Psyc 333 Study Guide: Causality, Experiments, Surveys, and Culture of Honor

Flashcards with vocabulary and definitions

Causality

Identifying what causes what; a tricky goal in psychology and science.

Randomized Experiment

A powerful way to infer cause-and-effect relationships between variables by randomly assigning participants to different groups.

Random Assignment

Assigning participants to different groups in an experiment with an equal chance of ending up in any group.

Independent Variable

The thing changed by the experimenter.

Dependent Variable

The outcome measured in an experiment.

Confounding Variables

Other factors that can influence outcomes in a study.

Internal Validity

How confident we can be that the independent variable truly caused the effect on the dependent variable.

Random Sampling

How you pick people from a population to be in your study; helps with generalizability to the broader population.

External Validity

Generalizing results to a population.

t-test

A test that checks if two means (averages) are significantly different from each other.

F-test

Often used in analysis of variance (ANOVA), a comparison of two groups.

Chi-square test

Tests used for categorical data.

p-value

The probability of getting your observed data (or something more extreme) if there really was no true effect.

Demand Characteristics

When participants pick up clues about the purpose of the study or what behavior is expected, and then alter their behavior.

Expectancy Effects (Observer Expectancy)

When a researcher's expectations unintentionally influence participants.

Pygmalion effect or Rosenthal effect

Another name for expectancy effects.

Double-blind designs

Where neither participants nor experimenters know who is in which group.

Systematic Error

Errors that bias the results in a particular direction.

Random Error

Error which just adds noise.

Field experiments

Experiments done in real-world settings to improve external realism.

Randomized controlled trials (RCTs)

Considered the “gold standard” for testing things like new therapies, medications, or interventions because of their strong causal inference.

Institutional Review Boards (IRBs)

Boards which oversee ethical guidelines.

Alternatives (Quasi-experiments and Matching)

Used when random assignment isn’t possible.

Propensity Score Matching

In a large observational study (no random assignment), analysts can mathematically create groups that are similar on a host of variables.

Converging evidence

Using multiple methods that converge on the same finding, because each method’s strengths can make up for the weaknesses of others.

Parsimonious

The simplest explanation that fits all the evidence.

Construct Validity

Did you actually measure what you intended to measure?

External Validity

Do the results generalize to other people, places, times?

Statistical Conclusion Validity

Did you use the right statistical methods, and do you have enough data to support your conclusions?

Internal Validity

Did you properly establish cause-and-effect without confounds?

Non-attitudes

When someone might give an opinion on a survey, but that opinion could be flimsy – it might change tomorrow, or it might be something they haven’t thought through.

The miracle of aggregation

A nickname for the Law of Large Numbers in the context of opinions.

Specificity matching

Attitudes predict behavior better when the attitude is measured at a level specific to the behavior.

Channel Factor

A small prompt or situational factor that helps “channel” an intention into actual behavior.

Culture of Honor

A social framework where people (particularly men) are highly sensitive to reputational slights and insults, and are expected to defend their honor – often with violence if necessary.

Effect size

A measure of how much something changes or how strong a relationship is, in standardized terms.

Cohen’s d

Measures the difference between two group means in units of standard deviation.

Expected Value (EV)

A formula from decision theory/probability. Expected value = probability of an outcome × value (or payoff) of that outcome, summed across all possible outcomes.

Law of Large Numbers

A principle stating that as the number of trials or sample size grows, the sample mean will get closer and closer to the true population mean, and results become more stable.

Stand your ground

A law where you do not have to retreat if threatened.

Demand Characteristics

Cues in an experiment that tip off participants to what the study is about or what behavior is expected from them.

Expectancy Effect

When a researcher unintentionally influences participants to behave in a way that confirms the researcher’s hypothesis.