Research Design: Theories and Hypothesis

• Research design covers theory vs. hypothesis, identifying designs like correlational, experimental, and descriptive studies, limitations to correlational studies, understanding moderators and mediators, and experiments

Theory

• Defined as a general and comprehensive idea about why variables are related
• Often described as a set of beliefs that explains/organizes facts
• Theories come from observations of the world, scientific articles, etc.
• Social Comparison Theory stipulates people compare themselves to others to evaluate themselves

Social Comparison Theory

• Upward social comparison involves comparing oneself to those perceived as better, which can motivate improvement, but also hurt self-esteem
• Downward social comparison involves comparing oneself to those perceived as worse, which can improve self-esteem

Hypothesis

• A specific, but tentative statement, subject to empirical testing.
• Often presents as an "If/then" statement about expected relationships between variables
• Can be supported or falsified

Research Designs

• Correlational studies measure two variables (predictor & criterion) to see how they covary; each variable is measured, but nothing is manipulated
• Experimental designs manipulate an independent variable and measure its effect on a dependent variable
• Descriptive studies examine only one variable without making an association claim

Key Terms

• Independent variable is what the experimenter manipulates
• Dependent variable is what the experimenter measures
• In correlational studies, the predictor variable correlates with the criterion variable

Descriptive Studies

• Aims to describe the way things are in order to report conditions
• Uses descriptive statistics such as frequencies, proportions, means, and standard deviations
• Examples include how much time teens spend on social media, the percentage of social media comparisons that are upward versus downward, and the average self-esteem of teens

Correlational Studies

• Measuring self-reported time spent on social media and self-esteem
• Correlation does not imply causation

Establishing Causation

• Covariance assesses if the variables are correlated
• Temporal precedence determines if one variable precedes the other in time
• Internal validity ensures the predictor variable alone leads to a change in the criterion variable.

Temporal Precedence

• Requires establishing the direction of causality, which can be achieved through a cross-lag design
• This design helps establish temporal precedence only when a statistically significant correlation is observed when social media use occurs before self esteem
• Same participants are tested for the correlation between variables at Time 1 and again at Time 2

Threats to Internal Validity

• Third variables can independently relate to both X and Y, making it seem like X and Y are related when they aren't
• Controlling for third variables statistically can reveal if a correlation remains significant, implying a genuine relationship
• If the correlation disappears when controlling for the variable, this suggests the third variable was the reason for the original correlation

Ruling Out Variables Statistically

• Achieved through multiple regression with the criterion variable as self-esteem
• Beta indicates the direction and strength of the relationship between the predictor and criterion variables
• Significance of each predictor is determined by controlling for all other predictors, such as time spent on social media which significantly predicts self-esteem (p = .025) even after controlling for possible third variables
• Multiple regression does not establish causation because temporal precedence cannot be confirmed and there may be uncontrolled 3rd variables

Explore research design, theory, and hypothesis with social comparison theory. Understand upward and downward social comparison and their impacts on self-esteem. Learn to differentiate among correlational, experimental, and descriptive studies.