Scribd
Upload
195 views2 pages

JAMOVI

Exploratory factor analysis (EFA) is used to identify underlying latent factors in a dataset. It reveals hidden factors that can be inferred from observable data. Latent factors represent psychological constructs that are difficult to directly observe. Principal component analysis is a similar data reduction technique. Confirmatory factor analysis tests a hypothesized model of how variables relate to latent factors. Reliability analyses like Cronbach's alpha test how consistently a scale measures a construct, with higher values indicating less error. McDonald's omega is an alternative reliability statistic that does not rely on the assumptions of Cronbach's alpha.

Uploaded by

Jenelyn Ponce Aguilo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
195 views2 pages

JAMOVI

Exploratory factor analysis (EFA) is used to identify underlying latent factors in a dataset. It reveals hidden factors that can be inferred from observable data. Latent factors represent psychological constructs that are difficult to directly observe. Principal component analysis is a similar data reduction technique. Confirmatory factor analysis tests a hypothesized model of how variables relate to latent factors. Reliability analyses like Cronbach's alpha test how consistently a scale measures a construct, with higher values indicating less error. McDonald's omega is an alternative reliability statistic that does not rely on the assumptions of Cronbach's alpha.

Uploaded by

Jenelyn Ponce Aguilo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

JAMOVI values between 0.9 and 1.

0 are
excellent.
Exploratory Factor Analysis (EFA)
- A statistical technique for identifying Principal Component Analysis (PCA)
underlying latent factors in a data - A data reduction technique which,
set. strictly speaking, does not identify
- A statistical technique for revealing underlying latent factors.
any hidden latent factor that can be - Produces a linear combination of
inferred from our observed data. observed variables.

Latent Factors Confirmatory Factor Analysis (CFA)


- Represent psychological - Unlike EFA, with CFA you start with
phenomena or constructs that are an idea - a model - of how the
difficult to directly observe or variables in your data are related to
measure. each other. You then test your model
against the observed data and
Checking Assumptions assess how good a fit the model is.

1. Sphericity Multi-Trait Multi-Method CFA


- Essentially checks that the variables - Both latent factor and method
in your dataset are correlated with variance are included in the model.
each other to the extent that they This is useful when there are
can potentially be summarized with different methodological approaches
a smaller set of factors. used for measurement and therefore
method variance is an important
2. Sampling Adequacy consideration.
- Checked using the Kaiser-Meyer
Olkin (KMO) Measure of Sampling Internal Consistency Reliability Analysis
Adequacy (MSA). - Tests how consistently a scale
measures a psychological construct.
KMO Index
- A measure of proportion of variance Cronbach Alpha
among observed variables that - Cronbach’s alpha (𝛼) is such a
might be common. statistic: a function of all the split-half
- If the KMO index is high (≈ 1), the coefficients for a scale. If a set of
EFA is efficient whereas if KMO is items that measure a construct (e.g.
low (≈ 0), the EFA is not relevant. an Extraversion scale) has an 𝛼 of
KMO values smaller than 0.5 0.80, then the proportion of error
indicate that EFA is not suitable and variance in the scale is 0.20. In other
a KMO value of 0.6 should be words, a scale with an 𝛼 of 0.80
present before EFA is considered includes approximately 20% error.
suitable. Values between 0.5 and 0.7
are considered adequate, values
between 0.7 and 0.9 are good and
Mcdonald's Omega Chapter 14: Connect
- An alternative is McDonald’s omega
(𝜔), and jamovi also provides this 1. Narrowly defined personality tests
statistic. Whereas 𝛼 makes the used in the context of
following assumptions: (a) no employment-related research and
residual correlations, (b) items have practice like the integrity test have
identical loadings, and (c) the scale been characterized as _____
is unidimensional, 𝜔 does not and is personality scales.
therefore a more robust reliability Criterion-focused occupational
statistic. If these assumptions are
not violated then 𝛼 and 𝜔 will be 2. A commonly used performance test
similar, but if they are then 𝜔 is to be in the assessment of business
preferred. leadership ability is the _____.
- Sometimes a threshold for 𝛼 or 𝜔 is Leaderless group technique
provided, suggesting a “good
enough” value. This might be 3. _____ has been defined as the
something like 𝛼s of 0.70 or 0.80 strength of an individual's
representing “acceptable” and identification with and involvement in
“good” relia- bility, respectively. a particular organization.
However, this does depend on what Organizational Commitment
exactly the scale is supposed to be
measuring, so thresholds like this 4. Wonderlic Personnel Test
should be used cautiously. It could Useful in screening individuals for
be better to simply state that an 𝛼 or jobs that require both fluid and
𝜔 of 0.70 is associated with 30% crystallized intellectual abilities.
error variance in a scale, and an 𝛼 or
𝜔 of 0.80 is associated with 20%. 5.
- If you are getting an 𝛼 coefficient
above 0.95 then this indicates high
inter-correlations between the items
and that there might be too much
overly redundant specificity in the
measurement, with a risk that the
construct being measured is
perhaps overly narrow.

Uniqueness
- The proportion of variance that is
‘unique’ to the variable and not
explained by the factors.

You might also like