multiTree is a java-based computer program for the analysis of multinomial processing tree models. multiTree provides parameter estimation, hypothesis testing, checks for identifiability, bootstrapping, and power analysis.
multiTree requires at least version 6.0 of the Java Runtime Environment installed on the target machine. The latest version of the Java Runtime Environment can be downloaded free of charge from www.java.com.
Please send questions/comments/suggestions to moshagen[at]uni-kassel.de.
multiTree for Linux
multiTree for Mac
Note: Since Apple considers me as a non-trusted developer, you need to change the default Gatekeeper settings. To do so, go to System Preferences -> Security & Privacy -> General, click the lock to make changes, and select "Allow Apps downloaded from anywhere". Alternatively, open the terminal and type "sudo spctl --master-disable". If you'd rather not want to change the Gatekeeper settings, you may also download the plain multiTree jar file and launch it via the terminal using the command "java -jar -XstartOnFirstThread multitree64.jar".
multiTree for Windows
Note: The multiTree userinterface may look disarranged on high resolution displays. This can be changed by applying a proper scaling to the UI elements as follows: Right-click on multitree.exe, select Properties, select the Compatibility tab, and then select the "Disable display scaling on high DPI settings" (set to System).
Help and Tutorials
Model Selection Based on Minimum Description Length (MDL / FIA)
multiTree allows to compute the Fisher information approximation (FIA), a model selection criterion based on the minimum description length principle similar to AIC or BIC. Essentially, FIA allows to make a trade-off between the fit and complexity of the models under consideration. FIA has the advantage that it takes the functional complexity of the models into account (e.g., how the parameters in the MPT model are connected and whether order constraints such as Do>Dn are included).
However, FIA is only an approximation that can fail if the number of observations is too small, leading to severely biased model selection. As a remedy, it should only be used if the total number of observations (usually the numer of responses times the number of participants) exceeds a lower bound. The following Excel sheet allows to compute this lower-bound sample size for FIA:
For more details see: Heck, D. W., Moshagen, M., & Erdfelder, E. (2014). Model selection by minimum description length: Lower-bound sample sizes for the Fisher information approximation. Journal of Mathematical Psychology, 60, 29-34. doi:10.1016/j.jmp.2014.06.002
You may use multiTree free of charge for academic and personal use. If you want to use multiTree for commercial applications, you need to contact the author. Although considerable effort has been put into program development and evaluation, there is no warranty whatsoever.