PSYC300 Notes- Research Designs.docx

3 Pages
Unlock Document

University of Maryland
PSYC 330
Andrea Chronis- Tuscano

Sources of error - Random Error- other influences causing effect - Self- esteem  likert scale  random error Other conceptual variables--^ (ex. Mood, self- promotion, filler questions, reverse score, modesty, narcissism) Reliability- freedom from random error, whatever it is you are measuring is consistent - Consistency across time o Test- retest (usually equivalent forms- same topic diff questions)  Taking GRE twice should get similar score- limits to how much improve  Can’t measure mood b/c is a state and changes over time o Use correlation to determine reliability - Consistency across items (internal consistency) o Split half- look at score of first 50 questions and score of second 50 questions, and correlation should be high o Coefficient alpha- Correlation of all items (question) on a test, want alpha closer to one - Consistency across judges (inter- rater reliability) o More than one judge evaluating something Actual score= true score+ random error Reliability= true score/ actual score - Proportion of the total that is true Increasing Reliability - Use good items o Apriori (pretesting) o Post hoc (item deletion) - Increase the number of items Construct Validity- is the test measuring what I think it’s supposed to measure - Subjective methods o Face validity- what appears relevant or right o Content validity- - Objective methods- correlations between similar type measures o Converging Operations- different measures of same conceptual variables should be correlated, want as close to 1  Ex. Self- esteem likert scale  Autobiographies o Discriminating operations- want low correlation close to 0 between conceptual variables  Ex. If self- esteem and optimism have high correlation then are measuring the same variables and not discriminating between them o Criterion Validity- paper and pencil test used to predict behavior  GRE measuring measure ability to perform well in college, I/O psychologists determine who’s going to be good at a job 10/15/12 Chapter 9 Correlational Research Designs Abstract 150 words (intro, method, results, discussion sentence) Line of best fit- Minimizes sum of square distances (sum of squares) Curvilinear r=0 Interpreting r - Assesses Linear relationships only - lrl effect size statistic, size of relationship - R^2= proportion of variance, turns out to be smaller, amount of variability in dependent variable explained for by independent variable o Ex. The proportion of aggressive behavior explained for by watching TV o Ex. r= .85 but r^2= .72 so only explaining .72 of variance, other .28 explained by something else  Explained (.82): how much of total is accounted for by correlation - Restriction of range (cor
More Less

Related notes for PSYC 330

Log In


Don't have an account?

Join OneClass

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

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.