Module 2 Research Methods

Scientific method steps

Observe/ask a question → form a hypothesis → design study and collect data → analyze results → draw conclusions, report, and replicate

Variable

Any factor that can change or vary in a study

Independent variable (IV)

Variable the researcher manipulates; the presumed cause

Dependent variable (DV)

Variable the researcher measures; the presumed effect of the IV

Theory

Well-developed set of ideas explaining observed phenomena

Hypothesis

Testable prediction, often "if-then"; derived from a theory

Deductive reasoning

General premise/theory → specific predictions (how hypotheses are formed from theories)

Inductive reasoning

Specific observations → general conclusions/theories

Falsifiable

Able to be disproven by results; required so a hypothesis can truly be tested

Operational definition

Exactly how a variable will be measured or manipulated in a study

Descriptive research

Describes behavior without testing cause; includes case studies, naturalistic observation, surveys, archival research

Case study

In-depth study of one or a few people; strength = rich detail, rare cases; weakness = can't generalize

Naturalistic observation

Watching behavior in its natural setting; strength = real behavior (high validity); weakness = observer bias, no control, time-consuming

Observer bias

Observations skewed to match the observer's expectations

Inter-rater reliability

Agreement among different observers recording the same event

Survey

Questionnaire or interview given to many people; strength = large data quickly; weakness = self-report may be inaccurate (social desirability), little depth

Archival research

Uses past records or existing data sets; strength = cheap, no reactivity; weakness = can't control what data exists or its quality

Longitudinal research

Same people measured repeatedly over a long time; strength = shows change over time; weakness = costly, slow, attrition

Attrition

Participants dropping out of a study over time

Cross-sectional research

Compares different age groups at one time; strength = fast, cheap; weakness = cohort (generational) differences confound results

Correlational research

Measures whether two variables are related, without manipulation; allows prediction

Correlation strengths

Allows prediction; can study variables that can't be ethically manipulated; real-world data

Correlation weakness

Correlation does NOT prove causation (can never tell us cause and effect)

Correlation

Relationship between two or more variables; one changes as the other changes

Positive correlation

Both variables move in the same direction (both up or both down)

Negative correlation

Variables move in opposite directions (one up, other down); not the same as no correlation

Correlation coefficient (r)

Number from -1 to +1; sign = direction, absolute value = strength (0 = no relationship)

Illusory correlation

Believing a relationship exists when it does not

Confirmation bias

Tendency to seek/notice evidence that supports beliefs and ignore evidence against them

Third variable problem

An unmeasured outside variable causes changes in both correlated variables, creating a false appearance of cause

Experiment

Researcher manipulates an IV and measures a DV while controlling other factors; only method that shows cause and effect

Requirements for an experiment

Manipulation of an IV, random assignment, and a control/comparison group

Experiment strengths

Determines cause and effect; high control

Experiment weaknesses

Artificial setting (may not generalize); some variables can't be ethically manipulated

Cause-and-effect relationship

Changes in one variable cause changes in another; only shown by experiments

Population

Entire group researchers want to learn about

Sample

Subset of the population actually studied

Random sample

Every population member has an equal chance of being selected; most representative

Representative sample

Sample that accurately reflects the population

Convenience sample

Participants chosen because they are easy to reach (e.g., college students); less representative

Generalize

Applying sample results to the larger population

Experimental group

Group that receives the manipulation (treatment/IV)

Control group

Group that does not receive the treatment; baseline for comparison

Random assignment

Every participant has an equal chance of being placed in any group; controls for pre-existing differences

Confounding variable (confound)

Uncontrolled outside factor that affects results, making it unclear whether the IV caused the change

Expectancy effects

When participant or researcher expectations influence results

Experimenter bias

Researcher's expectations skew the results

Placebo effect

Expectations or beliefs alone produce a change in experience

Placebo

Inactive substance or fake treatment given to the control group

Single-blind study

Participants don't know their group; researcher does

Double-blind study

Neither participants nor researchers know group assignments; guards against expectancy effects

Descriptive statistics

Numbers that summarize and describe data

Measures of central tendency

Describe the typical or center score (mean, median, mode)

Mean

Arithmetic average; most affected by outliers

Median

Middle score when scores are ordered; better with outliers/skewed data

Mode

Most frequent score

Measures of variability

Describe how spread out scores are (range, standard deviation)

Range

Highest score minus lowest score

Standard deviation

Average distance of scores from the mean; larger = more spread

Inferential statistics

Determine whether results can be generalized beyond the sample

Statistical analysis

Determines how likely any difference between groups is due to chance

Statistically significant

Result unlikely to be due to chance (usually p < .05, less than 5% chance)

Validity

Accuracy; does it measure what it is supposed to measure

Reliability

Consistency; gives the same results each time

Replicate

Repeating a study with different samples to check reliability

Peer review

Other experts evaluate a study's quality before publication

Institutional Review Board (IRB)

Committee that reviews and approves research with human participants

Informed consent

Participants told what to expect, risks, and implications, then agree to take part; can withdraw anytime

Deception

Purposely misleading participants to keep the study valid; must be justified

Debriefing

Telling participants the full truth (including any deception) after the study

Confidentiality

Participants' identities and data kept private

Institutional Animal Care and Use Committee (IACUC)

Committee that reviews research involving animals