Psychology 3401 Exam 3

Understand how confidence intervals relate to sample size and statistical significance.

- Bigger sample = smaller confidence interval - Statistically significant means > 5% chance of type 1 error (reject null when null is true) - usually, p < .05 - If 95% confidence interval for test statistic contains 0, not significant - if confidence interval does not contain 0, significant

Understand what it means to explain variability in a variable.

Describing how spread out or dispersed the data points are within a specific variable.

Be able to interpret correlations in light of the directionality and third variable problems.

Directionality: correlation does not imply causation, meaning you cannot definitively say which variable is causing the other; can only state they are related Third variable: just because two variables are related, does not mean one causes the other; there may be a third unknown variable causing the relationship

Distinguish between positive and negative correlations.

Positive Correlation: a statistical relationship between two variables that move in the same direction; when one variable increases, the other also increases; when one variable decreases, the other also decreases (ex: the more calories you eat, the more you weigh) Negative Correlations: a statistical relationship between 2 variables where one variable increases as the other decreases (ex: more you exercise, the less body fat)

Be able to evaluate correlational findings using the three causal criteria.

1. Covariance of cause and effect - a correlational establishes covariance of cause and effect 2. Temporal precedence - if we know that one variable preceded another, we can establish temporal precedence 3. Internal validity - examine whether study design adequately controls for potential confounding variables, ensuring that the observed correlation between variables is not due to other extraneous factors, and critically consider the direction of causality

Identify 2 issues with bivariate correlations.

1. Directionality Problem -Temporal Precedence 2. Third Variable Problem -Internal Validity

Identify 2 options for ruling out third variables.

1. Experiments - test what you want to discover 2. Multivariate Designs - assess the relationships among more than two variables simultaneously to account for as many potential third variables as possible

Identify 6 concerns in determining the statistical validity of correlational effects.

1. How strong is the relationship between two variables? 2. How precise is the estimate of the relationship? 3. Has the finding been replicated? 4. Could outliers be affecting the association? 5. Is there restriction of range? 6. Is the association curvilinear?

Understand how longitudinal studies fair in regards to the 3 causal criteria.

1. Is there a correlation? - yes, if p<.05 and/or 95% confidence interval does not contain 0 2. Is there temporal precedence? Yes, if cross-lag correlations are significant 3. Is there internal validity? -Maybe, maybe not

Identify and describe 3 possible patterns of cross-lag correlations.

1. variable 1 at time 1 shows significant correlation with variable 2 at time 2 (while variable 2 at time 1 doesn't show significant correlation with variable 1 at time 2) 2. variable 2 at time 1 shows significant correlation with variable 1 at time 2 (while variable 1 at time 1 doesn't show significant correlation with variable 2 at time 2) 3. both correlations are significant (mutually reinforcing)

Define multiple regression.

A statistical technique that computes the relationship between a predictor variable and a criterion variable, controlling for other predictor variables.

Define statistical control, and understand how it is used to rule out third variables.

-Statistical control: a method of using statistical procedures to remove the influence of extraneous variables in a study - allows researchers to isolate the relationship between the variables of interest and ensure that findings are accurate

Explain what it means for a correlation to be curvilinear.

- an association between two variables which is not a straight line; instead, as one variable increases, the level of the other variable increases and then decreases (or vice versa)

Understand relationship between correlation and causation.

- correlation does not imply causation - BUT causation does imply correlation

Understand how outliers can affect correlational findings.

- distort the calculated correlational coefficient, making the relationship between variables appear stronger or weaker than it actually is - an outlier can skew or pull the regression line their direction - bigger samples are less affected by outliers

Understand the importance of replication in correlational research.

- helps verify results, identify potential errors or biases, contribute to body of knowledge, and uphold the integrity of scientific process - type 1 error = claiming effect when none exists - if we're 95% confident, then 5% of the time we're wrong - be most skeptical of studies with small samples and large effect sizes

Explain how restricted ranges can mask correlations

- in bivariate correlation, the absence of a full range of possible scores on one of the variables, so the relationship from the sample underestimates the true correlation - correlation relies on variability, and restricted ranges reduce variability

Describe the relationship between predictor and criterion variables.

-A predictor variable is a variable that is expected to be used to predict the value of another variable (the criterion variable) -A criterion variable is a variable whose value is expected to be influenced by another variable (the predictor variable)

Describe cross-sectional correlations, auto-correlations, and cross-lag correlations.

-Cross-sectional: a correlation between 2 variables that are measured at the same time -Auto-correlations: the correlation of one variable with itself, measured at two different times -Cross-lag: a correlation between an earlier measure of one variable and a later measure of another variable

Articulate the theory-data cycle.

-Hypotheses: Researchers predict what else they should observe if the theory is correct -Data: Researchers observe something -Theory: Researchers propose a reason for those observations 1. you form a theory 2. based on that theory, you form a hypothesis 3. you test that hypothesis 4. if you fail to disprove it, add a point to that theories truthiness (the more points a theory has, the more confident we become that it is true)

Understand how longitudinal designs establish temporal precedence.

-Longitudinal Design: a study in which the same variables are measured in the same people at different points in time -allows researchers to determine occurs before the other, indicating potential causal relationships

Define mediation.

-Mediator: a variable that helps explain the relationship between two other variables -Answers the question: How does a change in X cause a change in Y?

Understand the difference between mediation and moderation.

-Mediators explain why X causes Y (ex: eating cheese causes upset stomach in lactose intolerant individuals because they can't fully digest the sugar lactose) -Moderators mean that X affects Y differently depending on the level of the moderator (if you are lactose intolerant, then eating cheese will give you an upset stomach) >either the same person is affected by X differently depending on their current state >or different types of people are affected by X differently

Know how to identify regression studies.

-Regression Buzzwords: "Controlled for" "Adjusted for" "Considering" "Accounting for" "Taking ____ into account"

Distinguish between bivariate and multivariate correlation.

Bivariate Correlation: an association that involves exactly two variables - reasons for using bivariate: if x causes y, x and y will be correlated; sometimes we only care about two variables; multivariate designs can be taxing for participants Multivariate Correlation: an association that involves more than two variables

Understand how researchers use pattern and parsimony to build a causal case.

By examining multiple correlational studies that consistently show a relationship between two variables, with the simplest explanation (parsimony) being that one variable directly causes the other

Understand the importance of testing for mediators.

helps researchers understand how variables influence each other, and it can help improve programs and theories

Understand how moderators can affect the relationship between two variables

moderator- a variable that, depending on its level, changes the relationship between two other variables - ex: if I'm hungry, the smell of food is intoxicating; if I'm full, the smell of food can be revolting (fullness level is a moderating variable that changes how the smell of food affects me)

Understand the r statistic and be able to identify small, medium, and large correlational effects.

r= Pearson's coefficient (measure of effect size in bivariate correlations) - ranges from -1 (perfect negative correlation) through 0 (no correlation) to 1 (perfect positive correlation) - .1 = small - .2 = medium - .3 = large - .4+ = very large