Different approaches to data analysis can reveal new, actionable information based on the same core data. One of the most effective ways to examine survey results is by segmenting respondents using cluster analysis.

Cluster analysis is a multivariate technique that groups respondents with similar behaviors, preferences, or characteristics to gain a greater understanding of the importance of the parallels and differences. This information can be used to add insight and narrow the lens for looking at information in different, specified ways.

Organizing Information

There are two ways to approach cluster analysis:

  • Hierarchical makes observations and joins clusters that remain together throughout analysis
  • Non-hierarchical is used to switch clusters around as the analysis proceeds, most often using K-means.

In order to use cluster analysis effectively, each cluster should have a good number of respondents, as too few reduces usefulness. Determine that the clusters make sense and that they are based on characteristics that are easy to understand and define propecia pills 5mg. Finally, validate the clusters by analyzing how they relate to other variables.

Choosing Criteria

The more relevant your chosen criteria, the more effective your cluster analysis can be. When developing pay structures, some of the best details to hone in on are: years of experience, size of company, location, degree, department sizeā€¦ (?) While the exact grouping should depend on the specific jobs you are trying to price for your own company, looking at the information through different stipulations can add perspective to your questions.
(MarketPay determine specific best features?)

Using Clusters to Analyze

Once you identify the clusters of respondents, you can display them in different ways to chart correlation and causation. Knowing the attributes and characteristics of different types of employees can help you make decisions about compensation packages that go beyond best guesses.

Presenting Results

One of the most important benefits of cluster analysis is the ability to draw conclusions based on a number of factors while easily relaying the results for maximum effect. The most effective software will give you access to charts and graphics that can effectively present the findings and best determine actionable results.

Custom tables enable you to easily create appealing and straightforward exhibits of your process and how you got to your conclusions. Condensing information in this simple way makes it easy to take in during a single presentation, and summarizes the hours of work put into your final decisions.

Cluster analysis is only one of many ways your experts can assess data to draw conclusions and build more comprehensive, competitive pay structures. Using MarketPay’s compensation software and different analysis strategies such as clustering, your enterprise can take the guesswork out of compensation.

Source: //www.edmeasurement.net/5244/SPSS%20survey%20data.pdf