CACTI is a GUI-driven tool used to code and study the delivery of Motivational Interviewing
The CASAA Application for Coding Treatment Interactions (CACTI) was developed as an instrument to facilitate accurate, sequential parsing and coding of human verbal interactions. CACTI has a simple interface for users to partition verbal interactions into discrete utterances that may then be coded by individual raters. Coding data for each utterance, including the beginning and ending times and rater codes, are stored sequentially in a text file that may then be processed for data analyses in statistical packages like GSEQ, SPSS, R, and SAS.
CACTI provides clear advantages over other coding software. First, CACTI allows a user to code complex verbal behavior without the need for the speech to be transcribed manually, which can save money and time. Second, CACTI saves all data digitally, which can be manipulated directly using other spreadsheet or statistical packages, thus reducing the cost and error associated with the manual data entry necessary for transcripts coded on paper. Third, CACTI retains the temporal sequence of codes, thus allowing researchers to assess not only the frequency of behavior counts, but also the patterns of sequences in which the behavior occurs. Fourth, CACTI allows coders to evaluate sessions “on the fly”, without requiring an additional pre-parsing step. For systems with a limited number of codes, CACTI may successfuly be used “live” during sessions.
CACTI is free and open-source, per the terms of the GNU Public License: http://www.gnu.org/licenses/gpl-3.0.txt. Anyone is free to download, modify, and redistribute the software and source code, which are currently available at https://github.com/jling-NM/CACTI/releases. Prior versions are currently available at https://code.google.com/p/ctcsu-player/. Note that this software is unsupported by UNM CASAA.
CACTI was initially developed by Alex Manuel, a computer scientist at the University of New Mexico, and Carl Staaf, a consultant at Robot Super Brain, in conjunction with Drs. Theresa B. Moyers and Jon M. Houck of the UNM Center on Alcoholism, Substance Abuse, and Addictions with funding from NIDA R01DA021227. Versions after 0.10 were developed by Josef Ling in conjunction with Jon Houck with funding from NIMHD R01MD009708.
1. Reduces money and time spent on transcription of lengthy interactions
2. Stores all data digitally, thus reducing error and time associated with manual data-entry
3. Is open source software, and thus free to use, edit, or redistribute for anyone and no entity
profits off of its use
4. Store sequential information for each utterance, allowing for more sophisticated data analyses
than simple frequency counts
5. Is easy to use
6. Saves paper necessary for printing lengthy transcribed conversations
7. Is adaptable to be edited to work with most coding systems, including those that use behavior
counts and global Likert-type ratings
8. Is developed by social scientists, for social scientists
9. Is Java-based, and thus should run nearly uniformly across operating systems
10. Coding interactions does not require pre-parsing, and utterance-by-utterance inter-rater reliability
can easily be computed using an alignment kappa available in freely-available software packages such as
the General Sequential Querier (GSEQ).
OS X
Download and mount DMG file. JRE 8 is included so an existing JRE is not required.
Windows
Download and run the EXE file. The Windows executable requires a pre-existing JRE 8 installation.
Debian Linux
Download and install the DEB file.
NOTE: The program does work with JRE versions above version 8.