Experimental Design and Causality in Research

Experiment

A structured investigation to establish causality.

Causality

Relationship where one variable influences another.

Independent Variable

Variable manipulated to observe effects on dependent variable.

Dependent Variable

Outcome measured to assess impact of independent variable.

Crowding

Condition of having many individuals in a space.

Confound

Extraneous variable that may affect results.

Internal Validity

Degree to which results are attributed to independent variable.

External Validity

Extent to generalize findings to broader populations.

Between-Subjects Design

Different subjects experience different conditions.

Within-Subjects Design

Same subjects experience all conditions.

Random Assignment

Randomly placing subjects into different conditions.

Matched Pairs

Subjects paired based on similar characteristics.

Counterbalancing

Method to control order effects in within-subjects design.

Operational Definition

Specific implementation of a variable for replication.

Sampling

Process of selecting subjects from a population.

Statistical Power

Ability to detect an effect if it exists.

Alpha Level

Threshold for determining statistical significance.

Statistical Test

Method used to analyze data and draw conclusions.

Cognitive Performance

Ability to process information and respond.

Population Sample

Subset of a population used for research.

Experimental Control

Maintaining conditions to isolate effects of independent variable.

Memory Test

Assessment measuring recall and recognition abilities.

Two-tailed test

Tests for differences in both directions.

Between-subjects design

Groups compared are independent from each other.

Independent samples t-test

Compares means from different groups.

Dependent samples t-test

Compares means from the same group.

Random assignment

Each participant has equal chance for conditions.

Pretest-posttest design

Measures before and after treatment.

Matched pairs design

Subjects matched on key characteristics.

Order effects

Performance influenced by condition sequence.

Practice effect

Improvement due to repeated exposure.

Fatigue effect

Decline in performance over time.

Contrast effect

Comparison alters response to conditions.

Complete counterbalancing

All possible condition orders tested.

Solomon four-group design

Evaluates pretest effects with four groups.

Experimental condition

Group receiving the treatment or intervention.

Control condition

Group not receiving the treatment.

Dependent variable (DV)

Outcome measured in an experiment.

Independent variable (IV)

Factor manipulated in an experiment.

Sample

Subset of population for study.

Population

Entire group from which samples are drawn.

Disguised pretest

Pretest hidden to prevent behavior change.

Random sampling

Selecting participants randomly from population.

Percent judged overweight

Outcome measure for body image study.

Complete counterbalancing

All possible orders are tested in experiments.

Partial counterbalancing

Random order assigned to each participant.

Latin square

Each condition appears in every position once.

Order effects

Influence of sequence on experimental outcomes.

Minimum N

Minimum subjects required for each condition.

Conditions

Different treatment variations in an experiment.

Orders

Different sequences of presenting conditions.

Algorithm for orders

Systematic method to generate condition sequences.

Even number of conditions

Specific arrangements for conditions with even counts.

Odd number of conditions

Specific arrangements for conditions with odd counts.

Test subjects

Participants needed for valid experimental results.

6 orders

Total sequences for 3 conditions in complete counterbalancing.

24 orders

Total sequences for 4 conditions in complete counterbalancing.

120 orders

Total sequences for 5 conditions in complete counterbalancing.

Random order effects

Ensures balanced treatment conditions across participants.

Condition appearance

Each condition must appear in various positions.

Treatment conditions

Different experimental setups to compare effects.

Control condition

Baseline group for comparison against experimental group.

Between-subjects design

Different participants for each condition.

Within-subjects design

Same participants across all conditions.

Research question influence

Design choice dictated by the study's focus.

Cozby and Rawn (2016)

Reference for Latin square instructions.

Subjects multiple of 6

Necessary for valid complete counterbalancing.

Subjects multiple of 24

Necessary for valid complete counterbalancing with 4 conditions.