You are on page 1of 20

FACTOR ANALYSIS

Introduction
Factor Analysis is an interdependent technique hence the basic purpose is to reduce the collective information. Interdependent techniques need no dependent

variable and no independent variable, all the


variables in the technique are interdependent.

An interdependent variable is the variable that


depends both variables and is independent of others as well.

General Linear equation of any interdependent technique or analysis can be like this 1.X + 2.Y = 5000 Where, n = Coefficient of nth variable. X and Y = Variables (These variables are neither dependent nor independent, they are interdependent).

Factor Analysis
Factor Analysis is primarily used for data reduction or structure detection.

1.

The purpose of data reduction is to remove redundant (highly correlated) variables from the data file, perhaps replacing the entire data file with a smaller number of uncorrelated variables.

2.

The purpose of structure detection is to examine the


underlying (or latent) relationships between the

variables.

Requirements
1. The researcher should be sure to include at

least 5 variables.

2. Sample size must be at least 50 observations,


no factor analysis can be done if sample < 50.

preferably sample size is 100 or more


observation/cases.

3. Each variable must at least have 10 cases that is 10:1, some suggest 5:1.

4. The sample must have more observations than variables.

5. The variables should be quantitative at the interval or ratio level. Categorical data (such as religion or country of origin) are not suitable for

factor

analysis.

Data

for
can

which

Pearson
be

correlation

coefficients

sensibly

calculated should be suitable for factor analysis.

6. The data should have a Bivariate normal distribution for each pair of variables, and observations should be independent.

7. The factor analysis model specifies that variables are determined by common factors (the factors estimated by the model) and unique factors (which do not overlap between observed variables).

8. The computed estimates are based on the


assumption that all unique factors are

uncorrelated with each other and with the


common factors.

Assumptions
1. A basic assumption of factor analysis is that some

underlying structure does exist in the set of selected


variables. 2. The sample is homogeneous with respect to underlying factor analysis. 3. Variables should be interrelated.

4. The researcher must ensure that the data matrix has


sufficient correlations to justify the application of factor analysis.

Factor Analysis Methods


For Data Reduction:

The principal components method of extraction begins by


finding a linear combination of variables (a component) that accounts for as much variation in the original variables as possible. It then finds another component that accounts for as much of the remaining variation as possible and is

uncorrelated with the previous component, continuing in


this way until there are as many components as original variables.

Usually, a few components will account for most of the variation, and these components can be used to replace the original variables. This method is most often used to reduce the number of variables in the data file.

Factor Analysis Example

Factor Analysis Methods (Contd)


For Structure Detection.

Other Factor Analysis extraction methods go one step further by


adding the assumption that some of the variability in the data cannot be explained by the components (usually called factors in other extraction methods). As a result, the total variance explained by the solution is smaller; however, the addition of this structure to the factor model makes these methods ideal for examining relationships between the variables.

Key Terms
1. Anti-Image Correlation Matrix: It is the matrix representing the extent to which factors explain each other. It also shows the MSA value for each variable. 2. Bartletts test of sphericity: It is statistical test for the overall significance of all correlations among the variables within the correlation matrix. (Ho: the correlation matrix is identical

matrix, therefore Ho should be rejected, i.e. the sig value


should be <0.05).

3. KMOs measure: It is the statistical measure that tests whether partial correlations among variables are small. It basically tells us about the feasibility of the factor analysis. (Ho; the partial correlations among the variables are sufficient for the applying factor analysis, therefore it should be accepted). For KMO sig value must be > 0.5.

4. Cluster Analysis: An interdependent technique with the objective of


grouping variables based on their similar characteristics. Similar to Q-Factor Analysis.

5. Common Variance: Variance shared with other variables in the factor analysis.

6. Communality: Total amount of variance that


an original variable shares with all other

variables.
7. Correlation Matrix: Table showing

interrelationships among variables.

Key Terms (contd)


8. Dummy Variable: A binary metric variable that is used to represent the single category of non-metric variable. 9. Eigen Value: Column sum of square loading for factors. It represents the variance accounted for by factor. Also known as Latent Root. 10. 11. Factor: A set of variables that are highly inter-correlated. Factor Loadings: The correlation between the original variables and the factors, and the key to understanding the nature of a particular factor. 12. 13. Factor Matrix: Table showing factor loadings of all variables for each factor. Factor rotation: Process of manipulation or adjusting the factor axis to achieve more simple and pragmatically more meaningful factor solution. 14. Measure of sampling adequacy: It is a measure of appropriateness to apply factor analysis for both individual and entire correlation. Value above 0.5 is appropriate. 15. Measure Error: Inaccuracy occurs due to error in measurement scale, data entry error, etc.

Key Terms (contd)


16. 17. Multicollinearity: Extent to which a variable is explained by other variable in the analysis. R-Factor Analysis: Analyzing the relationship among the variables to identify the latent dimensions (Factors). 18. Specific Variance: Variance of each variable unique to that variable and not explained or associated with other variables in the factor analysis. 19. Summated Scales/Composite Analysis: Method of combining several variables that measure the same concept into a single variable in an attempt to increase the reliability of the measurement. 20. Orthogonal Factor Rotation: Factor rotation in which the factors are extracted so that their axes are maintained at 90degrees. 21. Oblique Factor Rotation: Factor rotation in which the factors are extracted without maintaining their axes at 90degrees. 22. Common Factor Analysis: Factor model in which factors are based on reduced correlation matrix.

23.

Component Factor Analysis: Factor model in which factors are based on total variance.

Key Terms (contd)


24. Factor Score/Absolute Value: Composite measure/summated scale created for each observation on each factor extracted in factor analysis. 25. 26. Factor Structure Matrix: Table showing the correlations between variables and factors. Surrogate Variable: Selection of single variable with highest factor loading to represent factor in data reduction stage. 27. 28. 29. 30. P-value: It is the probability of calculated statistical test or calculated significance level of the test. Type-1 Error: Rejection of Null hypothesis when it is true (). Type-2 Error: Acceptance of Null hypothesis when it is false (). Varimax Method: An orthogonal rotation method that minimizes the number of variables that have high loadings on each factor. This method simplifies the interpretation of the factors. 31. Quartimax Method: An orthogonal rotation method that minimizes the number of factors needed to

explain each variable. This method simplifies the interpretation of the observed variables.
32. Equimax Method: An orthogonal rotation method that is a combination of the varimax method, which simplifies the factors, and the quartimax method, which simplifies the variables.

Practical example
SPSS review Path to open SPSS and to reach target file. Car sales example. Understanding the data description. Measurement scales. Analyze----Dimension reduction-----FA. Data selection. Option selection.

You might also like