Class Notes (835,539)
Canada (509,225)
POL101Y1 (1,148)
Lecture 6

POL 322 week 6 lec

6 Pages
Unlock Document

Political Science
Michael Painter- Main

POL 322 week 6 lec Oct 22, 2012 Reliability and Factor Analysis types of scales and indexes -scales focus on hierarchy -indexes focus on addition -eventually take each individual variable and add together (super variable) -can change ordinal into interval-like -Brians et. al don’t really distinguish scales from indexes dimensional -fills concept -traditional index -elements in a broader concept and we are trying to tap into what they are hierarchical -pass thresholds -Guttman scales -may weigh responses -ex. political participation -scale -we do things in a particular realm because they are easier to do, then later on we may get more advanced ***Non-Association*** -when put indicators together that may not fit -ex. vulnerability -immigrants, poor, old -index -ex. social economic status -education, wealth, class etc. that may be tied but does not have to be a strong tie, they don’t have to go together but we are still interested in their relationship POL 322 week 6 lec Oct 22, 2012 -ex. measuring vulnerability -ex. just because you’re an immigrant doesn’t mean you’re old though we may still be considering both variables -we are forcing variables together -different from dimensional because it does not give strong results/ ties like dimensional Cronbach’s Alpha Score -test of reliability -do indicators measure concept consistently? -high association? -Cronbach’s score- 0 to 1 -closer to 1, more consistent are indicators, stronger pattern -every person who gets one right, gets all them right if the score is 1 -0 = no pattern at all -threshold- generally .60 -but not always utilized -suggests there is enough association between variables that they go together -hope all variables have ties to each other -do the components consistently measure a broader concept? --> are they reliable -nominal variables are bad in this case because they have no order Factor Analysis (aka Exploratory Factor Analysis) -examines dimensionality through creation of mathematical factors -based on how much factors explain variance of indicators -do some of the strong pattern variables go better together in the set than others in the strong pattern? ex. 2 out of 5 variables are always associated while the other 3 are only usually associated -those two variables may represent an even broader theme POL 322 week 6 lec Oct 22, 2012 -the different dimension (groups of strongly associated variables) are factors -if relationship between factors is very strong, they should explain the broad theory better than another of the individual variables -bunch of variables, test them, it tells us do they associate? if so then we get factors that tell us they go so well together that putting them together tells us so much more than knowing one of them on its own Selection of Variables -must be ordinal, interval/ ratio, dummy -must be set of variables that hold common concept -i.e. dimensional approach -we take variables that we think have some strong relationship because we think that they represent some valuable broader relationship Factor Extraction -2 types of factors: - principle factoring - principle components factoring (principal components analysis) -basic difference: PCF assumes that total variance explained by factors under investigation -PF doesn’t -PF more conservative estimate -we prefer these
More Less

Related notes for POL101Y1

Log In


Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.