Correlation vs. Causation in Social Science

The Marshmallow Test is an experiment used to measure a child's ability to delay gratification. A child is given a marshmallow and told they will receive a second one if they wait a certain amount of time before eating the first one.

The Marshmallow test is believed to be a measure of self-control, which can predict success in later life - better academic performance, better social skills and well-being.

The original Marshmallow Test studies focused on a small group of children, primarily from affluent families in Stanford.

A correlation exists when two or more things occur together. But, it doesn't mean one causes the other.

Causation means one thing directly influences another. Cause and effect.

The Marshmallow Test shows a correlation between delaying gratification and later academic success, BUT it doesn't prove causation.

Later studies show that factors like parental affluence could also affect academic success, not just the ability to delay gratification.

When evaluating research, careful consideration of potential alternative explanations is crucial for drawing accurate conclusions.

A factor or variable that is not included in the study but can influence the relationship between the independent and dependent variables.

Two or more explanatory variables are intertwined, making it unclear which variable is responsible for the observed effect.

The cause must occur before the effect.

Changes in the independent variable are accompanied by changes in the dependent variable.

A justifiable explanation for why two variables are related.

Establishing that the independent variable, and ONLY the independent variable, caused changes in the dependent variable; rival explanations must be ruled out.

The independent variable is directly manipulated to observe its effect on the dependent variable.

A study design that minimizes bias and ensures that the results are due to the independent variable, not other factors.

A situation where two or more factors influencing a response variable are intertwined, making it unclear how much each contributes.

Occurs when the effects of two or more variables on a response variable cannot be distinguished from each other.

Repeating an experiment under the same conditions to check for consistency in the results.

Obtaining similar results in different studies, even with different researchers, methods, or locations.

Obtaining similar results when different teams use the same input data, analysis steps, and conditions.

Ensuring research can be trusted by replicating the experiment with different teams or methods.

Ensuring data is collected and analyzed according to scientific principles, minimizing bias and promoting reliable results.

A systematic approach to acquiring knowledge based on empirical evidence and logical reasoning.

Systematic errors in thinking or perception that influence how we process information, leading to inaccurate conclusions.

A bias that favors information confirming pre-existing beliefs, while ignoring or downplaying contradictory evidence.

The tendency to focus on successful cases while neglecting failed ones, drawing misleading conclusions from incomplete data.

The trustworthiness and validity of research findings, ensuring they accurately reflect the phenomenon being studied.

The tendency to be influenced by factors (e.g., personal beliefs, expectations) that are not directly related to the research question.

The process of critically evaluating scientific claims and findings to ensure their validity and reliability.