PROPERTY OF BUSINESS MANAGEMENT SOCIETY
Simple linear only uses one predictor. • Summarize the data with fewer variables
that collectively explain most of the variation
A linear regression model postulates that in the data.
the relationship between response and • Use the principal components in other
predictor is linear. unsupervised learning methods such as
clustering.
Cross-validation is the technique used to
evaluate the predictive models by Exploratory Factor Analysis
partitioning the sample.
Exploratory factor analysis is a statistical
Exploratory factor analysis is a statistical technique that is used to reduce data to a
technique used to reduce data to a smaller smaller set of summary variables and to
data underlying theoretical. explore the underlying theoretical structure
of the phenomena. It is used to identify the
The goodness of fit measure is a general structure of the relationship between the
term that basically tries to summarise the variable and the respondent.
predictive strength of the model as a whole.
-is generally used to discover the factor
The typical measures are structure of a measure and to examine its
(i) Coefficient of determination (R-squared) internal reliability.
(ii) Standard error of estimate
(iii) simple regression model EFA is often recommended when
researchers have no hypotheses about the
DATA REDUCTION ANALYSES nature of the underlying factor structure of
their measure.
FA, EFA, PCA, or CFA
-All of them are interdependence methods
in which no single variable or group of Exploratory Factor Analysis
variables is defined as being independent or There are some fairly standard pieces of
dependent. The statistical procedure information to include:
involves the analysis of all variables in the
data set simultaneously so the goal of these 1.Theoretical underpinnings for the area you
interdependence procedures are to uncover are studying, and specifically for the
structure by the grouping of variables constructs that you are interested in
uncovering through EFA.
Principal Component Analysis
2. A description of the sample (e.g.
Principal Components Analysis (PCA) is demographic information, sample size,
a dimension-reduction technique used to sampling method).
reduce the number of variables without
losing important information.
Page 1 of 2
PROPERTY OF BUSINESS MANAGEMENT SOCIETY
3. A description of the type of data used measurement theory and then validate it
(e.g., nominal, continuous) and descriptive with CFA so the focus is on the "validity" of
statistics. our latent constructs.
4. Describe how you went about testing the In CFA a requirement is a prior selection of
assumptions for EFA. Details regarding variables on the basis of established theory
sphericity checks and measures of sampling and to hypothesize beforehand the number
adequacy should be reported. of factors in the model. We most commonly
use the CFA measurement model to
5. Explain what FA extraction method (e.g. validate multi-item constructs such as the
maximum likelihood) was used. items to measure a construct e.g.
Explain the criteria and process used for satisfaction.
deciding how many factors were extracted
in the final solution, and which items were
selected. Clearly explain the rationale for
key decisions during the EFA process.
6. Explain what rotation methods were
attempted, the reasons why, and the results.
7. Final factor loadings should be reported
in the results, in a table. This table should
also report the uniqueness (or communality)
of each variable (in the final column). Factor
loadings should be reported with descriptive
labels in addition to item numbers.
Correlations between the factors should
also be included, either at the bottom of this
table or in a separate table.
Confirmatory Factor Analysis
Confirmatory Factor Analysis (CFA) is
generally part of a procedure such as
Structural Equation Modeling (SEM)
conducted via software such as LISREL,
AMOS, MPLUS, etc.
The SEM model typically includes two
different sub-models:
1) the measurement model (CFA) and the
structural model (SEM). In the
measurement model, we specify the
Page 2 of 2
