Research Methods - Interactions Pt1 - Factors and Levels

Understanding Research Terminology

Introduction to Research Factors

• The video aims to clarify common terminology used in research studies, particularly focusing on factors and their levels.

• Examples of research questions are presented, such as comparing the effectiveness of exercise versus cognitive behavioral therapy (CBT) on depression.

Types of Variables in Research

• The discussion highlights independent variables (IV) and dependent variables (DV), with examples illustrating how groups are compared based on these variables.

• It is noted that some setups involve genuine experiments with random assignments, while others may be correlational studies without manipulation.

Understanding Factors and Levels

• All discussed setups involve comparing two groups, indicating that the top variable is categorical with two categories.

• The term "factor" is introduced to describe the variable by which scores are grouped, such as treatment type or distraction level.

Exploring Factor Levels

• A factor can have multiple levels; for instance, car size can be categorized into compact, midsize, and full-size cars.

• When measuring factors with inherent ordering (like car sizes), they may be assessed on an ordinal scale while DVs could be measured on a ratio scale.

Complex Studies Involving Multiple Levels

• An example involving different medications for cholesterol treatment illustrates a single factor (pill type) with three levels: new drug, old drug, placebo.

• The concept of factors extends beyond three levels; researchers can study factors split into four or more levels affecting various outcomes like hunger level.

Understanding Factors and Levels in Research Design

Operationalizing Gender Variables

• The operationalization of gender in studies can include categories such as man, woman, non-binary, or genderqueer. This categorization defines the levels of the gender variable being studied.

Measuring Education Levels

• Education level can be operationalized with varying numbers of levels (e.g., five levels in one study versus nine in another), which reflects how researchers group participants based on their educational attainment.

Example: Oxygen Consumption Measurement

• In a study measuring oxygen consumption (V2 max), participants' scores are taken before and after a training regimen to assess its effectiveness. This creates two levels for the training factor: pre-training and post-training.

Between Subjects vs. Within Subjects Design

• A factor can be classified as between subjects (each participant experiences only one level) or within subjects (each participant experiences all levels). For example, comparing safety scores across car sizes is a between subjects design.

Statistical Analysis Techniques

• When analyzing data from a between subjects factor, an ANOVA (Analysis of Variance) is commonly used to compare three or more groups. It generalizes the T-Test for multiple comparisons.

• A within subjects ANOVA is utilized when each participant provides data across all conditions, allowing for repeated measures analysis. This contrasts with the independent measures T-Test that compares only two groups.

Conclusion on Statistical Measures

• Regardless of whether using a between or within subjects ANOVA, researchers calculate an F value to determine if there are significant differences among the levels of factors being studied.

Study Design: Single vs. Multiple Factors

Understanding Factorial Studies

• The discussion begins with the concept of a study involving a single independent variable, focusing on how variations in this factor can affect the dependent variable.

• It is emphasized that while studies can be conducted with one factor, it is also feasible to explore multiple factors simultaneously.

• The ability to test multiple independent variables allows researchers to analyze not only individual effects but also interactions between different levels of these factors.

• This approach provides insights into which combinations of factors yield significant differences in the dependent variable, enhancing the depth of analysis.

• Overall, factorial designs are highlighted as powerful tools for understanding complex relationships within data.