Factor analysis in market research is a statistical method used to uncover underlying dimensions, or factors, in a dataset.
By examining patterns of correlation between variables, factor analysis helps to identify groups of variables that are highly interrelated and can be used to explain a common underlying theme.
Factor analysis can be best used for complex situations where there are many data sets and potentially many variables to summarize.
In this blog, we will explore the different types of factor analysis, their benefits, and examples of when to use them.
What is Factor Analysis?
Factor analysis is a commonly used data reduction statistical technique within the context of market research. The goal of factor analysis is to discover relationships between variables within a dataset by looking at correlations.
This advanced technique groups questions that are answered similarly among respondents in a survey.
The output will be a set of latent factors that represent questions that “move” together.
In other words, a resulting factor may consist of several survey questions whose data tend to increase or decrease in unison.
If you don’t need the underlying factors in your dataset and just want to understand the relationship between variables, regression analysis may be a better fit.
Things to Remember for Factor Analysis
There are a few concepts to keep in mind when doing factor analysis. These concepts guide the way that factor analysis is applied to certain projects and it’s interpretation.
- Variance: This measures how much values are off from the average. Since you want to understand the influence of the factors, variance will help understand how much variance there is in the results.
- Factor Score: This is a number representation of how strong each variable is from the original data. This is also related to a specific factor and can also be called the “component score”. It helps determine which variables are most changed by factors and which are most important.
- Factor Loading: this is typically a coefficient for correlation related to the variable/factor combination. Higher factor loadings means there is a stronger influence on the variables.
What are the Different Types of Factor Analysis?
When discussing this topic, it is always good to distinguish between the different types of factor analysis.
There are different approaches that achieve similar results in the end, but it's important to understand that there is different math going on behind the scenes for each method.
Types of factor analysis include:
- Principal component analysis
- Exploratory factor analysis
- Confirmatory factor analysis
1. Principal component analysis
Factor analysis assumes the existence of latent factors within the dataset, and then works backward from there to identify the factors.
In contrast, principal component analysis (also known as PCA) uses the variables within a dataset to create a composite of the other variables.
With PCA, you're starting with the variables and then creating a weighted average called a “component,” similar to a factor.
2. Exploratory factor analysis
In exploratory factor analysis, you're forming a hypothesis about potential relationships between your variables.
You might be using this approach if you're not sure what to expect in the way of factors.
You may need assistance with identifying the underlying themes among your survey questions and in this case, I recommend working with a market research company, like Drive Research.
Exploratory factor analysis ultimately helps understand how many factors are present in the data and what the skeleton of the factors might look like.
The process involves a manual review of factor loadings values for each data input, which are outputs to assess the suitability of the factors.
Do these factors make sense? If they don’t make adjustments to the inputs and try again.
If they do, you often move forward to the next step of confirmatory factor analysis.
3. Confirmatory factor analysis
Exploratory factor analysis and confirmatory factor analysis go hand in hand.
Now that you have a hypothesis from exploratory factor analysis, confirmatory factor analysis is going to test that hypothesis of potential relationships in your variables.
This process is essentially fine-tuning your factors so that you land at a spot where the factors make sense with respect to your objectives.
The sought outcome of confirmatory factor analysis is to achieve statistically sound and digestible factors for yourself or a client.
A best practice for confirmatory factor analysis is testing the model's goodness of fit.
This involves splitting your data into two equal segments: a test set and a training set.
The next step is to test the goodness of fit on that training data set, which includes applying the created factors from the training data set to the test dataset.
If you achieve similar factors in both sets, this then gives you thumbs up that the model is statistically valid.
How Factor Analysis Can Benefit You
1. Spot trends within your data
If you are part of a business and leveraging factor analysis with your data, some of the advantages include the ability to spot trends or themes within your data.
Certain attributes may be connected in a way you wouldn’t have known otherwise.
You may learn that different customer behaviors and attitudes are closely related. This knowledge can be used to inform marketing decisions when it comes to your product or service.
2. Pinpoint the number of factors in a data set
Factor analysis, or exploratory factor analysis more specifically, can also be used to pinpoint the right number of factors within a data set.
Knowing how many overarching factors you need to worry about allows you to spend your time focusing on the aspects of your data that have the greatest impact.
This will save you time, instill confidence in the results, and equip you with more actionable information.
3. Streamlines segmenting data
Lastly, factor analysis can be a great first step and lead-in for a cluster analysis if you are planning a customer segmentation study.
As a prerequisite, factor analysis streamlines the inputs for your segmentation. It helps to eliminate redundancies or irrelevant data, giving you a result that is clearer and easier to understand.
Here are 6 easy steps to conducting customer segmentation. Factor analysis could fit nicely between Step 3 and Step 4 if you are working with a high number of inputs.
When We Recommend Using Factor Analysis
Factor analysis is a great tool when working with large sets of interconnected data.
In our experience, it’s designed to help companies understand the hidden patterns or structures within data collected. It can simplify complex information, which is especially important when managing numerous variables.
For example, if your company gathered extensive customer feedback through surveys, factor analysis can transform those responses into more manageable, meaningful categories.
Factor analysis helps condense variables like customer satisfaction, product quality, and customer service into broader factors like 'product satisfaction' or 'customer experience'.
Examples of Performing a Factor Analysis
With so many types of market research, factor analysis has a wide range of applications.
Although, employee surveys and customer surveys are two of the best examples of when factor analysis is most helpful.
1. Employee surveys
For example, when using a third party for employee surveys, ask if the employee survey company can use factor analysis.
In these surveys, businesses aim to learn what matters most to their employees.
Because there is a myriad of variables that impact the employee experience, factor analysis has the potential to narrow down all these variables into a few manageable latent factors.
You might learn that flexibility, growth opportunities, and compensation are three key factors propelling your employees’ experiences.
Understanding these categories will make the management or hiring process that much easier.
2. Customer surveys
Factor analysis can also be a great application when conducting customer satisfaction surveys.
Let's say you have a lot of distinct variables going on in relation to customer preferences.
Customers are weighing these various product attributes each time before they make a purchase.
Factor analysis can group these attributes into useful factors, enabling you to see the forest through the trees.
You may have a hunch about what the categories would be, but factor analysis gives an approach backed by statistics to say this is how your product attributes should be grouped.
Factor Analysis Best Practices
1. Use the right inputs
For any market research study in which you plan to use factor analysis, you also need to make sure you have the proper inputs.
What it comes down to is asking survey questions that capture ordinal quantitative data.
Open-ended answers are not going to be useful for factor analysis.
Valid input data could involve rating scales, Likert scales, or even Yes/No questions that can be boiled down to binary ones and zeros.
Any combination of these questions could be effectively used for factor analysis.
2. Include enough data points
It is also imperative to include enough data inputs.
Running factor analysis on 50 attributes will tell you a whole lot more than an analysis on 5 attributes.
After all, the idea is to throw an unorganized mass of attributes into the analysis to see what latent factors really exist among them.
3. Large sample sizes are best
A large sample size will also convey more confidence when you share the results of the factor analysis.
At least 100 responses per audience segment in the analysis is a good starting point, if possible.
Contact Drive Research to Perform Factor Analysis
Drive Research is a national market research company that can help with lots of market research projects. Advanced methods like factor analysis are a part of our wheelhouse to get the most out of your data.
Interested in learning more about our market research services? Reach out through any of the four ways below.
- Message us on our website
- Email us at [email protected]
- Call us at 888-725-DATA
- Text us at 315-303-2040
Tim Gell
As a Senior Research Analyst, Tim is involved in every stage of a market research project for our clients. He first developed an interest in market research while studying at Binghamton University based on its marriage of business, statistics, and psychology
Learn more about Tim, here.
