear relation to -1.00 perfect negative relation inverse relationship if scores on one variable are high, they are low on other amount of alcohol consumption and grades
Background image of page 1

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

View Full DocumentRight Arrow Icon
Concrete example we can all picture: Height and weight correlate about .70—strong correlation Correlation coefficient not only tells you about the direction of relationship (by sign it has) but also about its magnitude (or strength—how big is r) .40 generally considered as strong correlation in social sciences .25 moderate and pretty typical of effect sizes the sign of correlation only tells you about the direction but nothing about its strength -.78 stronger relationship than .43 (if no sign, assume +) Scatter plot--graph scores on two variables for entire sample for each participant plot point that indicates how s/he scored on two measures looked at graphs that show correlations that differ in strength and direction (see also book p. 151) positive vs negative stronger—more tightly clustered ( r = .00 vs 1.00)
Background image of page 2
only index of LINEAR association curvilinear relationship not captured need to plot your data!!! Factors that can Distort Correlation Coefficients correlation coefficient you compute could be an over- or an underestimation of the true degree of relationship b/w two variables Need to check three factors before safely interpreting your r (1) Restricted Range Understand figures in book p. 149 Homogeneous sample (whose scores fall w/in narrower range) produce smaller correlations than those obtained from a heterogeneous sample (with wider range of scores) when have Restricted range—misled into concluding that two vars are only weakly correlated or not at all however, if a broader range of scores had been sampled, strong relationship would have emerged THUS=>examine your raw scores so you can see whether range of scores might be restricted
Background image of page 3

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

View Full DocumentRight Arrow Icon
(2) Outliers Scores that are obviously deviant from remainder of data that one can question whether they belong in data set at all Score considered outlier if more than 3 SDs from mean of data (what do 3 SDs mean/imply?) two kinds of outliers-- figure p. 151 in book
Background image of page 4
Image of page 5
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 04/22/2008 for the course PSYCH 31:043 taught by Professor Klohnen during the Spring '08 term at University of Iowa.

Page1 / 23

Chapt 6 - Correlation are two things related or associated...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online