This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: A Quick Introduction to Factor Analysis From Data Analysis: A Statistical Primer for Psychology Students by Edward L. Wike: ... [T]here are elaborate extensions of r to multivariate data ... another case is k variables (factor analysis) . Fortunately, for both the student and author, none of these extensions will be considered here. From Linear Algebra and its Applications by Gilbert Strang: ... [T]he technique is so much needed that, even starting as an unwelcome black sheep of multi- variate analysis, it has spread from psychology into biology and economics and the social sciences. From linear algebra: A “matrix” is a rectangular (in this discussion, usually square) array of numbers; e.g., 3 5 7 1 2 3 - 1 3 3 . 5 2 2- 1 We can add two matrices of the same dimensions, by adding corresponding entries; and we can multiply a matrix by a number, by multiplying all the entries in the matrix by that number: 1 2- 1 +- 2 1 2 3 =- 1 3 2 2 3 1 2- 1 = 3 6- 3 A square matrix is a “diagonal matrix” if the only nonzero entries (if any) are on the “main diagonal” (upper left to lower right): 5- 1 0 The “transpose” of a matrix is the (new) matrix obtained turning the rows into columns (and vice versa): 3 5 7 1 2 3 T = 3 1 5 2 7 3 - 1 3 3 . 5 2 2- 1 T = - 1 3 3 . 5 2 2- 1 If a (square) matrix A has the property that A T = A (as this one does), then A is called “symmet- ric”. This just amounts to saying that, for each i and j , the entry in row i and column j is equal to the entry in row j and column i , i.e., the matrix is symmetric about its main diagonal. Relevant example: Suppose we have several variables, the scores on 4 exams, one in math, one in physics, one in English, and one in history — one list of 4 scores for each student in a class. Taking the exams 1 in pairs, we can find a correlation coefficient for each pair. We then form a matrix with the rows and columns labelled M(ath), P(hysics), E(nglish) and H(istory); the entry in the X row and Y column is the correlation coefficient between X and Y. The result is a symmetric matrix (since the correlation between X and Y is the correlation between Y and X) with 1’s on the main diagonal (since the correlation between X and X is 1):...
View Full Document
This note was uploaded on 03/09/2008 for the course CALC 1,2,3 taught by Professor Varies during the Spring '08 term at Lehigh University .
- Spring '08