Class Notes (835,293)
Psychology (2,710)
PSYC 2001 (163)
Lecture

12 Pages
128 Views

School
Department
Psychology
Course
PSYC 2001
Professor
Guy Lacroix
Semester
Fall

Description
Correlational research Correlational research - What are correlations? • Scatterplots • Correlations and causality - What are regressions? • y = mx + b • Multiple regressions - Examples • Siddiqui, West, and Stanovich (1998) Correlational research - The purpose of correlational research is to establish that a relation exists between variables and to describe the nature of this relationship. • Rubin’s (1970) love scale and “being in love” • Reliability coefficient is .87 to 1.0 (Bowker et al., 2009) Pearson correlation coefficient - The Pearson correlation (r) coefficient is the statistic that indicates the degree to which two variables are related to one another in a linear fashion. - Range r is between – 1.00 and + 1.00 - Positive correlation (r > .00). Positive, direct relationship between two variables - Negative correlation (r < .00). Negative, inverse relationship between two variables No relation between the variables Perfect relation between the variables Strong relation between the variables Weak relation between the variables How can we tell that a correlation is significant? - What is the probability that a “lucky result” was obtained? - The probability is calculated using the Student’s t statistic. Critical values of r Number of participants Minimum r that is significant (for a p value of .05) 10 .52 20 .37 30 .30 40 .26 50 .23 60 .21 80 .18 100 .16 1000 .05 Correlation matrix - When multiple correlations are used, the results are typically shown in a correlation matrix. Index 1 2 3 1. Grades 1.0 - - 2. Performance IQ .65 1.0 - 3. Hours of study .34 .01 1.0 Correlations and causality - ACorrelation does not imply causality! - Three criteria must be satisfied to conclude that one variable causes another: • Covariation. There is a correlation between variables X and Y. • Directionality. Variable X (the cause) precedes Variable Y (the effect). • Elimination of extraneous variables. No other variable may be responsible for the relationship between X and Y. Directionality problem - Correlational studies do not allow us to determine which variable is the cause and which variable is the effect. - Self-esteem  school achievement (Stanovich example 2007) Third-variable problem - Although a correlational study may establish that two variables are related, it does not necessarily mean that there is a direct relationship between them. Linear regressions - Another way of looking at the relation between two variables. - The best fitting linear function is estimated and then used to “explain” the changes in one variable using a second variable. - Criterion. The variable to be explained. - Predictor. The variable used to predict, to “explain”. - Correlation between women owning a toasting and being on contraceptives; not causal y = mx + b Effect size for regressions - Using the regression equation, we know that the predictor (x) explains a 31% change in the criterion (y). 2 - r . The measure of effect size for regressions. Multiple regression - Regressions allow for multiple predictors. - Example. Suppose you want to explain grades using both IQ and number of hours of study. Index 1 2 3 1. Grades 1.0 - - 2. Performance IQ .65 1.0 - 3. Hours of study .34 .01 1.0 Multiple regression - R. Measure of relation for multiple variable. - R . The measure of effect size for multiple regressions Hierarchical multiple regressions - What if I want to see the relation between study time and grades independently of IQ? - Mathematically, the variance explained by IQ is removed after step 1. Hence, the variance explained by study time is “unique”. - Final F= probability that the result is due by chance • Percentage that results could be lucky; is it significant or not. 0.12 explains 12% of the leftover variable that has not be explained by IQ Criterion = Grades Variable R  R 2 Final F Step 1 1. IQ .65 .42*** 26.82*** Step 2 2. Study time .73 .12** 20.82*** Author Reco
More Less

Related notes for PSYC 2001
Me

OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Join to view

OR

By registering, I agree to the Terms and Privacy Policies
Just a few more details

So we can recommend you notes for your school.