Quasi-Experimental and Correlational Designs

Flashcards on Quasi-Experimental and Correlational Research

Quasi-Experimental Design

Research where conditions cannot be manipulated or controlled, resembling real experiments but lacking essential elements like random assignment.

Subject Selection in Quasi-Experimental Design

Selecting subjects based on pre-existing conditions to compare behavioral differences or observe natural situations.

Natural Experiments

Experiments where the independent variable may be manipulated, but participants are not randomly assigned.

Internal Validity in Quasi-Experimental Designs

Validity that is typically low in quasi-experimental designs due to the lack of random assignment.

Ex Post Facto Studies

Investigates effects of subject variables (e.g., age, gender, trauma) without manipulation, grouping subjects based on pre-existing characteristics.

Non-Equivalent Groups Design

Compares effects of different treatments on pre-existing groups, requiring measurement of attributes that threaten validity to ensure comparability.

Nonequivalent Control Group Design

Similar to pretest-posttest control group design but without random assignment, where selection bias is a major concern.

Longitudinal Studies

Measures behavior over different time points in the same group, but is time-consuming and faces participant retention issues.

Cross-Sectional Studies

Compares groups already at different stages at a single point in time, faster than longitudinal studies but requiring more participants and having lower statistical power.

Pretest/Posttest Design

Measures behavior before and after a natural event or treatment, often used with comparison groups to improve validity but susceptible to practice effects.

Correlational Research

Non-experimental design to assess statistical relationships between variables without manipulation or control of confounds.

Uses of Correlational Research

Used to establish relationships among behaviors, predict behaviors, and show associations between antecedents and outcomes, but cannot establish cause-effect relationships.

Importance of Correlational Research

An initial step in inferential research; if there's no correlation, then no causation is possible, generating ideas for future experiments.

Descriptive Goal of Correlational Research

Describes relationships between variables, e.g., Does rim size correlate with narcissism scores?

Predictive Goal of Correlational Research

Predicts behavior based on relationships, e.g., If Jason scores high on a narcissism test, can we predict his rim size?

Simple Correlations

Relationship between pairs of scores from each subject, analyzed using the Pearson Product-Moment Correlation Coefficient (r).

Pearson Product-Moment Correlation Coefficient (r)

Ranges from -1.00 to +1.00, indicating positive, negative, or no relationship between variables.

Correlation vs. Causation

Correlational studies cannot determine the causal direction between variables.