POL222 - week 6, 7, and 8

3 types of experiments

Laboratory experiments - Conducted in controlled settings, often with convenience samples

Field experiments - Conducted in real-world environments, participants are real individuals, treatments are real interventions, DV measures real responses

Survey experiments - Embedded within surveys, using random treatment assignments in questions, randomly sampled and sample of convenience

Internal Validity

Confidence that the relationship observed between X and Y is truly causal

Strengthened by random assignment, hidden intention, well-designed stimulus

Weakened by participants knowing the purpose of the experiment, poorly isolated treatments

External Validity

Confidence that the causal relationship can be generalized beyond the study

Strengthened by representative samples and realistic stimuli

Weakened by artificial setups and non-representative participants (e.g., university students)

Experimental Stimulus

operationalization of the independent variable (treatment or manipulation).

A good stimulus isolates the causal effect, while a poor one can introduce bias or reduce realism

Experimental Intention

whether participants know the purpose of the experiment.

If they do, they may alter behavior (to please or sabotage the researcher), reducing internal validity

hiding experimental intentional can cause ethical concerns because Deception can violate informed consent

Meta Analysis

quantitative synthesis of previous studies on the same topic

It standardizes effect sizes, adjusts for differences (e.g., sample size, uncertainty), and computes an overall effect estimate

Systematic Review

Qualitative or mixed-method synthesis summarizing all evidence comprehensively

Weakness of Field Experiments

Ethical difficulties, sometimes have weaker control and risk of non-compliance

How to address non-compliance

using instrumental variable analysis (IV) to estimate the treatment effect among those who actually received the intervention

Clientelism

A political exchange where material favors are offered to individuals or groups in return for votes

Programmatic Policy

Politicians seek votes by offering broad benefits, like policy programs or public goods, available to all eligible citizens

Social Desirability Bias

When people underreport socially undesirable behaviors or attitudes in surveys (e.g., prejudice, corruption, discrimination)

→ People give socially desirable, not truthful, responses

Experiments help overcome this bias by indirectly measuring true attitudes or behaviors

Intergroup Contact Theory

Intergroup/social contact interventions promote cooperation between groups under equal, cooperative conditions

Expected effects:

↑ Knowledge, ↓ Anxiety, ↑ Empathy

↓ Out-group prejudice and discrimination