• Shuffle
    Toggle On
    Toggle Off
  • Alphabetize
    Toggle On
    Toggle Off
  • Front First
    Toggle On
    Toggle Off
  • Both Sides
    Toggle On
    Toggle Off
  • Read
    Toggle On
    Toggle Off
Reading...
Front

Card Range To Study

through

image

Play button

image

Play button

image

Progress

1/30

Click to flip

Use LEFT and RIGHT arrow keys to navigate between flashcards;

Use UP and DOWN arrow keys to flip the card;

H to show hint;

A reads text to speech;

30 Cards in this Set

  • Front
  • Back
Why can 5 measures of IQ give a more complete description and a more accurate concept of IQ?
Most constructs, such as IQ, Agression, and motivation are not fully described by a single measure.

Each measure may identify part of IQ that others are not including. Therefore, a composite is better than any single measure.

Redundant information can be combined into one variable.

To have greater chance of covering a construct, measure multiple times, there is less chance that part of the construct will be left out and you get broader coverage.
How does cluster analysis treat variables?
Each whole variable is assigned to a single cluster. Five variables get 5 clusters. Highly correlated variables are combined in the same column. Z Scores are averaged at the bottom and 5 different variables can be combined to 3 or even less.
How does cluster analysis help build a theory of intelligence?
Start with IQ (the concept you want to measure) as a circle. Then list variables that contribute to getting a measure of IQ, such as Verbal intelligence, Math intelligence, Social Intelligence. Combine Variables that are similar and highly correlated. The ones that are left over will give you an idea of the major components of intelligence.
Why is Cluster Analysis helpful in combining variables?
Each variable goes with its most highly correlated neighbor. Helps you decide if you are measuring many things or just a few. If verbal performance and social end up in the same cluster, IQ is very general. If not, you end up with descriptive facts about what intelligence is.
What is the difference between multivariate prediction methods and multivariate covariation analysis?
The prediction branch uses predictor variables and criterion variables and the multivariation covariation side uses multiple measures on each subject.
What is the purpose of multivariation covariation analysis?
Take multiple measures on subjects and see if each variable is measuring a different construct or if they redundantly cover the same territory.
In multivariation covariation analysis, what can be concluded if you measure 5 variables for each subject and the variables are highly correlated with each other?
The measures are all measuring the same underlying trait or ability.

Whatever quality makes you score high on one variable, makes you score high on another variable.
In multivariate covariation analysis, what can be concluded if variables 1, 2, and 3 are highly correlated with each other, but not correlated with variables 4 and 5, which do show correlation?
This means that 5 measurements are really testing only a couple of underlying constructs. Scores 1, 2, and 3 are 1 quality of the trait (IQ) and 4 and 5 are others (self esteem and social)
How does multivariate covariation analysis help simplify a theory?
Several measurements are taken. If most of the measurements are highly correlated, intelligence becomes 5 things instead of 20. Use the power to determine how many different constructs make up intelligence.
How can a psychological construct such as agression be measured effectively?
Most psychological constructs are best described by measures using multiple indicators. Measure agression by hitting, pushing, threatening, and insulting. Multiple indicators cover the construct more completely.

Then use multivariate covariation to combine the indicators into a single variate.
How does combining variables relate to sample size?
If you can use multivariate covariation analysis to combine 20 predictors down to 5, you can use a smaller sample and get more reliable results.

Algebraically collect redundant information in many variables to a small number of variates.

3 variates can predict a criterion as well as 20 redundant variables.

Gain: fewer subjects needed and greater reliability.
Cluster analysis, Principal components, and Factor Analysis treat variables differently in what way?
C.A. - groups variables together based on similarity. It never divides a variable into parts.

PC and FA distribute each variable across several variates.
What is the difference between how Cluster Analysis, Principal components, and Factor Analysis treat Variance?
CA and PC work on all the variance but FA focuses only on the redundant variance in the set.
Cluster Analysis treats variables as ______ and analyzes _____ variance.
wholes, total
Principle Components Analysis _______ variables across _______ and analyzes ______ variance
distributes
variates
total
Factor Analysis _______ variables across _______ and analyzes ________ variance.
distributes
variates
ONLY common
In multivariate covariation analysis, if each variable is highly correlated with each of the others that are measuring a construct, what can be concluded?
They are probably all reflecting teh same underlying quality or trait.

Even though you are using 5 measures, you are learning only one piece of information about the subjects.
T/F
Cluster Analysis divides a variable into parts
False
T/F
Principal Components and Factor Analysis both distribute each variable across several variates.
True
Why does Factor Analysis focus on only common or redundant variance?
Part of the measure may be related to something else such as social class or impulsivity. Leave it out so it doesn't influence the analysis and conclusion.
What is a principal component?
A principal component is NOT a simple variable. It is a variate made by combining parts of each variable.
What number do you get if you square a principal component and add across a row?
1.0
What number do you get if you square a principal component and add them down a column?
The part that it plays in the variance. Divide by #of variates to get %
What is special about the variates that are created from the parts of the variables in principal components?
They are Uncorrelated.
What is factor analysis?
Use Factor Analysis to reduce a large number of variables to a smaller set.

This allows for a better understanding of the dimensions underlying the initial variables.

The goal is to use the smallest number of explanatory concepts to explain the maximum amount of common variance in a correlation matrix.
What is the significance of the first Principal component?
It has the greatest variance.

It is the mathematically constructed variate that best represents the redundancies among the original variables.
What is done in Principal Components Analysis?
Convert correlated variables to noncorrelated variables and let the few largest stand in for the larger number of original variables.
What is multiple regression?
Use a set of predictors to predict a single criterion. Combine into a variate and compute Pearson Coefficient. See if predictor gives a close number to actual value.
What is a weakness of Multiple Regression?
Multiple Regression does not tell you how the predictors influence the criterion. Something else is needed to develop a model of how the predictors work together to influence the result.
What is Path Analysis and how does it complement Multiple Regression?
State in advance how the predictors work together to predict an outcome.

Parent education influence study time and notes, but they don't influence Parent Education. Multiple Regression helps in telling how much influence, but not how.