Factor analysis is the collective name given to a group of statistical techniques that are used primarily for data reduction and structure detection. In modern business we are inundated with data. New data, old data, new types and formats of data – we are literally drowning in the stuff. And having too much data can be just as useless and debilitating as having too little.
Factor analysis can reduce the number of variables within data to help make it more useful. This reduction in variables within the data also makes it much easier to detect a structure in the relationships between those variables which makes the variables easier to classify.
This is made possible because of a key concept – that multiple observed variables have similar patterns of responses because they are all associated with a latent variable, i.e. a variable that isn’t or can’t be measured. For example, people are likely to respond similarly to questions about personal income, education and occupation because those questions are all associated with the latent variable of socioeconomic status.
It originated in psychometrics and is often used in behavioural and social sciences, marketing, product management and operations research. It is particularly useful when you have large quantities of data to analyse and draw insight from.
You would consider using factor analysis if you need to analyse and understand more about the interrelationships among a large number of variables and to explain these variables in terms of their common underlying dimensions or factors.
For example, if you have gathered a treasure trove of quantitative and qualitative data about your customers and what they think and feel about your product offerings, this is potentially very useful. But only if you can unravel the interdependencies and appreciate what variables affect what outcome. This can be very difficult when there are a lot of potential variables and a lot of potential outcomes.
Factor analysis can help, condensing the information contained in a number of original variables into a smaller set of dimensions (factors) with a minimum loss of information.
Factor analysis can help you extract insights from huge data sets. It can also help you to identify causal relationships that could direct strategy and improve decision making. It can help you to answer:
Say you were going to use factor analysis for marketing purposes; there are four basic steps:
If you want to know more about factor analysis and how to use it you can explore the links at the end of this chapter. Alternatively there are many commercially available factor analysis tools on the market that can help you.
Factor analysis can be used to improve employee engagement because it allows you to analyse the structure of the interrelationships or correlations among a large number of variables by defining a set of common underlying dimensions, known as factors.
If you realise that your staff turnover is too high but you are unsure why, you may conduct exit interviews and initiate an employee survey. But the data alone may not tell you very clearly what is happening in the business. By identifying all the salient attributes that you can think of that may be causing the high turnover you can then use factor analysis to assess the correlations and identify patterns that can help you to solve or at least reduce the problem.
This is particularly useful because both objective and subjective attributes can be used provided the subjective attributes can be converted into scores. Often in issues such as staff turnover or high absenteeism it’s the qualitative subjective data that holds the key to solving the issue. This technique can also identify relationships and latent dimensions or constraints that direct analysis may not.
So long as you have the right statistical program factor analysis is very accessible. It’s easy and inexpensive.
That said, its ultimate usefulness depends heavily on the researchers’ ability to collect a sufficient and relevant set of attributes. If important attributes are missed or ignored then the value of the analysis is significantly compromised and may lead to poor decision making.
To learn more about factor analysis and how to use it see for example:
18.118.163.207