|
| Technology in Applied Behavior Analysis |
| Tuesday, June 1, 2004 |
| 9:00 AM–10:20 AM |
| Republic B |
| Area: TPC/DDA; Domain: Applied Research |
| Chair: Michael J. Cameron (Simmons College) |
| Discussant: Raymond G. Romanczyk (Binghamton University) |
| Abstract: Learning Objectives
Participants will lean about the interface between MATLAB and task analysis research
Participants will learn about the variables that both hinder and facilitate the completion of a task analyzed progam
Participants will learn about the value of data sonification and the production of soundgraphs |
| |
| The MATLAB Profiler: A Model for Task Analysis Research |
| NED GULLEY (MathWorks) |
| Abstract: There is a great deal of similarity between writing computer code and developing a task analysis for instruction. Both require a line-by-line display of information and both must be designed for optimal performance. One difference, however, is that computer code can be easily analyzed. The MATLAB Profiler shows computer programmers where code is slow and offers specific suggestions about how to facilitate the speed of a program. The Profiler evaluates computer code against a wide array of variables and on an a priori basis, flags specific lines that will hinder overall performance. The Profiler serves as an excellent model for evaluating task analyzed programs. |
| |
| Technology and Task Analysis Research |
| MICHAEL J. CAMERON (Simmons College) |
| Abstract: Teaching skills by way of a task analyzed program is standard practice in the field of applied behavior analysis. Despite the fact that behavior analyst have clear guidelines on how to develop a task analysis, there is a lack of literature on how to evaluate the quality of a task analyzed program. The MATLAB Profiler provides behavior analysts with a conceptual model for program evaluation. Within this presentation, the effect of an a priori analysis of a wide variety of variables will be demonstrated. Specifically, task analyzed programs, evaluated pre and post analysis will be displayed to demonstrate the need for an a priori program evaluation. |
| |
| I Can Identify That Data Path in Four Notes |
| ROBERT L. SHAPIRO (Simmons College) |
| Abstract: In addition to recognizing data patterns in a visual format, it has been shown that with minimal training subjects are able to recognize data patterns presented aurally by a computer. Similarly, data have occasionally been used to create musical pieces. This is a natural leap, as musical pitches and sounds are based upon a logarithmic scale, taking into account such variables as topography (instrumentation), frequency (expressed in Hertz), magnitude (volume), duration (length of notes), and latency (onset of notes). However, existing musical pieces based on data center around the eight-note major scale, omitting the other five notes in the thirteen-note chromatic scale (as well as countless microtonal notes prevalent in the music of other cultures) and therefore not accurately representing data. This study used data gleaned from a variety of behavior analytic research projects and converted these data into soundgraphs, talking into account all thirteen notes of the chromatic scale. A comparison of visual versus technology generated aural data will be presented, along with implications for extending the use of soundgraphs to interpret numerical data. |
|
| |