1 / 13

Factor Analysis I

Factor Analysis I. Principle Components Analysis. “Data Reduction”. Purpose of factor analysis is to determine a minimum number of “factors” or components that can explain a maximum amount of variance in a set of survey items.

Télécharger la présentation

Factor Analysis I

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Factor Analysis I Principle Components Analysis

  2. “Data Reduction” • Purpose of factor analysis is to determine a minimum number of “factors” or components that can explain a maximum amount of variance in a set of survey items. • In the absence of a conceptual model, the process can become entirely “machine driven” and uninteresting.

  3. Factors • X1 an observable response to a survey item. • Fs are unobservable factors determined from shared variance found with in the sample and the n items. • U is the unique variance attributable only to that survey item. • A factor analysis of n items will generate at least n factors.

  4. “Communalities” show the proportion of variance in each items explained by the factors (or shared with other items). • Extraction of .862 shows that response to ‘Count Per Box’ shows that 86.2% of the overall variance in the item can be explained by the principle components (or factors).

  5. Eigenvalues will sum to n, the number of items. • Three components have eigenvalues over 1.00, indicating they account for more an “expected amount” for completely independent items. • Subtotals of eigenvalues/n =% of Variance explained.

  6. Component Matrix • “Default” output with each extraction technique, of an “unrotated” set of factors.

  7. “Rotation” • You’re attempting to produce a set of factor loadings to fit your conceptual presentation of the factors. • Do you believe the factors to be completely independent or orthogonal: Varimax. • Do your believe that responses could indicate shared or correlated traits: Oblimim (“Oblique”)

  8. Rotated Solution: Varimax • New eigenvalues and % of variance explained—you have created new factors to meet your assumption of the rotation. • Here is how factor analysis sometime is treated with skepticism befitting paranormal psychic phenomena, such as “voodoo.”

  9. Rotated Solution “Oblimin” • Cannot make claims about % of variance explained, total of eigenvalues exceeds n.

  10. Varimax rotation Oblimin rotation “New component matrices”

  11. Additional “Oblimin” Output

  12. Two Major Extraction Methods • Principle Components—best for data reduction • Common Factor, or Principal Axis Factoring • SPSS includes: • Unweighted least squares • Generalized least squares • Maximum likelihood • Alpha factoring • Image factoring

  13. Farm Credit Services Results • The 17 questionnaire items were originally designed to capture unique dimensions of satisfaction a customer with a lender. • What does the explained variance of a factor analysis illustrate?

More Related