SI-CHAID®
SI-CHAID® 4.0 performs tree based analyses on a categorical dependent variable using the CHAID (CHi-squared Automatic Interaction Detector) algorithm. An extension to multiple dependent variables is possible using an interface to Latent GOLD® where a CHAID analysis is performed following the estimation of any latent class model in Latent GOLD® to profile the resulting latent segments based on demographics and/or other exogenous variables (see product description below for more details).
SI-CHAID® is licensed on a perpetual basis (does not expire). Discounts for multiple users are automatically applied to your order (2-4 licenses – 15%, 5-9 licenses – 20%, 10+ licenses – 25%).
To order SI-CHAID® together with a Latent GOLD® license, please click here.
Description
The Define part of the program is used to set up a CHAID Definition (.chd) file with the File → New command, or alter the specifications of an existing .chd file with File → Open. The typical setup includes the selection of the dependent variable, the predictor variables, the combine-type of the predictors, and various options for growing the tree (stopping rule, significance levels, etc.). Define may also be used to enter or modify scores for the categories of the dependent variable when the ordinal algorithm is specified. The model specifications, which are saved with a .chd extension, can be inspected with a text editor (Notepad, for example).
The Explore part of the program allows you to grow or alter a SI-CHAID® Tree, automatically or interactively, using the settings given in a previously saved (.chd) file. It can also be used to produce crosstabulations, gains charts, and if-then-else source code statements that can assist in scoring your data file.
Our flagship modeling tool Latent GOLD® 6.0 provides a direct link to SI-CHAID® 4.0. With this option, a CHAID Definition (.chd) file is automatically generated immediately following model estimation which can then be used as input to SI-CHAID® 4.0. (See Latent GOLD® Tutorial #4, and Choice Tutorial #1A).
SI-CHAID® has the following unique features:
Tree nodes have detailed information which may be customized using the Tree Node Display panel. From this view the SI-CHAID model may be modified by growing, pruning, or restoring previously saved tree branches or by rearranging category groupings. Operations on the tree take place on the “current” node which is the highlighted (active) node. Clicking on a node makes it the current node. Multiple Tree Diagram windows may be open, each displaying different node contents or other customized views.
The Gains chart provides various tabular representations of the terminal nodes (segments) from the SI-CHAID tree which may be customized using the Gains Items panel. Multiple Gains Chart windows may be open, each with its unique customized appearance.
The table provides tabulation of a single predictor by the dependent variable. The cell entries can be customized using the Table Items panel.
Often segmentation is desired that is predictive of not one but multiple criteria. For example, in database marketing, dependent variables might include 1) response to the most recent mailing (responder vs. nonresponder), 2) response to past mailings, 3) the amount spent, 4) profitability, and possibly others. It is now possible to obtain CHAID segmentation trees that are predictive of multiple dependent variable criteria. In addition, these dependent variables may be continuous, ordinal, nominal, or count variables, or any combination of these!
The groundbreaking article “An Extension of the CHAID Tree-based Segmentation Algorithm to Multiple Dependent Variables”, Magidson and Vermunt (2005) shows this is possible. The key is to use latent classes as a proxy for the multiple dependent variables. This can be done with Latent GOLD® 5.1/5.1/4.5 when the dependent variables are used as indicators in a latent class cluster or factor model, or it can be done with Latent GOLD® Choice 5.1/5.0/4.5 when the dependent variables are choices obtained from a discrete choice study.
Each of our flagship modeling tools Latent GOLD® 5.1/5.0/4.5 and Latent GOLD® Choice 5.1/5.0/4.5 provide a direct link to SI-CHAID® 4.0. With this option, a CHAID Definition (.chd) file is automatically generated immediately following model estimation which can then be used as input to SI-CHAID® 4.0.
To see how this works:
- Read the article (Magidson and Vermunt, 2005)
- View a tutorial where a CHAID tree is developed that is predictive of 11 dependent variables in conjunction with Latent GOLD® 4.5 (SI-CHAID® Tutorial #4)
- View a tutorial from Latent GOLD® where a CHAID analysis is used with a discrete factor model to better understand the factors (Latent GOLD® Tutorial #4)
- View a tutorial from Latent GOLD® Choice (Choice Tutorial #1A)
System Requirements
SI-CHAID® is designed to operate on XP/Vista, Windows 7/8, Windows 10 or Windows 11
System Requirements: 16MB Drive Space, 512MB of RAM.
Input files: SPSS system files, delimited text files.
Tree nodes have detailed information which may be customized using the Tree Node Display panel. From this view the SI-CHAID model may be modified by growing, pruning, or restoring previously saved tree branches or by rearranging category groupings. Operations on the tree take place on the “current” node which is the highlighted (active) node. Clicking on a node makes it the current node. Multiple Tree Diagram windows may be open, each displaying different node contents or other customized views.
The Gains chart provides various tabular representations of the terminal nodes (segments) from the SI-CHAID tree which may be customized using the Gains Items panel. Multiple Gains Chart windows may be open, each with its unique customized appearance.
The table provides tabulation of a single predictor by the dependent variable. The cell entries can be customized using the Table Items panel.
Often segmentation is desired that is predictive of not one but multiple criteria. For example, in database marketing, dependent variables might include 1) response to the most recent mailing (responder vs. nonresponder), 2) response to past mailings, 3) the amount spent, 4) profitability, and possibly others. It is now possible to obtain CHAID segmentation trees that are predictive of multiple dependent variable criteria. In addition, these dependent variables may be continuous, ordinal, nominal, or count variables, or any combination of these!
The groundbreaking article “An Extension of the CHAID Tree-based Segmentation Algorithm to Multiple Dependent Variables”, Magidson and Vermunt (2005) shows this is possible. The key is to use latent classes as a proxy for the multiple dependent variables. This can be done with Latent GOLD® 5.1/5.1/4.5 when the dependent variables are used as indicators in a latent class cluster or factor model, or it can be done with Latent GOLD® Choice 5.1/5.0/4.5 when the dependent variables are choices obtained from a discrete choice study.
Each of our flagship modeling tools Latent GOLD® 5.1/5.0/4.5 and Latent GOLD® Choice 5.1/5.0/4.5 provide a direct link to SI-CHAID® 4.0. With this option, a CHAID Definition (.chd) file is automatically generated immediately following model estimation which can then be used as input to SI-CHAID® 4.0.
To see how this works:
- Read the article (Magidson and Vermunt, 2005)
- View a tutorial where a CHAID tree is developed that is predictive of 11 dependent variables in conjunction with Latent GOLD® 4.5 (SI-CHAID® Tutorial #4)
- View a tutorial from Latent GOLD® where a CHAID analysis is used with a discrete factor model to better understand the factors (Latent GOLD® Tutorial #4)
- View a tutorial from Latent GOLD® Choice (Choice Tutorial #1A)