Useful Facts about Factor Analysis

While some may have heard of the technique “factor analysis” many remain unclear about exactly what it is. How does factor analysis figure into decision making? How is it applied? What is it used for?  What are “latent factors”? While these terms may sound complicated or otherwise cumbersome the fact is that the ideas behind factor analysis are rather straightforward.

What Does Factor Analysis Do?

We are all living, breathing statistics. That may sound cold, but it’s a fact. We are all consumers of media, news, information, products, and services. Because all people really are to statisticians is statistics, it’s helpful to be able to strip away unnecessary information. Take people; as complex and individual and amazing though they are; and try to reduce them down to the lowest common denominator. In doing this the statistical “dimensionality” is reduced. Rather than complex octagons we are now simple boxes.

Latent Factors

Another important thing which this technique addresses are the “latent factors.” What does that mean? Well anything latent; a feeling, an impression, a hunch; is something that can’t be measured. So when you’re doing market research there are things you can know; name, age, gender, ethnicity, income level, occupation; and things you can’t know; passions, intelligence, motivating factors, upbringing. Still with factor analysis we’re able to group these individuals accordingly by the responses they give. As some have keenly observed “the observable data doesn’t create the underlying factor; the underlying factor creates the observable data.”

How Does it Help

One thing which many people conducting surveys do is ask questions which have no real merit unless observed repeatedly. If you asked 100 people in a restaurant with 35 tables and 100 chairs and 1 waitress how their experience was; likely all of them would say terrible. However if the one waitress was covering for 30 others, then the data you got for this one particular day about this one particular waitress would not stick in any real way. Diners experience on the next day when the place was fully staffed would be far different.

As the body giving the survey you want to be dealing with facts. Cold, hard, facts; things which can be correctly transposed to different settings with the same result. A plane takes off from tremendous speed; a helicopter takes off from a stationary position. These are things which cannot be debated. If you’re trying to group results together, factor analysis can really help. You are dealing with measureable items, your latent observation has been borne out and you can group these respondents accordingly.

Return Shoppers

You can also get measurable results if you’re asking the right questions. Using this technique you should be able to accurately peg a customer’s likelihood that they will purchase a product or service after using it once. If this technique was found to be accurate over an extended period of time it would be immeasurably valuable to all businesses who are trying to discover a products survivability in the marketplace. Using a tool like factor analysis to determine which products to push on full force and which to abandon would vastly change the landscape of the marketplace.

Of course whether or not this remains true over the longer haul; whether this techniques implementation has any lasting chance of thinking before and ahead of consumers, is still up for debate. Still with so much potentially riding on factor analysis’ success or failure, it’s something we all should try and get our heads around sooner rather than later. Nobody wants to be behind the 8-ball on the next big, sure thing!

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>