Intro to Marketing Research: Factor Analysis Exploratory Factor Analysis Purpose Data Reduction Detecting Underlying Structure Principal Components vs Common Factor

What It Is

Exploratory Factor Analysis (EFA) is a statistical technique used to identify the underlying structure of a large set of variables by reducing them to a smaller number of factors. This method helps researchers understand the relationships between variables and identify patterns that may not be immediately apparent. A classic example of EFA in marketing research is the study by Kaiser (1960) on the factor analysis of consumer attitudes towards different brands of soap. By applying EFA, researchers can identify the underlying dimensions of consumer preferences, which can inform marketing strategies and product development.

Key Terms & Concepts

• Exploratory Factor Analysis (EFA): A statistical technique used to identify the underlying structure of a large set of variables by reducing them to a smaller number of factors.
• Factor: A common underlying dimension or theme that explains the correlations among a set of variables.
• Eigenvalue: A measure of the amount of variance explained by each factor (e.g., an eigenvalue of 1 indicates that the factor explains 100% of the variance).
• Factor loading: The correlation between a variable and a factor (e.g., a factor loading of 0.8 indicates a strong relationship between the variable and the factor).
• Communality: The proportion of variance in a variable that is explained by the factors (e.g., a communality of 0.6 indicates that 60% of the variance in the variable is explained by the factors).
• Kaiser's criterion: A rule of thumb for determining the number of factors to retain, which suggests retaining factors with eigenvalues greater than 1.
• Scree plot: A graphical representation of the eigenvalues, which can help researchers determine the number of factors to retain.
• Principal Components Analysis (PCA): A type of factor analysis that is similar to EFA, but does not assume that the factors are correlated.
• Common Factor Analysis (CFA): A type of factor analysis that assumes that the factors are correlated and that the variables are measured with error.
• Cronbach's alpha: A measure of the reliability of a set of variables (e.g., a Cronbach's alpha of 0.8 indicates high reliability).
• Sample size: The number of observations in the dataset (e.g., a sample size of 100 indicates a relatively small dataset).
• Assumptions of EFA: The variables should be measured on an interval or ratio scale, the data should be normally distributed, and the variables should be correlated.

Common Misunderstandings

• Misunderstanding: EFA is a type of descriptive research.
• Correction: EFA is a type of exploratory research, which aims to identify patterns and relationships in the data.
• Misunderstanding: PCA and EFA are the same thing.
• Correction: PCA is a type of factor analysis that is similar to EFA, but does not assume that the factors are correlated.
• Misunderstanding: EFA assumes that the factors are correlated.
• Correction: CFA assumes that the factors are correlated, but EFA does not make this assumption.

Quick Application / Identification

A marketing researcher wants to identify the underlying dimensions of consumer preferences for different brands of coffee. The researcher collects data on 10 variables, including price, taste, and packaging. Which statistical technique would be most appropriate for this study?

Answer: Exploratory Factor Analysis (EFA) Explanation: EFA is a statistical technique that can help the researcher identify the underlying dimensions of consumer preferences, which can inform marketing strategies and product development.

Last‑Minute Revision

• EFA is used to identify the underlying structure of a large set of variables.
• A factor is a common underlying dimension or theme that explains the correlations among a set of variables.
• Eigenvalue is a measure of the amount of variance explained by each factor.
• Factor loading is the correlation between a variable and a factor.
• Communality is the proportion of variance in a variable that is explained by the factors.
• Kaiser's criterion suggests retaining factors with eigenvalues greater than 1.
• Scree plot is a graphical representation of the eigenvalues.
• PCA is a type of factor analysis that is similar to EFA, but does not assume that the factors are correlated.
• CFA assumes that the factors are correlated and that the variables are measured with error.
• Cronbach's alpha is a measure of the reliability of a set of variables.
• Sample size is the number of observations in the dataset.
• EFA assumes that the variables are measured on an interval or ratio scale.
• EFA assumes that the data are normally distributed.
• EFA assumes that the variables are correlated.
• ⚠️ EFA is not suitable for small datasets (less than 100 observations).
• ⚠️ EFA is not suitable for variables that are not measured on an interval or ratio scale.