Chapters10 Up to factor analyses 20

1
Chapters10 (Up to factor analyses) /20

• Class 4

2
Quantitative Descriptive Studies

• Survey Research
• Data Reduction
• Correlation research

3
Survey Research Design Issues

• Matching the survey design to the researchers
  questions
• Defining the sample
• Quality of survey instrument(s)
• Data Analyses

4
Classification or Data Reduction Research Designs

• Factor Analysis and Cluster Analyses
• Purpose Organize data by categorizing it into
  simpler underlying structures or subgroups
• Assumption There are several latent
  constructs that account for the relationship
  among the many variables examined.

5
Factor Analysis

• It examines the inter-relation among a large
  number of items to identify sub-groups that
  correspond to underlying psychological dimensions
  
• Exploratory
• No a priori specification of specific dimensions
  or factors
• Confirmatory
• Researcher specifies a priori number of factors
  and items that will be associated with each
  factor.

6
Exploratory FA Procedures

• Identify theoretical construct(s)
• Select items to assess each construct
• Select type of exploratory factor analyses
  (principal component widely used)
• Decide on number of factors to extract (theory,
  scree test, eigenvalues)
• Factor rotation method (to distribute variance
  among factors) orthogonal or oblique

7
Exploratory FA Procedures

• Name the factors Examine factor loadings to
  determine factors associated with each factor
• simple structure
• conceptually meaningful set of items
• psychometric properties of factor scores
• Conduct confirmatory factor analyses with
  different sample to verify stability of factor
  structure

8
Eigen values variances extracted by the factors
9
Scree Plot
10
Communalities/Factor Loadings

• Communalities
• Proportion of each item's variance that can be
  explained by the factors. It is also noted as h2
  and can be defined as the sum of squared factor
  loadings for each item.
• Factor loadings
• The correlations between the item and the factor
  - possible values range from -1 to 1. 

11
Table 1 CDS Item Number and Factor Loadings
Item number Factor 1
Factor 2 Factor 3
Factor 4 Lack
of Lack of Approach-
Perceived
Structure Support
Approach Barriers


Note. Two items, 11 and 17, were not
assigned to any factor because they had high
loadings on two or more factors.
12
Communality

• Communality of Item 13
• .792 .132 .132 .042 .6595
• The four factors explain 66 of the variance of
  item 13
• Items with low communalities do not work well
  with the factor structure

13
Quality Items

• Deal with only one central thought
• Are precise
• Are brief
• Avoid awkward expressions
• Avoid irrelevant information
• Use positive language
• Avoid double negatives
• Avoid terms like All or None
• Avoid indeterminate terms like frequently or
  sometimes

14
Effect size d

• d M1 - M2 / sd
• of one sample if variances are
  homogeneous
• d M1 - M2 / pooled sd
• Sd pooled v (s1² s2²) / 2
• the square root of the average of the
  squared standard deviations.