DIRECT AND INDIRECT MEASURES OF STUDENT BEHAVIOR
Functional behavioral assessment can be a time-consuming process, one that usually is best accomplished in stages. As discussed in Step 2, the functional behavioral assessment process may begin with a series of initial direct and indirect observations (e.g., using a scatterplot) and/or discussions with adults or students who have witnessed the behavior (e.g., functional interviews). An examination of the information from these observations and interviews may suggest specific times and settings in which to conduct more thorough observations (e.g., during a specific academic subject or class period). These subsequent observations would lead the IEP team to develop an hypothesis statement regarding the factors that are most predictive of the students behavior (e.g., a science lesson that requires lengthy silent reading of technical material). Both direct and indirect measures of student behavior are described more thoroughly in this section.Direct Assessment Direct assessment consists of actually observing the problem behavior and describing the conditions that surround the behavior (its context). This context includes events that are antecedent (i.e., that occur before) and consequent (i.e., that occur after) to student behaviors of interest. There are several tools to select from in recording direct assessment data. Each has its particular strength. IEP teams should consider what they want or need to know about the presenting behavior and select direct observation strategies and recording tools accordingly. A description of the most commonly used tools and the kinds of data they can help gather follows.
Scatterplots. Often, initial observations can be accomplished through the use of a scatterplot (see Appendix A for sample scatterplot forms). The purpose of a scatterplot is to identify patterns of behavior that relate to specific contextual conditions. A scatterplot is a chart or grid on which an observer records single events (e.g., number of student call-outs) or a series of events (e.g., teacher requests and student responses) that occur within a given context (e.g., during teacher-led reading instruction, at lunch, on the playground). Scatterplots take various forms, depending on the behavior of interest and its social/physical context. Some require observers to sequentially record (by category) various events (e.g., format of instruction, teacher behavior, student/peer responses, likely purpose of student reaction).
ABC charts. Another way to observe student behavior is with an Antecedent-Behavior-Consequence (ABC) chart (also referred to as an Antecedent-Response-Consequence or ARC chart) (see Appendix B for examples of ABC charts). This approach allows an observer to organize anecdotal or descriptive information on the students interactions with other students and adults in such a way that patterns of behavior often become clear. A modified ABC chart might be individualized to contain several predetermined categories of teacher or peer antecedent behavior, student responses, and consequent events, along with space for narrative recording of classroom observations.
Using scatterplots and ABC charts together. By using the ABC procedure, the student may be observed in settings and under conditions where the behavior is most likely and least likely to occur. A scatterplot to chart the relationship between specific types of instruction and the students appropriate/inappropriate responses may also be developed.
A scatterplot can be developed to observe and record the relationship between a specific set of classroom variables (e.g., teacher lecture and student off-task behavior) or playground behaviors and to analyze a particular situation. For instance, out-of-seat behavior might be measured in increments of 1-5 minutes, while fights on the school bus may be recorded daily (e.g., critical incident reports). Furthermore, student behavior may be a function of specific teacher-pupil interactions (e.g., there may be a relationship between teacher reprimands and student outbursts). Observing and recording teacher-pupil interactions may lead to a better understanding of the relationship between these factors of classroom interactions. Both the ABC and scatterplot procedures are useful in identifying environmental factors (e.g., seating arrangements), activities (e.g., independent work), or times of the day (e.g., mornings) that may influence student behavior.
Both ABC and scatterplot recording procedures are useful not only in identifying problem behavior, but also in identifying the classroom conditions that may trigger or maintain the students behavior. It is also important to observe situations in which the student performs successfully so that IEP teams can compare conditions and identify situations that may evoke and maintain appropriate rather than inappropriate behavior (e.g., in science class as opposed to language arts class). In this way, it is possible to get a clearer picture of the problem behavior, determine the critical dimensions of the behavior, write a precise definition of the behavior, select the most appropriate assessment tools, and develop an effective intervention plan for changing the behavior.
As we already mentioned, multiple measures of student behavior and its social/ environmental contexts usually produce more accurate information than a single measure. This is especially true if the problem behavior serves several functions or purposes that may vary according to circumstance. In our previous example of Mandys wisecracks, making inappropriate comments during lectures may serve in some instances to get her something (e.g., peer attention). In another classroom, the same behavior may help her to avoid something (e.g., being called on by the teacher). Information gathered through repeated observations of Mandy across settings will enable the IEP team to distinguish among the various purposes for her inappropriate remarks.
Amount versus quality of behavior. Different types of behavior may require different data collection techniques. For example, it is important to know how often a behavior occurs (e.g., call-outs); in this case, a system that yields the number of behaviors, or frequency measure, is appropriate. At other times, knowing how long the behavior occurs is more relevant (e.g., out-of-seat), so that a duration measure becomes more useful. Furthermore, the usefulness of documenting the severity or intensity of a behavior is evident when the IEP team tries to measure other disruptive behaviors. To say that Charles was upset two times yesterday may not reflect the fact that he succeeded in disrupting instruction in the entire middle school wing for a total of 45 minutes.Severity of Disruptive Behavior Rating Rubric
In some cases, it is useful to report the severity and measure of a behavior using a rubric to capture the magnitude and/or amount of variation in the behavior. This is true with regard to both student and adult behavior. That is, a student tantrum may be minor or extreme and of short or long duration. Teacher reprimands might be insignificant except when they are repeatedly and loudly delivered to the student for an extended amount of time. The following rubric could be used to observe and record the severity of a students disruptive behavior.Indirect Assessment We know that student behavior usually is related to the context in which it occurs. However, the IEP team will not always be able to directly observe all the events that bring about or maintain specific student behavior. So-called "setting events" (sometimes referred to as slow triggers) can exist within the classroom (e.g., Charles is asked to join a new reading group), or be far removed from it but still exert a powerful influence over student behavior (e.g., Charles has an argument with another student at the bus stop before school). External events of this nature may increase the likelihood of conflict in the classroom, especially if the student is struggling academically and/or dislikes the subject matter. These setting events (or specific antecedents for the behavior) often may not be directly observable. In other cases, the behavior may be serious but not occur frequently enough in settings accessible to adults to be readily observed (e.g., verbal or physical aggression). In these instances, the behavior must be assessed by using indirect measures.
Methods of indirect assessment. Indirect or, as it is sometimes called, informant assessment, relies heavily on the use of interviews with teachers and other adults (e.g., bus drivers, cafeteria workers, office staff) who have direct contact with the student. (See Appendix C for a sample interview form.) In addition, a semi-structured interview with the student, himself, could provide insight into the students perspective of the situation and yield a more complete understanding of the reasons behind the inappropriate behavior. It may be useful to follow the same interview format with both the student and significant adults (e.g., special and regular classroom teachers, support personnel) and to compare these two sources of information. Even elementary aged students can be credible informants, capable of sharing accurate information about contextual factors that influence their behavior. Indirect measures can yield valuable information, but they usually are not as reliable as direct observation measures. For this reason, IEP teams must be careful not to put too much faith in information derived from informant accounts alone. Examples of interviews conducted with teachers and students to help determine the likely function of a students behavior are included in Appendix D.
Surveys or questionnaires are another source of indirect information. For example, a Problem Behavior Questionnaire can be administered to one or more teachers who have day-to-day contact with a student of concern (see Appendix E for sample Problem Behavior Questionnaire forms). Recalling a typical behavioral episode, teachers read 15 statements and circle a number on the questionnaire that corresponds to the percent of time each statement is true for that student. A second form is used for recording and interpreting the responses from everyone who completed a questionnaire for that student. Any item marked with a three or above on this profile form suggests the potential function of the problem behavior. If there are two or more statements scored as three or above (i.e., (50% of the time) under a particular sub-column (e.g., escape under peers or attention under adults), then it may indicate a possible primary function of the behavior.
In collecting information regarding the context of a behavior problem, it is important to understand that contextual factors may include certain affective or cognitive behaviors, as well. For instance, Juan repeatedly acts out and is verbally threatening during instruction when given lengthy and difficult assignments. Even so, it may not be the assignment itself that triggers the acting-out behavior. Rather, it may be the fact that he knows he doesnt have the skills necessary to complete the work that prompts an anticipation of failure or ridicule. Or, he may have a family member who is critically ill; therefore, he finds it difficult to concentrate.Accuracy of Behavior Measurement There are a number of ways that accuracy in observing and recording student behavior and the social/environmental conditions that surround it can be jeopardized. Common problems include:
In the end, the usefulness of functional behavioral assessment depends on the skills and objectivity of the persons collecting the information. Accordingly, if the information is to be helpful to IEP teams, it must be reliable and complete information about the behavior. Those conducting the functional behavioral assessment must: a) clearly define the behavior of concern and regularly review that definition; b) have sufficient training and practice to collect observation and interview data; c) select the most appropriate assessment procedure(s) for both the behavior and the context; d) collect information across time and settings using multiple strategies and individuals; and, e) conduct routine checks of the accuracy of observer scoring/recording procedures.