Exploratory Factor Analysis
What to include in an exploratory factor analysis
The following three actions should be taken when finishing an exploratory factor analysis:
1. Determine the number of factors
Setting the number of factors in the theory you are testing is the first step in conducting an exploratory factor analysis. For instance, you might take into account variables like patient locations, length of treatment options, and patient ages in each group if you want to test the hypothesis that patients with anxiety respond better to behavioral therapy than medication. Plotting the factors will show how much they vary once you know how many there are. This may help you understand which factors most likely correlate.
2. Select a method for extraction
The researchers’ next step is to carry out a second factor analysis, which aids in determining the loadings for each factor they’ve chosen. Researchers then select a mathematical strategy to find this. There are various methods they can choose from, including:
3. Choose a method for rotation
The researchers’ final step is to rotate their recently extracted loadings. As a result, they can produce the highest loadings and eliminate lower ones, which helps them simplify the structure as much as possible. Researchers may use either an orthogonal or oblique rotation type. Researchers frequently use the orthogonal type by default, which makes the assumption that factors don’t correlate with one another. Oblique type assumes that variables are related to one another and how they are related.
What is exploratory factor analysis?
Psychologists use exploratory factor analysis, a statistical technique, to create psychometric tests. It can be used by researchers to develop questions about their research topics, comprehend the relationships between variables, and find latent variables. This approach places a strong emphasis on the interrelationships between common factors and manifest variables. Additionally, it aids researchers in making connections between variables and indicators prior to moving on to the next phase of their study.
EFA vs. CFA
EFA and CFA are both used by researchers to complete various tasks related to their research processes. Here are the primary similarities and differences between the two:
Researchers use factors they’ve gathered by evaluating principal output components during exploratory factor analysis. These are used by them to gauge the accuracy of their internal metrics. These might not be related variables, but they can also aid researchers in evaluating the quality of each individual item. Researchers may load any item on any factor. They can select these loadings using a variety of estimators, such as maximum likelihood. Researchers can group correlating variables and indicators using EFA in order to evaluate the validity of the internal factor structure.
To determine how trustworthy internal measures are, researchers conduct confirmatory factor analysis using only theoretical factors. Even if two factors are unrelated, they can still be used by researchers to assess item quality. Researchers specify factors structures to use CFA to identify the categories of item loads. This enables researchers to relate potential factor structures to observed data. Researchers frequently select factor loadings based on the most likely scenario. Researchers can use CFA to restrict factor correlations, limit loadings to particular relationships, reduce the range of measurement error correlations, compare alternative models, contrast multiple factor structure groups, and test secondary factor models.
Assumptions of EFA
When using exploratory factor analysis, researchers assume that every set of variables they can observe has underlying factors. They also believe that these frequently unobserved variables can shed light on the nature of the relationships between their observed variables. Additional presumptions used by researchers when conducting an exploratory factor analysis are listed below:
What is exploratory factor analysis?
Exploratory factor analysis (EFA) is typically used to identify a measure’s factor structure and assess its internal consistency. When researchers don’t have any theories about the nature of the underlying factor structure of their measure, EFA is frequently advised.
What is an example of exploratory factor analysis?
Human physical characteristics like height, weight, and pulse rate are examples of measured variables. Typically, researchers would have a large number of measured variables that they would assume to be connected to a smaller number of “unobserved” factors.
What is the difference between confirmatory and exploratory factor analysis?
All measured variables are related to each latent variable in exploratory factor analysis. However, confirmatory factor analysis (CFA) allows researchers to define the number of factors needed in the data as well as the relationship between each measured variable and latent variable.
What is exploratory factor analysis SPSS?
Exploratory factor analysis looks for underlying factors or variables that could explain the correlations between a set of observed variables.