Since 2002, 4 million visitors plus:
hit counters
search engine optimization service

  Appletcollection Vertical Menu java applet, Copyright 2003 GD

Dumont, R., & Chafouleas, S. (1999) Conducting behavioral observations: Some technical support?, Communiqué, Vol. 27, 7, 32-33

 

Since the proposed IDEA regulations now require careful analysis of behavior in certain prescribed situations, school psychologists are looking for more efficient ways to conduct and collect data from behavioral observations. The use of computers and particularly specific software may aid in the collection and analysis of this data. Two recently published software programs, !Observe and the Behavior Observation Assistant (BOA), purport to make the collection, organization, and storage of behavioral observations trouble-free. These programs have been used and critiqued. Information specific to each program is provided, as well as summary comments regarding the use of computer software for data collection.

!Observe software ($149), Psychsoft, Inc. (1-800-536-4996) http//www.psycsoft.com

!Observe software is available for all Windows and Mac platforms. The software is also available for the Apple Newton (110 or better). The !Observe computer program (version 1.0b4) was reviewed using a Macintosh PowerMac 7100/66, a Power Book 160, a Newton Message Pad 130, and an IBM Dell portable. Installation on each computer was straightforward and flawless. Simply copying the appropriate files to the Mac hard drive or running the Setup program on the IBM had the files and programs installed in seconds. During this review, we found that use of the Apple Newton (or any similar IBM based palm size computer) had a number of distinct advantages over the laptops and desktop computers. Due to its small size and unobtrusiveness, the Newton was ideal for taking to a classroom to record behaviors. Since familiarity with the mouse, trackball, or touchpad is necessary for input of the data, the Newton, with its stylus input, made using the software efficient and easy.

Although a short 20+ page manual accompanies the software, anyone semi-computer literate will find this software very user friendly. The data collection process is built around customizable templates that we found to be extremely easy to produce. Once familiar with the program, in less than 5 minutes, these reviewers were able to create a rather elaborate template for comparing various on-task / off-task behaviors for a subject and a control. Templates, either provided with the software (ED, MR, ADHD, Critical Incidence, Autism, Mental Status, and a Functional Assessment) or created by the user, are simply groups of "behavior buttons" that are used to tabulate which behaviors will be observed. The templates that come with the program were useful as examples of what could be easily done with the !Observe software. We found that it was so easy to create templates that we never used any of the preprogrammed ones. Each template may have up to 24 "behavior buttons which may be color-code buttons to aid collection. Whenever a behavior of interest is noted, the observer simply clicks on the appropriate button using the stylus, mouse, track pad, or pointer device. Each behavior may be grouped into some broader variable (i.e., Class and Category). For example, a button labeled "hitting" might be coded as belonging to a negative class and a physical category while another button labeled "quiet work" might be coded as belonging in the positive class and the nonverbal category.

Cues are available to serve as a reminder to enter your observation data. Time interval may be set to any whole-second increment , and total observation time can be set in any whole minute increment. Reminders are provided by either a Flash (the background behind the headings of the capture window change color every time an interval passes) or a Beep (the system beep will sound when an interval passes). Users may choose to have both or neither reminders on. Although the author lists a number of data collection options, we found the program capable of being configured to aid in a variety of observation formats. Specifically, frequency counts, interval and duration recording, and momentary time sampling were all readily available. In order to conduct each type of observation, the user simply needs to plan at which point data is to be recorded, and then consider the collection format when interpreting the summary report.

The program automatically saves the observation data. The summary report lists each of the observed, coded behaviors followed by their class, a count of the number of times each of the specific behavior buttons was clicked, and percentage of the time the button was pushed in relation to all other buttons. In addition, it is possible to obtain information regarding rate per minute and duration (in seconds) of each behavior. Users may also elect to have observations summarized as a data stream, listing each behavior in the order of recording along with the exact time (8:43:44 PM) that the behavior occurred. !Observe provides two forms (Standard or Interval) for summarizing the data. Standard form records every behavior button pushed during a set interval while Interval form tracks every different button pushed within an interval. If, for example, within a single 10 second interval the "On Task" behavior button is pushed three times and the "Off Task" behavior button 2 two times, !Observe would record this as On Task - 60%, Off Task - 40% using Standard form and On Task 50% (one interval) and Off-Task 50% (one interval) using interval form. These reviewers saw little utility in the information provided by the Interval summaries. Bar and Pie charts are available.

Behavior Observation Assistant, ($99.00) Bunger Solutions (972- 424-9647) http:www.bungersolutions.com

The Behavior Observation Assistant (BOA) is available for IBM or IBM compatible platforms, and requires Windows 3 or later to operate. The software comes on 5 3 ½ inch disks, and is easily installed in minutes. Technical support is available through an online help section or via email. The accompanying manual is easy to read and provides a number of useful examples. Although the largest feature of the program is data collection and summary, the Behavior Observation Assistant also provides a template for creating behavior intervention plans, and reference tools related to behavior management (i.e., manifestation determination, checklists of possible reinforcers).

The setup for data collection is easily accomplished through the use of an on-screen toolbar. Data collection itself involves the following steps: subject setup, selection of behavior(s), and collection of the data. Subject setup is done by first selecting a group (i.e., Mr. Smith's classroom) and then an individual (Johnny) within the group. Preprogrammed target behaviors are provided (i.e., hitting, vision not properly directed, speaks without permission). These may be easily selected and edited by clicking on each or the observer may edit or create totally new behaviors. Observation location (i.e., gym) and setting can be specified. Up to 6 behaviors can be recorded when observing 1 subject, however if more than 1 subject is involved in the observation, only 1 behavior may be recorded. To actually collect data, the observer presses preset function keys that relate to the selected behaviors. Pressing the escape button allows the observer to make comments or correct errors. However, doing this halts the observation, thus, providing inconsistencies in observation real time when comparing sessions.

The authors report a number of data collection options, including interval recording, time sampling, frequency recording, and duration recording. The choice of observation type (i.e., baseline, intervention, follow-up) is also provided as an option. Observation time periods and other collection information can be modified according to need, however, range limitations may be present depending on the type of observation conducted. For example, time can be selected from 5 seconds to 1 minute when using interval recording while time periods from 1 to 10 minutes are available when using time sampling. It should be noted that the BOA's time sampling procedures are not equivalent to momentary time sampling procedures described by Saudargas and Lentz (1986). The BOA stops recording time until either a response is entered or 10 seconds have elapsed before moving on, thus potentially providing inconsistency in the collection across sessions. Data from each observation is saved, and may be viewed in a summary report. The summary report presents all background information regarding the session (i.e., Who was observed?, How were they observed?, How long were they observed?), and combines the observation data in table format. Reports may be exported into a file format (i.e., word processor) different from BOA. Observation data is automatically saved and sessions may be organized through the file management button on the toolbar. Graphing options are not available, and data across observation sessions cannot be combined into one report.

Summary Comments

In summary, the !Observe and the Behavior Observation Assistant provide a variety of data collection and behavior management options in a relatively easy to use format. Subject and behavior setup is easy to accomplish, and the behavior choices provided should fit the needs of most school psychologists. In order to become fluent with how to record the data, we found it necessary to practice a number of times the use of the input device. For the BOA, if 6 behaviors have been coded into the template, it may be difficult to remember which function key corresponded to which behavior. With the !Observe software, the buttons themselves are labeled, making the input very smooth. However, we did notice that if we used templates with many labeled buttons, we had some difficulty keeping accurate track of behaviors simply because we had too many buttons. The more we used the two programs, the less of a problem we had. For both programs, the user must carefully consider how data collection is configured in order to understand the type of observation procedure he/she is using, particularly when comparing information across subjects and observations. For example, the !Observe manual states that interval observations are conducted, but the description of the data collection actually suggests momentary time sampling. The BOA describes time sampling, but this procedure is not the same as the momentary time sampling conducted by the !Observe. Bell and Beedle (1993) provide a review of data collection techniques and terminology. For both programs, the observation summary reports are easy to read, and save time by doing all tabulations for the observer, but multiple observations need to be combined by hand. Finally, for the BOA, although it can be helpful to have references related to behavior management and a behavior intervention plan template, the observation data are not linked to the template. For both programs, technical assistance, by phone and/or e-mail, was quick, professional, and accurate.

Is it worth using computer software programs to make behavioral observations? The answer is - it depends. First, use of the software in school settings is may be limited by the technical requirements. Clearly, one needs at least a laptop computer to fully utilize either software program (we can't imagine anyone hauling a desktop computer into a classroom). Until students were familiar with our presence, carrying a laptop computer into a classroom invariably caused a stir and a number of raised eyebrows. Nonetheless, the laptop served its purpose and the observations went smoothly. Second, usefulness depends on the purpose of data collection and desired efficiency. If the purpose is to collect frequency counts of a behavior, it may not be more efficient to bring out a computer when paper would do just as well. For more complex data collection techniques such as momentary time sampling, use of the software by an observer may facilitate accuracy of the collection given the potential auditory and visual reminders, and save time with data tabulation. However, since the observation reports do not easily combine data across observations, the observer does not save time with the analysis of behavior patterns. In summary, we were able to identify a number of positive features of each of the reviewed software programs which may be useful in the school setting. Software selection is best made after careful analysis of the purposes for observation.

References

Saudargas, R. A. & Lentz, F. E. (1986). Estimating percent of time and rate via direct observation: A suggested observational procedure and format. School Psychology Review, 15, 36-48.

Bell, D. R. & Beedle, B. B. (1993). Observing and Recording Children's Behavior. Kendall/Hunt Publishing

Ron Dumont is Associate Professor of Psychology and Director of the MA and Psy.D. programs in School Psychology at Fairleigh Dickinson University. Ron is also a contributing editor for the Communiqué

Sandy Chafouleas is Assistant Professors of Psychology and director of the SUNY-Plattsburgh School Psychology program..