Using a Bar Code Reader to Enter Behavioral Observations (1988)
© 1988, 2013 by Dallas Denny & Wendy McAmis
Source: Denny, Dallas, & McAmis, Wendy. (1988). Using a bar‑code reader to enter behavioral observations into a portable computer. 1st Annual Conference, Tennessee Association for Behavior Analysis, Blountville, TN, 27‑29 October, 1988. (Received honorable mention in competition).
Using a Bar Code Reader to Enter Behavioral Observations Into a Portable Computer
Initial Impressions (1) (2)
By Dallas Denny
George Peabody College of Vanderbilt University and Greene Valley Developmental Center
And Wendy McAmis
Greene Valley Developmental Center
(1) Reprints are available from the first author at P.O. Box 115, Afton, Tennessee 37616
(2) The authors would like to thank Dr. Jim Fox of George Peabody College of Vanderbilt University for editorial assistance with the manuscript
We connected a hand held bar code reader to an inexpensive portable computer. We wrote a simple BASIC program to enable the computer to time bar code entries and save them as RAM files and used the BCR to record behavioral data in a variety of settings in a large residential facility for persons with mental retardation. Data were entered by scanning bar codes which were assigned to predetermined categories of behavior. Certain bar codes proved to be unscannable, but the system performed reliably. However, we are not cetain entry directly from the keyboard or with paper and pencil would not in most instances be faster and easier. It would be useful to formally contrast keyboard entry with entry via the BCR in a variety of settings; additionally, it would be desirable to compare the performance of this microcomputer based system with that of dedicated hand held scanning devices and paper and pencil entry.
Introduction
Portable electronic devices (event recorders) for the collection of behavioral data have been available since the mid 1970s. Their introduction freed researchers from dependency upon mainframe computers and mechanical tabulating devices and made collection of behavioral data in field settings more feasible (Martin & Bateson, 1986). In the past five or six years, some researchers have elected to use microcomputers instead of dedicated devices like the OS-3 and Datamyte (Bernstein & Livingston, 1982 Deni, 1987; Deni, Szijarto, Eiler, & Fantauzzo, 1983; Denny & Fox, 1987; Denny, Fox, & McEvoy, 1986; Flowers, 1982 Ormsby, Deitz, & Cole, 1986; Reith & Bahr, 1987; Repp, Harman, Felce, van Acker, & Karsh, 1987; Saudargas, 1987). Typically, these microcomputers run programs the investigators or their colleagues have written or are commercial products (Storey & Homer, in press). Reports of these systems have generally consisted of descriptions of the hardware and software involved and of the contexts in which the programs have been utilized. Martin & Bateson (1986) and Bakeman & Gottman (1986) have discussed the merits and liabilities of using event recorders. However, despite their widespread use, the advantages and disadvantages of using either dedicated devices or microcomputers for collection of behavioral data have not been empirically determined.
In the departments of Psychology and Special Education at George Peabody College of Vanderbilt University, the TRS-80 Model 100 portable computer has been used extensively for collection of behavioral data. Although both the hardware and software have performed well, some data collectors have been unwilling or unable to come to grips with keyboard entry of data (Jim Fox, 1987, personal communication). Consequently, we sought alternative methods of data entry. Knowing bar code readers (BCRs) had been successfully used to collect behavioral data (Eiler, 1987), we decided to purchase the bar code wand accessory for the Model 100 and test it in an applied setting. The purpose of this report is to give our initial impressions of the use of a microcomputer based BCR to collect behavioral data in field settings.
Method
Hardware and Software
The TRS-80 Model 102 portable computer is readily available at Radio Shack Computer Centers arid Retail Stores for less than $600.00 (the Model 102 has replaced, but is functionally identical to the computer we used, the Model 100). The Model 102 features a full-sized keyboard, up to 32k of battery-maintained RAM memory, built-in modem, real time clock, and BASIC programming language. There is also a BCR port. The Model 102 weighs about 3 pounds, and can operate for up to 20 hours on four AA alkaline batteries. Its features make it especially suitable for use in field settings.
A BCR (Radio Shack Catalog Number 26-1183) is available at Radio Shack Computer Centers and Retail Stores, selling for less than $100.00. Software provided with the BCR enables the Model 102 to read several different types of bar codes, including UPC-A, which by Federal law appears on many consumer products.
Setting
Greene Valley Developmental Center is a large residential facility (population approximately 750) for individuals with mental retardation. it is located near Greeneville, in Upper East Tennessee.
Applied behavior analysis (behavior modification) is the predominant mode of habilitative treatment at Greene Valley. Behavioral techniques are used both to accelerate desirable behaviors and to decelerate undesirable behaviors. Analysis of the effectiveness of treatment is based upon data collected with paper and pencil. The facility has recently begun to realize the need for improvement of behavioral data collection techniques, especially in the evaluation of the application of (rarely used) aversive procedures. There is also a facility-wide impetus for the use of microcomputers.
Subjects
Because the variables of interest were related to the method of data collection, the first and second authors were the subjects of this research. Both authors had had considerable experience with direct observation and recording of behavior. The first author was thoroughly familiar with the use of microcomputers for collection of behavioral data (Denny, Fox, & McEvoy, 1986; Denny & Fox, 1987), and was accomplished in the BASIC programming language. The second author had a general familiarity with microcomputers, especially in their use with persons with mental retardation. The first author had owned a Model 100 computer for years, and was thoroughly familiar with it. The second author had no experience with the Model 100. Neither author had prior experience with BCRs.
Prior approval for this research was obtained from Green Valley’s Research, Publication, and Training Committee.
Procedure
A scanning sheet was made by cutting out pre-existing bar codes and pasting them on a sheet of paper. Predetermined behaviors were assigned for each bar code.
The authors then used the Model 100 with BCR to collect behavioral data for a variety of ongoing programs for adult men and women with severe and profound mental retardation. When a behavior was noted, the observer (a member of the staff) depressed a key on the side of the BCR and moved it rapidly over the corresponding bar code. Data were entered both in real time and at the end of pre-specified intervals. Examples of the settings and data were recorded are: appropriate social interactions during a club meeting (in this case, recording was done in real time), and recording of self-injurious face scratching and headbanging in a woman undergoing reduction of psychotropic medication (in this case, recording was done from videotape which had been shot in her bedroom). The BCR was also used to enter data for the Aberrant Behavior Checklist, which is used routinely at the facility to screen for problem behaviors.
The computer could be cradled in the crook of an arm, laid on a horizontal surface, or placed in a backpack, saving the observer from carrying a three-pound weight. Holding a clipboard with the scanning sheet attached and the bar code wand (which looked like a fat pen) was topographically similar to paper and pencil entry. We found it most convenient to lie the computer and the scanning sheet on a flat surface.
Reliability
For purely practical reasons, interobserver agreement data were not taken. We could afford only one Model 102.
Results
We were interested in determining whether the BCR had a substantial advantage over paper and pencil. Was entry with the BCR faster, more accurate, and less troublesome?
Character substitution is the misreading of a bar code in which a number or letter is substituted for another. Manufacturers of hand-held BCRs advertise character substitution errors of fewer than 1 per 10,000 characters scanned (Eiler, 1987). In thousands of scans, we did not identify any character substitutions. On the other hand, however, the computer often failed to record passes over certain bar codes; some bar codes were unreadable even with repeated scanning (this is analogous to supermarket situations in which a clerk finally must manually enter information into the cash register). The simple feedback system (the computer beeped to indicate a successful scan) cued the observer when a scan was not successful. Of course, if the incorrect code was scanned, erroneous entry would occur, even though the equipment was functioning properly.
The time for successful scans decreased rapidly with practice. For example, on the first use of the BCR, it took the first author, while sitting at a desk, more than 30 seconds to make the first scan. It took more than three minutes to make only 40 scans. After a month of regular use, the first author was able to scan as many as 116 bar codes in one minute. Of course, this speed could not be obtained while actually observing behavior. We found we could scan rapidly enough to keep up with all but the most rapidly occurring behaviors.
Since the computer matches patterns based on contrast between light and dark areas, anything which reduces that contrast will negatively affect the ability to scan (Eiler, 1987). Photocopies of bar codes were less likely to be successfully scanned than originals. Also, repeated scanning caused the bar codes to rapidly deteriorate. Placing a sheet of clear plastic over the bar code seemed to slow its deterioration.
It was necessary to visually locate bar codes before they could be scanned. This of course necessitated looking away from the subject. There were occasions when targeted behaviors may have occurred while our eyes were on the scanning sheet.
Conclusions
Our impression is that use of a computer-based BCR is a suitable method for collecting some types of behavioral data. The necessity of looking away from the keyboard suggests the BCR may be more practical for collection of interval rather than continuous data. The provision of mechanical guides would provide tactile cues for positioning the BCR and might reduce the need to look at the scanning sheet.
Although the BCR proved to be practical for entry of behavioral data in a variety of settings, its advantages over other forms of entry remain unclear. For that matter, the advantages of event recorders have not been systematically investigated. Bakeman & Gottman (1986) noted “…there is a satisfying physicality about pencil marks on paper. The whole record can be seen easily and parts of it modified with nothing more complicated than an eraser. Although paper can be lost, it almost never malfunctions” (p. 66). Martin & Bateson (1986) caution that computers can introduce complexity and expense in situations in which paper and pencil would do as well.
There are obvious unanswered questions about event recorders—accuracy, expense, time budgets, and reliability of equipment—and perhaps less obvious questions concerning the time necessary to train observers, the effects of instrumentation upon subjects, and the ability to handle complex observation codes. The issue of interobserver agreement is of critical importance.
We have begun to look empirically at some of the respective advantages and disadvantages of event recorders and paper and pencil.
References
Aberrant Behavior Checklist. Slosson Educational Publications, Inc., P.O. Box 280, East Aurora, New York 14052.
Bakeman, P., & Gottman, J. M. (1986). Observing interaction: An introduction to sequential analysis. New York: Cambridge• University Press.
Bernstein, D. M.,& Livingston, C, (1982). An interactive program for observation and analysis of human behavior in a long-term continuous laboratory. Behavior Research Methods & Instrumentation, 14, 231-235.
Deni, P. (1987). Computer assistance for observational research in laboratory or field settings. Perceptual and Motor Skills, 65, 298.
Deni, R., Szijarto, K., Eiler, A., & Fantauzzo, C. (1986). BASIC programs for observational research using the TRS-80 Model 100 portable and Model 4 computers. Behavior Research Methods & Instrumentation, 15(6), 616.
Denny, D., & Fox, J. (1987). Collecting and analyzing data with the TRS-80 Model 100 and the Behavioral Observation Code Computerized (BOSCO). Paper presented at The Twentieth Annual Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Gatlinburg, Tennessee, 25-28 March.
Denny, D , Fox, J , & McEvoy, M. (1986). Using the TRS-80 Model 100 portable computer as a data collection device. Poster presented at The 11th Annual Meeting of The Association for Behavior Analysis, Columbus, Ohio, 24-27 May.
Eiler, J. M. (1987). Bar Code Applications in Medical Records. Paper presented at the Association for Systems Management 1987 Northwest Systems Conference & Seminars on Computing Excellence, Seattle, WA, 18 September.
Flowers, J. H. (1982). Some simple Apple II software for the collection and analysis of observational data. Behavior Research Methods & Instrumentation, 14(2), 241-249.
Martin, P., & Bateson, P. (1986). Measuring Behaviour. New York: Cambridge University Press.
Ormsby, D.E., Deitz, D.E.D., & Cole, K.B. (1986). A microcomputer system for recording and analyzing social interactions in the natural environment. Poster presented at The Nineteenth Annual Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Gatlinburg, Tennessee, 13-15 March, 1986.
Reith, H,, & Bahr, C. (1987). Using Epson HX-20 portable microcomputers to collect observational data in secondary, special education classrooms. Paper presented at The Twentieth Annual Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Gatlinburg, Tennessee, 25-28 March, 1987.
Repp, A., Harman, M., Felce, D., van Acker, R., & Karsh, K. L. (1987) A real-time, parallel entry, portable computer system for observational research Paper presented at The Twentieth Annual Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Gatlinburg, Tennessee, 25-28 March, 1987.
Saudargas, R. (1987). A handheld computer system for classroom observation: Application of the Hewlett-Packard 71B handheld computer. Paper presented at The Twentieth Annual Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Gatlinburg, Tennessee, 25-28 March.
Storey, K.,& Homer, P.H. (in press). Putting behavioral units back into the stream of behavior: A consumer report. Behavior Therapist.