CLASS MATERIALS

What are the differences between types of probability and nonprobability samples?

Probability sampling has randomization as every element has an equal chance to be selected into the sample for generalizability, and is used with a sampling frame

Non-probability sampling has no randomization, is common in social sciences, and is used when no population list exists. 

Probability sampling

has randomization as every element has an equal chance to be selected into the sample for generalizability, and is used with a sampling frame

Non-probability sampling

has no randomization, is common in social sciences, and is used when no population list exists

Simple random sampling (Pros & Cons)

Pros: 

• Controls selection/systematic bias

• Sample is comparable to population 

Cons:  

• Less efficient 

• Weak sampling frame 

Cluster sampling (Pros & Cons)

Pros: 

• Useful when a sampling frame is not available

Cons: 

Quota Sampling (Pros & Cons)

Pros: 

Cons: 

• Does not involve randomization

Benefits of a pilot study

pre- and posttests

testing out the study to refine it?

Issues present with quasi-experimental designs

• No random assignment

• Selection bias

• Selection-maturation

Threats to internal validity (CHEST MIST)

CHEST

• Contamination

• History

• Experimental Mortality

• Selection bias

• Testing

MIST

• Maturation

• Instrumentation

• Statistical regression

• Treatment misidentification

Contamination

Awareness of group differences; demoralization

History

Events that happen outside the experiment; uncommon

EX: The Great Depression may have influenced productivity level identified by researchers of the Hawthorne studies; civil rights movement

Experimental Mortality (Attrition)

Participants who drop out are systematically different from those who remain

Selection bias

When participants are chosen in a way that leads to a non-representative sample, either by voluntary participation, their inherent characteristics, or method of data

Strive for a diverse and representative sample, ensure random sampling, and consider participant eligibility.

Testing

Performance is influenced by prior exposure to the same or similar tests due to familiarity

Maturation

Natural changes can occur independent of any intervention; pronounced in longitudinal studies

EX: long-term studies on drug rehabilitation; police reform

Instrumentation

Changes in measurement tools (questions, scales, collection method)

Statistical regression

Outliers

Treatment misidentification

Treatment/ intervention studies is incorrectly defined, applied, or identified

EX: Stanford prison experiment: Guards exaggerated their behaviors, with Zimbardo too involved; Hawthorne effect; Clever Hans, drug dogs

Know the placement of the pre- and post-test and treatment (X)