Psych 2300: Chapter 8 Notes
Psych 2300: Chapter 8 Notes Psych 2300
Popular in Research Methods In Psychology
Popular in Psychlogy
This 3 page Class Notes was uploaded by MadsSwart on Wednesday July 6, 2016. The Class Notes belongs to Psych 2300 at Ohio State University taught by Emanuele Rizzi in Summer 2016. Since its upload, it has received 41 views. For similar materials see Research Methods In Psychology in Psychlogy at Ohio State University.
Reviews for Psych 2300: Chapter 8 Notes
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 07/06/16
Chapter 8: bivariate correlational research - When you see two things that are related, and you think are related, but actually have nothing to do with one another To be causal - Covariance - Internal validity - Temporal precedence Correlational strength - As you get a measurement change by this much on this chart or whatever - You get the same ratio change on the other chart for the other variable - This means that they are changing at the same rate o Either moving at the same amount in the same direction [perfect correlation o Or moving at the same amount in the opposite direction [perfect correlation - When they don’t move at the same rate; imperfect correlation – this is what r actually does; what r is saying is that this is how much [the magnitude] or proportion of changes in each other - X percent of change in this can predict the change in this - The more correlated things are, the closer they will fall to the line, and the stronger the correlation is; closer to 1 T test and anova are used to measure causal but can also be used to measure association Association – measured not manipulated Association does NOT equal causation Not the test statistic that matters; how the test was measured Associations with categorical variables - Causal claim version - Assign conditions ; manipulate - Comparison is the same - To compare marital satisfaction among people who met naturally vs. online you would have to force people to marry someone online vs. in person [that’s what you would be manipulating/controlling] - So even though you are still performing a t test, the causal claim isn’t best Interrogating association claims - Construct validity o How well was each variable measured o Evaluate every variable of interest; every variable you are trying to discuss o Address for each variable EX: take significant other out to scary movie or romantic movie? Two bridges: scary bridge and control bridge with attractive female Can which bridge you cross assess heightened emotional state o Construct Validity of Study Thematic Apperception Test [TAT] & Follow-up Call [behavioral] Interrater Reliability; multiple measures show convergence: validity Heart rate risen; automatic so good measure of sexual arousal TAT: show ambiguous picture and describe picture o Someone aroused gives more detailed explaination Variables: type of bridge crossed (scary vs. control) – categorical Arousal / attraction (TAT score & % call back) TAT not a good way to measure; if you don’t believe in the way the variable is measured, then the claim cannot be made. Call back could be a better measure - Statistical validity o How well do the data support the data support the conclusion? > Accuracy of Measurement >how important/meaningful are the results? o Effect Size Strength of an association; falling more in line with each other [spread] Smaller spread = greater effect size Stronger correlations = more accurate predictions Predictive and Causal are NOT the same thing o Statistical Significance P-values probability that you’re correlation IS FALSE The lower the probability, the more likely that it’s true In world/truth there is no correlation – so should show no correlation Larger effect size = likelier to be statistically significant o Outliers Most dangerous when you have a low amount of data Less effective with a large number of data points Pulls data in its direction Be aware when dealing with association claims Evaluate statistical validity/outliers rationally o Range Restriction Having as much of the scale as possible helps you to know for sure that the correlation/relationship you are observing is representative of the ENTIRE set of data o Curvilinear Relation May be predictive in a way other than a straight line Could be going up and then down etc. Correlations tell you how well data fits on a line; not if there is a correlation - Internal Validity o Can we make a causal inference from an association? o Do we assess internal validity for associative claims? – No; only for causal Why is the relationship not causal? Causality is tempting, but need to be able to differentiate between the 2 o Third-variable problem: two variables seem causal/related when there is a third variable making it seem like there is a relationship must be correlated to both other variables o Directionality Problem: you can’t change the direction of the variables and get the same result[order matters] o Spurious association: not due to an actual relationship between variables; correlation exists because of some other third variable don’t have anything in common with each other - External Validity o To whom can the association be generalized? o In associative claims No random, no fowl Not using random sampling does not reject association Relationship between variables may still generalize WITHIN other populations Pearson’s R: Strength and Direction of Correlation Associative Relationship: 1- Related variables 2-measured variables
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'