Factor Analysis is Your Friend. AnnMaria De Mars, PhD. The Julia Group & 7 Genera>on Games

Size: px
Start display at page:

Download "Factor Analysis is Your Friend. AnnMaria De Mars, PhD. The Julia Group & 7 Genera>on Games"

Transcription

1 Factor Analysis is Your Friend AnnMaria De Mars, PhD. The Julia Group & 7 Genera>on Games

2 WHY? Imagine this

3 What exactly were you planning on doing with that? Let s say you have a massive pile of data

4 You Could: 9% of adolescents reported blah blah blah 23% of adults said blah blah blah

5 The Problem: 1. Boring! No one is going to read each one. 2. Cannot conduct rela=onal analysis of each variable sta=s=cal sin 3. Individual items are notoriously unreliable.

6 SUBSCALES? Do you own guinea pigs? Do you have any stuffed animals?

7 Don t Be Scared of Factor Analysis! Conceptually, it s prejy simple Image Source: scojyoungpsalm37.blogspot.com

8 Factor analysis is for 1. Revealing pajerns of interrela=onships among variables 2. Detec=ng clusters of variables 3. Reducing a large number of variables to a smaller number of variables, the factors of factor analysis.

9 You can factor analyze anything test scores, individual items on a test measurements of various dimensions (i.e. height or weight) agricultural measures like yield of a rice field Socioeconomic measres

10 Is this a fair test? You have studied for a final exam in Biology 101. There is one ques=on, What is the rela=onship between respira=on and photosynthesis? Your child is in fi[h grade. Her weekly spelling test consists of one word.

11 You already understand this True Variance vs. Error Variance Variability & Validity

12 Ques>ons Image Source:

13 How do I Decide on the number of factors Interpret factors With SAS Enterprise Guide

14 How to do Factor Analysis using SAS Enterprise Guide

15 A Brief Overview of the Process 1. Open a data set, run a factor analysis, and observe the data s fit. 2. If necessary, run a correla=on analysis to create a dataset to analyze 3. If necessary, make modifica=ons and run your model once or twice more

16 A complete project

17 Our data From the 500 Family Study Hundreds of ques=ons answered Example uses 42 items asked adolescents regarding parent communica=on, rules, decision- making

18 FILE > OPEN > DATA

19 Select the variables Hold down shi= key to select more than one at a Ame

20 TASKS > MULTIVARIATE > FACTOR ANALYSIS

21 Look at your log first! When you get your results, do NOT look at your results first. Be smarter than most people and look at your log. To do that you click on the tab that says LOG

22 What if you see this? WARNING: 123 OF 465 OBSERVATIONS IN DATA SET WORK.SORTTEMTABLESORTED OMITTED DUE TO MISSING VALUES. If we didn t have a lot of people missing data, we could skip the next few steps, but hey, that s life.

23 Tasks > Describe > Summary sta=s=cs Drag and drop to select variables to analyze

24 Why factor analyze the correla>on matrix? The default for SAS is to delete a record if it is missing ANY of the variables. My first analysis was missing 120 records but no single item was missing for more than 49 people

25 Select the variables you want by clicking on them and pressing the blue arrows in between the panes

26 Step Two: Create a Correla>on Matrix Dataset DATA SET > TASKS > MULTIVARIATE > CORRELATIONS.

27 Output matrix as SAS dataset of type=corr: OUTPUT DATA > SAVE OUTPUT > DATA > RUN

28

29 TASKS > MULTIVARIATE > FACTOR ANALYSIS

30 Select the variables Hold down shi= key to select more than one at a Ame

31

32 So Now What does all of this Factor Analysis crap mean, anyway?

33 What exactly is a factor? Recall, a factor is some underlying trait that is measured indirectly by the items you measured directly. AKA dimension reduc@on technique

34 How many factors? 3 possibili=es

35 Eigenvalue The amount of variance in the individual measures explained by the factor. Square the loadings in the factor pajern and add them up. The total is the eigenvalue. Predic=on: At least one person who reads this will do exactly that and be surprised that I am right. Contrary to appearances, I do not make this s*** up.

36 Eigenvalue Common criterion for deciding the number of factors is Minimum eigenvalue greater than 1. Makes sense but

37

38 Method 3 Parallel analysis criterion (macros available Google it)

39 Now do it again

40 What do the factors mean?

41 Important Point One The correla=on of a variable with a factor is called the loading. loadings can be posi=ve or nega=ve.

42 Important point two To ease interpreta=on we d really like to have simple structure variables load close to 1.0 on one factor and close to zero on the others.

43

44 No, it is not party in the pool, sorry. Next step: iterate

45 Secret to factor analysis Finding a solu=on that is defensible BOTH sta=s=cally and theore=cally Nice first step to structural equa=on modeling. In fact, it IS the first step.