Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
PDAM #9
| Question | Answer |
|---|---|
| What are the 3 Assumptions for Factor Analysis | 1, Sampling Adequacy = do variables share enough variance 2, Sphericity = isnt correlation matrix and identity matrix 3, Multicollinearity = the correlation should be unperfect |
| What is a identity matrix | matrix with 1 in the diagonal otherwise with zeros only |
| How can we test sampling adequacy (3) | 1, with Kaiser-Meyer-Olkin (KMO) 2, check if the variables are suitable for Factor Analysis because they share enough common variance 3, give index between 0- not suitable at all and 1-totally suitable SHOULD BE AT LEAST =0.5 |
| How to test Sphericity (3) | 1, Bartlett´s Test of Sphericity 2, compare covariance matrix to identity matrix 3, H0: Cov.Mat. = Ide.Mat H1: Cov.Mat ≠ Ide.Mat |
| How to test multicollinarity (2) | 1, we calculate the determinant of the correlation matrix 2, the determinant should be higher that 0.00001 |
| What are two ways to determine the appropriate number of factors | 1, Eigenvalues - Kaiser´s criterion 2, Scree plot - Cattel´s criterion |
| How to determine number of factors with Scree Plot (3) | 1, we build a scatterplot with eigenvalues for number of factors 2, we find the point of inflexion 3, Cattel´s Criterion = Point of inflexion - 1 |
| How to determine number of factors with Kaiser´s criterion (3) | 1, we create a matrix with factors and their eigenvalue 2, Eigenvalue = "How many unit of variables does this factor explain" 3, Factors with Eigenvalue over 1 should be retained |
| When to use Kaiser´s criterion over Scree Plot (4) | 1, less than 30 variables 2, sample size more than 250 3, communatilies after extraction are less than .7 for every variable 4, mean communality (across all variables) above 0.6 |
| What is rotating the factors | we change the coordinate system NOT DATA to make it easier to overview (like the Vienna Subway Line model) |
| What are two ways to rotate the axes | 1, Orthogonal rotation = factors remain perpendicular 2, Oblique rotation = factors can move closer or away from others |
| What is idea and most common method of A, Orthogonal rotation B,Oblique rotation | A, Orthogonal rotation Idea = treats factors as uncorrelated Methods = Varimax B,Oblique rotation Idea= treats factors as correlated Methods = Direct oblimin |
| What is factor allocation (3) | 1, we allocate the variable to factor with highest shared loading / correlation 2, factor loadings of at least 0.5 or 0.4 are substantial 3, the lower the loading the higher sample we need for significance |
| How to measure factor consistency (2) | 1, Cronbach´s Alpha 2, more that 0.7 is acceptable |
| How many samples do we need for a Factor analysis | rule of thumb = bare minimum of 10-20 observation per variable |
| What is difference between variance and kurtosis | variance = amount of spread kurtosis = extremeness of tails relative to spread |
Created by:
semester1.IBWL