Decision Trees

 

Understanding interactions of explanatory variables
for some variables of interest

Decision Tree Analysis is an exploratory method which explains some variable of interest without using a strict regression model.

It falls half-way between the categories of Segmentation and Key Drivers Analysis.

It is designed primarily for situations where:

  • The variable of interest is a two category (Yes / No) or multi-category (choose one option from 3+) variable.
  • Most of the potential predictors are categorical, rather than numeric scales.
  • We are interested in understanding the interactions between predictors, rather than their overall effects.
  • A simple model is required based on observed data, rather than a regression model.
  • We want to identify small groups in the population, which contain a high proportion of cases, belonging to some category of interest.

The most well-known form of decision tree analysis is known as CHAID (Chi-squared Automatic Interaction Detector). Predictors are built up in layers in a tree-like fashion with a root, nodes and branches. For multi-category predictors, CHAID creates optimally merged categories in order to simplify the resulting model and reduce the number of branches.

These models are often useful in their own right. For example, “What combination and interaction of demographic category typifies someone who buys brand X?”. They are also useful as an exploratory lead-in for other methods to help identify which categories of potential predictor should be merged and combined in, say, a Drivers Analysis.

Back to Statistical Toolkit
Case Studies

  • Employment Tribunals

    We have worked on several CHAID analyses to determine the combination of features of employment…

Key Factors

  • Identify small groups in the Population which account for high proportions of behaviour or interest
  • Intuitive segments defined by combinations of existing variables
  • Optimal merges and combinations
  • Ideal for categorical data (e.g. demographics)

 

  • Latest Resources

    • Choice Modelling – A New Approach to Dual Response Author - Gary Bennett

      In preparing for a conference paper, we realised that the choice community as a whole has adopted an unrealistic method for dealing..

      View Full Article

    • Extensions to our Correlated Component Regression (CCR) Methodology Author - Gary Bennett

      Correlated Component Regression is now our preferred methodology for predictive modelling. It provides more reliable, stable and robust models particularly in the..

      View Full Article

    • Factor Analysis in a Regression Context Author - Gary Bennett

      As well as providing everything you need for a Key Drivers analysis, our CCR methodology enables us to analyse the underlying components..

      View Full Article

    • Gary and his team really understand the research business, are friendly and helpful, give advice on methods when I’m floundering and (importantly for us non-statisticians!) explain methods and outputs in layman’s terms.  I know, whenever they’re involved in a project, it will be completed on time and accurately, with easily digestible outputs.  Gary himself has the experience and gravitas to be extremely credible in front of clients.  For all these reasons, The Stats People is always my first port of call when complex statistical analysis is needed.

      Eileen Sutherland maru / edr

    • Working with Gary and his team is very easy; they provide statistical analysis in a way that is simple to understand and meets our research objectives completely. The team not only provides the statistics, but also a thorough recommendation and explanation of why this is the best approach to take and how the data should be interpreted. This experience with The Stats People has been second to none.

      Sheetal Padania Open Plan Research

    • Gary and the team are always very responsive – making time to talk through ideas and solutions informally. I have a great deal of confidence that they will always recommend the most appropriate and pragmatic approaches to achieve our objectives and every project we have worked together on over the past few years has been a success in that regard.

      Alistair Kuechel IFF Research

    • Gary and the team are one of the most innovative providers of stats services there is. Our clients get great benefit from our partnership with The Stats People too. I am happy to recommend Gary and the team to colleagues as often as I can.

      Phil Dunn Planning Shop International

    • Gary is a world class statistician with both a broad and deep knowledge of the statistical tools available to market researchers, and how they can best be applied to give our clients accurate and actionable insight.  I have worked with Gary and his team for many years and value his expert advice, commitment and innovative approach.

      Chrissie Wells Breaking Blue Research

    • Gary and his team are integral to our quantitative work.  They have a deep understanding of market research in general, and also the specific challenges of healthcare research.  They also have the knack of presenting complex ideas in a clear and logical way; we often include them in discussions with our clients at proposal, design and analysis stages.  Above all, they are a pleasure to work with.

      Guy Pascoe Purdie Pascoe

    • Log In

    The following log in is for clients who have purchased our Data Visualisation Services.
    Enter your log in user or email and password

    Click here for more information on our interactive graphs.

    Log In

    CORExpress® Register

    Enter your details on our register for more information on CORExpress® Software and training courses.
    • This field is for validation purposes and should be left unchanged.