Does Latency in Recording Data Make a Difference? Confirming the Accuracy of Teachers’ Data

Abstract

The effects of latency on the accuracy of data recorded by three special education teachers were examined in this study. Teachers recorded data on the target behaviors of three students with varying disabilities. The accuracy of data recorded was assessed during three time periods: immediately after the target behavior occurred, at the end of the school day, and the following school day. A multielement design was used to evaluate data accuracy. Results were interpreted to confirm that data recorded immediately after a behavior occurred were more accurate and reliable than data documented at the end of the school day or the start of the following school day. In addition, data recorded by each teacher had a mean agreement of 97% or above for the time period immediately after a student’s behavior occurred. Furthermore, each teacher reported that it was beneficial to record data immediately after the target behavior occurred. Implications and future research directions are provided.

Keywords

data collection latency accuracy reliability direct measurement

The Individuals With Disabilities Education Improvement Act (2004) requires special education teachers to develop Individualized Education Programs (IEPs) containing measurable annual goals and short-term objectives, and to include a statement of how progress toward meeting goals and objectives is being met, 34 C.F.R. § 300.347(a)(7). Concomitantly, the No Child Left Behind Act emphasizes that teachers implement scientifically based practices and procedures supported by research (Burns & Ysseldyke, 2009). Although educators are under tremendous pressure to engage in evidence-based practices, it is often ignored (Boardman, Arguelles, Vaughn, Hughes, & Klingner, 2005; Simonsen et al., 2010). Evidence-based practices are those that include scientifically based activities or those with data that support their work (U.S. Department of Education, 2002). The “evidence” in evidence-based practice is of particular importance to special education teachers as the data-collection methods used in classrooms have the potential to sustain the effective instructional practices currently used or immediately identify ineffective practices. Specifically, data play a role in the development, implementation, and promotion of evidence-based practices (McDuffie & Scruggs, 2008). Teachers gather assessment data to increase student success and document the effectiveness of an instructional practice over time, lending support to a particular method (e.g., evidence; Babkie & Provost, 2004). As such, special education teachers must become knowledgeable about their role in the research process and become active and accurate data collectors (Simonsen et al., 2010).

Making appropriate instructional decisions to enhance student progress in achieving IEP goals and objectives requires teachers to regularly document student performance (Brown, Snell, & Lehr, 2006; Horner, Sugai, & Todd, 2001). Inaccurate data recording may lead to incorrect conclusions about student progress and intervention effectiveness (Parson & Baer, 1986) and affect student achievement (Wolery, Bailey, & Sugai, 1988). Although data are necessary to provide evidence for decision making, special education teachers frequently report difficulties with this process (Sandall, Schwartz, & Lacroix, 2004). Researchers of several studies cited special education teachers’ belief that data collection is valuable (Cooke, Heward, Test, Spooner, & Courson, 1991; Ysseldyke, Nania, & Thurlow, 1985). Yet, teachers report infrequently or rarely engaging in data collection or using data to change their instruction (Farlow & Snell, 1989; Rathgen, 2006; Walton, 1985). Special education teachers identified numerous data-collection barriers, including a lack of knowledge and skills for how and when to collect data, time management for collecting data, analyzing data, and appropriately using the data for instructional decision making (Sandall et al., 2004; Ysseldyke et al., 1985). In fact, rather than relying on data, teachers reported making educational decisions based on their own judgments, memories, individual observations, and the influence of conversations and collaborations with fellow teachers (Boardman et al., 2005). Although individual judgments and observations are important factors in the data-collection process, accurate recording increases the likelihood that instructional decisions will ultimately result in improved instruction and student progress.

One recommended practice for accurately measuring student learning requires direct observation of student performance (Cooke et al., 1991; Farlow & Snell 1989). To make accurate judgments about instructional effectiveness and improve student performance, teachers should frequently record and regularly analyze those data (Cooke et al., 1991; Farlow & Snell 1989; Horner et al., 2001). Unfortunately, many teachers find direct observation and data collection to be too time-consuming and a task that interferes with their teaching responsibilities (Jones, 2009; Munger, Snell, & Loyd, 1988). Measuring student behaviors only once provides limited information, whereas measuring performance repeatedly and over time allows teachers to more accurately determine the effectiveness of instruction and student learning (Gable, Arllen, Evans, & Whinnery, 1997; Horner et al., 2001). Thus, data reliability can be directly influenced by measurement frequency, suggesting that it may be necessary to record student behaviors more often (Fuchs, 1986; White & Haring, 1980). In addition, teacher’s judgments about student performance may be less accurate than those based on direct observation (Lewis-Palmer, Sugai, & Larson, 1999; White & Liberty, 1976). Data might be inaccurate if they are recorded at some point after a student completes the target behavior when there is no permanent product (Lewis-Palmer et al, 1999; Logan, 1991). However, when permanent products are unavailable, not only do the frequencies with which data are recorded become noteworthy but so does the latency between student behavior and data collection.

Only one study could be found in which latency as a factor in data accuracy was examined when recording data. Logan (1991) examined whether data recorded by four special education teachers at three different time periods (e.g., immediately after a 10-min instructional sequence, after a 30-min instructional session involving two additional students, and at the end of the day) after their students with profound disabilities completed targeted behaviors would be accurate. Results were interpreted to conclude that data were most accurate when recorded following the 10- and 30-min instructional sessions. When teachers waited until the end of the day to record student performance levels, accuracy decreased. Logan concluded that teachers did not need to engage in simultaneous data collection while teaching. Rather, data were most accurate when the latency between behavioral occurrence and recording was brief. Logan noted that although the performance levels of students with profound intellectual disabilities may not fluctuate much across sessions or days, future investigators should examine the data-collection accuracy of teachers who serve students with moderate and severe disabilities, whose behaviors may change more frequently.

We sought to replicate Logan’s examination of latency effects on the accuracy of data recorded by special education teachers. However, unlike the Logan study where teachers worked individually with students who experienced profound intellectual disabilities, we focused on students with mild and moderate intellectual and behavioral disabilities who were engaged in individual and group classroom activities throughout the school day. In the current study, three special education teachers recorded data on target students in each of their classes. For each student, three target behaviors were recorded at different times (e.g., immediately after the behavior occurred, at the end of the school day, at the start of the following school day) to confirm the impact of latency on data accuracy. Students were engaged in typical classroom activities, including individual, small group, and one-on-one learning sessions with peers and adults.

Method

Participants

Three special education teachers, Suzie, Neelu, and Karol (pseudonyms), participated in this study. Each was selected to participate based on their (a) holding a current teaching license, (b) employment as a special education teacher, and (c) willingness to participate. Each held at least a bachelor’s degree in special education and taught secondary students (ages 12–19 years) with varying disabilities. All reported that they were actively engaged in regularly recording data based on student instructional and behavioral objectives. Table 1 presents a demographic summary for each teacher participant.

Table 1.

Teacher and Student Demographic Information.

Teacher/student	Degree	Number of years teaching	Classroom type	Number of students	Type of data collection	Disability	Age (years)
Teacher
Suzie	BA/Ed minor: reading	12	Communication/moderate and severe	8	Individual student (1×–2×/week)
Neelu	BA/K-12 SE, LD, Mild/Mod	8	Academic instruction	13	Individual student (1×–2×/week)
Karol	MS/K-12 SE	2	Functional life skills	12	Individual student (1×–2×/week)
Student
Jaxon						DS, MoID	19
Pete						ODD, ADHD	14
Daniel						BPD, MID	13

Note: BA/Ed = Bachelor of Arts in Elementary Education; BA/SE = Bachelor of Arts in Special Education; MS/SE = Master of Science in Special Education; LD = emphasis in learning disabilities; Mild/Mod = emphasis in mild/moderate disabilities; DS = Down’s syndrome; MoID = moderate intellectual disability; ODD = oppositional defiant disorder; ADHD = attention deficit hyperactivity disorder; BPD = bipolar disorder; MID = mild intellectual disability.

Suzie

Suzie taught students identified with communication and intense learning needs (e.g., students who experienced moderate/severe autism, students who were nonverbal, students who were considered legally blind). Her students ranged in age from 12 to 19 years. She taught at the school for approximately 2 years. Her previous experience included teaching first and fourth grades. Prior to her current position, she served as a media specialist in several public schools for 10 years.

Neelu

Neelu taught academic skills (e.g., language arts/English, math, science) to students, ages 12 to 15 years, identified with oppositional defiant disorder, conduct disorder, attention-deficit/hyperactivity disorder, and mild intellectual disabilities. She had worked in the present school for more than 4 years prior to the study. Her previous teaching experience included working as a substitute teacher, paraprofessional, and elementary schoolteacher in a self-contained classroom for students with emotional disorders for more than 3 years.

Karol

Karol taught functional life skills to students, ages 13 to 19 years, identified with mild and moderate intellectual disabilities. She had worked in the present school for 2 years prior to the study. Preceding her current position, Karol had no prior teaching experience.

Three students who attended each teacher’s class also participated in this study. Jaxon, Pete, and Daniel (pseudonyms) were selected as target students from whom teachers would collect learning and behavioral data during this investigation based on (a) being a student in one of the participating teacher’s classrooms, (b) parental consent, and (c) their willingness to participate based on informed verbal assent. Jaxon was a student in Suzie’s class, Pete attended Neelu’s class, and Daniel was a student in Karol’s class. Table 1 provides demographic information on each of these students.

Settings

Baseline and intervention sessions took place within each teacher’s classroom in a residential school located in a suburban midwestern community. The residential school specialized in providing year-round treatment, and educational and vocational services to school-age students (6–21 years) who were dually diagnosed with developmental and emotional disabilities. Each teacher’s classroom consisted of no more than 15 students and contained a desk or table for each student. Each classroom contained two windows, a closet, and a computer. In addition, Suzie’s classroom had a water table allowing students to play with water toys or other materials (e.g., rice, beans, sand) for tactile stimulation. Two paraeducators were present in each classroom setting.

Dependent and Independent Variables

The dependent variable in this investigation was the accuracy at which each teacher recorded student performance data. The independent variables addressed latency. Specifically, three independent variables were randomly alternated that determined the time at which each teacher recorded student learning and behavioral data: immediately after the behavior was performed, at the end of the school day, or the morning of the following day. Teachers identified three specific behaviors for their students and randomly assigned the time period in which a particular behavior would be targeted for data collection. Information regarding when data were collected for each target behavior per student is presented in Table 2.

Table 2.

Target Behaviors and Data Recording Time Periods.

	Data recording time period
Student	Immediately	End of day	Following day
Jaxon	Refusing to follow directions after an initial directive	Not following directions during calendar activities	Number of requests (asking for food, people, books, etc.)
Pete	Out of bounds (leaving classroom without permission)	Instigating peers (talk to or touch in ways that irritated them)	Not following directions
Daniel	Not following directions	Inappropriate boundaries (touched peers or staff without their permission)	Negative behaviors (e.g., yelling at peers/staff, hitting walls or desk, pushing over or kicking chairs)

Experimental Design

A multielement design was used to illustrate the accuracy of teacher data for each student’s target behaviors. This design was selected as it allowed for an immediate comparison between the three data-collection time periods to determine if one resulted in greater accuracy than another (Barlow & Hersen, 1984; Kazdin, 1982). Independent variables were alternated and counterbalanced to control for carryover effects. Each teacher was asked to collect data during each of the three designated time periods (e.g., immediately after a behavior occurred, at the end of the school day, or the start of the following school day after behavior occurred). To determine the order of data collection, each time period was represented by a number. For example, the time period “immediately after a behavior occurred” was assigned the numbers 0, 3, and 6. Likewise, the time period “at the end of a school day” was assigned the numbers 1, 4, and 7, and the time period “at the start of the following day” was assigned the numbers 2, 5, and 8.

Each teacher was asked to select a number between 0 and 8. If they chose 0, 3, or 6, they would begin collecting data immediately after the behavior occurred. If they chose 1, 4, or 7, they would begin collecting data at the end of the day. Furthermore, if they chose 2, 5, or 8, they would begin collecting data at the start of the following day. Suzie chose the number 0 and started data collection immediately after the behavior occurred. Neelu chose the number 7 and started data collection at the end of the school day. Karol chose the number 8 and started data collection at the start of the following day.

Data Collection

Event recording was used to document the number of times students exhibited target behaviors. Teachers recorded data at a designated time period (e.g., immediately after a behavior occurred, at the end of the school day, or the start of the following school day after behavior occurred). Investigators collected data only during the instructional periods. During analysis, the numbers recorded by teachers and investigators were compared. Event recording was selected as it was the most direct and accurate way to reflect the number of times student’s target behaviors occurred (Alberto & Troutman, 2012).

Procedures

Preintervention

During this phase, investigators reviewed data-collection procedures with teachers. Teachers were video recorded as they taught their target students during the same instructional period each day for a minimum of 5 days. They were asked to engage in activities as they normally would within their classroom (e.g., one-on-one, individual group work, or small-group activities) that included recording data on their student’s behaviors. The teachers chose the instructional (e.g., observational) period because, it was assumed, they would have insight regarding when the student would exhibit possible target behaviors. Specifically, the teachers had knowledge of what periods of the day a student exhibited a behavior in an excessive manner (e.g., after lunch because the student was tired, during transitions). Each teacher’s observational period was approximately 45 min in length (range = 35–45 min; µ = 41.87 min). Teachers identified three specific target behaviors per student, operationally defined, and recorded data on those behaviors.

Intervention

This phase lasted 15 to 25 days for each teacher. Each teacher collected data on three regularly occurring target behaviors for their student (see Table 2). Teachers collected data on behaviors of target students while other students were in the classroom. Teachers were engaged in one-on-one activities, such as spelling, with target students, which allowed target students to work alone with the teacher. In addition, teachers engaged in individual group work or small-group activities, such as math, English, and physical education, with other classmates. Thus, target students were required to interact with their classmates during data recording.

Due to the nature of special education classrooms, the conditions varied under which teachers recorded data in their classrooms. For example, Suzie observed her students during the early morning (9:00–9:35 a.m.) and often began the instructional period with small-group instruction, whereas paraeducators worked individually with two students. After 15 min, the entire class came back together and Suzie engaged students in direct instruction. Neelu observed her students in the early afternoon (12:45–1:15 p.m.) and frequently engaged them in direct instruction for a majority of the period, reserving the last 15 min for individual work. Paraeducators were available as students needed them (i.e., when they raised their hand for help). Karol observed her students during the latter part of the morning (10:00–10:45 a.m.) and engaged them in small-group activities for the duration of the instructional period, whereas paraeducators led some small groups. Thus, each teacher used varying classroom formats and conditions while recording data. A summary of the conditions each teacher used in her classroom while recording data is presented in Table 3.

Table 3.

Description of Classroom Conditions During Teachers’ Instructional Periods.

Condition	Suzie	Neelu	Karol
Instructional period	9:00–9:35 a.m.	12:45–1:15 p.m.	10:00–10:45 a.m.
Type of instruction	Small group	Direct instruction	Small group
Number of students	6	13	6–8^a
Activity	Calendar time	Spelling	Mathematics

The number of students in her small group varied daily due to the group assignment students received.

Interobserver Agreement and Treatment Fidelity

To ensure reliability, interobserver agreement (IOA) data were collected across all of the intervention sessions by two trained independent observers. Agreement was calculated using a total percentage agreement method by dividing the smaller number of events observed by the larger number and then multiplying this number by 100%. IOA averaged 95% and ranged from 90% to 100% for Suzie (Teacher 1). Agreement was 100% during each IOA session for Neelu (Teacher 2). Finally, IOA averaged 90% for Karol (Teacher 3), with a range of 85% to 95%.

Treatment fidelity measures confirmed the teachers’ correct use of the data collection across each time period. To ensure the validity of intervention, a task analysis was used to record the percentage of steps (e.g., observe the behavior and mark the instance on the data-collection sheet) each teacher correctly followed when collecting data on the target behaviors for each student (see the appendix for task analysis). Treatment fidelity measures were conducted during 30% of intervention sessions. For Suzie (Teacher 1), treatment fidelity was 97% (range = 94%–100%), and 100% for Neelu (Teacher 2) and Karol (Teacher 3).

Results

Figures 1 to 3 illustrate the accuracy of each teacher’s data when recorded immediately after a behavior occurred, at the end of the school day, and at the beginning of the following school day. Visual analysis revealed that data recorded immediately after a student behavior occurred were more accurate than those recorded at the end of the day or the following day for all three teachers. In addition, all teachers had more agreement per session with the researchers during the time period immediately after the target behavior occurred than any other time period.

Figure 1.

Percentage of agreement per session per time period in which Suzie gathered data on student’s target behaviors.

Figure 2.

Percentage of agreement per session per time period in which Neelu gathered data on student’s target behaviors.

Figure 3.

Percentage of agreement per session per time period in which Karol gathered data on student’s target behaviors.

Suzie

Suzie’s data were more accurate and reliable when collected immediately after the target behavior occurred. Figure 1 illustrates the percentage of agreement per session per time period on the target behaviors of her student. The mean agreement on the number of target behaviors exhibited by the student was 98% during the time period immediately after a behavior occurred. The mean agreement for the time period at the end of the school day was 74%, whereas it was 75% for the time period at the start of the following day. The mean difference between immediately after the behavior occurred and at the end of the day was 24%, whereas the mean difference was 23% between immediately after the behavior and the next day. The mean difference between end of the day and data recorded the next day was 1%. These differences along with visual analysis suggest that the time period immediately after a student’s behavior occurred proved to be more accurate and reliable when compared with the other time periods.

Neelu

Neelu’s data were more accurate and reliable when collected immediately after the target behavior occurred. Figure 2 illustrates the percentage of agreement per session per time period on the target behaviors of her student. The mean agreement for the number of target behaviors exhibited by the student was 100% during the time period immediately after a behavior occurred. The mean agreement for the time period at the end of the school day was 71%, whereas it was 14% at the start of the following day. The difference between the means between immediately after the behavior occurred and the end of the school day was 29%, whereas the mean difference was 86% between immediately after the behavior occurred and the following day. The mean difference between the end of the day and data recorded the next day was 57%. These differences along with visual analysis suggest that data recorded immediately after a student’s behavior occurred proved to be more accurate and reliable when compared with the other time periods.

Karol

Karol’s data were more accurate and reliable when collected immediately after the target behavior occurred. Figure 3 illustrates the percentage of agreement per session per time period on the target behaviors of her student. The mean agreement for the number of target behaviors exhibited was 97% immediately after a behavior occurred. The mean agreement at the end of the school day was 37%, whereas it was 43% the following day. The difference between the means between immediately after the behavior occurred and the end of the school day was 60%, whereas the mean difference was 54% between immediately after the behavior occurred and the following day. The mean difference between the end of the day and data recorded the next day was 6%. These differences suggest that the time period immediately after a student’s behavior occurred provided more accurate and reliable data when compared with the other time periods.

Social Validity

When asked whether they were surprised at the results of the study, each teacher stated that they were not. Teachers reported that it was beneficial and made more sense to record data immediately after a behavior occurred, rather than waiting until the end of the day. They stated it was more beneficial to them to record data immediately after the behavior occurred because it was “fresh” in their minds and more accurate. However, one teacher stated that recording data immediately after a behavior occurred was not always the most feasible method. She stated she would “often have to stop in the middle of her lesson to write down the data she was recording.”

All three teachers expressed experiencing difficulties remembering what data to record while dealing with other issues in their classrooms (e.g., dealing with student behaviors, teaching lessons). They reported it was difficult to remember the data they were supposed to record at the end of the day and from the previous day. They found it less beneficial to record data at the start of the following day because they forgot what data to record and often guessed at the number of times they thought their student exhibited a target behavior the previous day.

Each teacher reported that she would attempt to record data immediately after students’ behavior occurred in the future. All three noted that collecting data immediately after a behavior occurred made it easier for them to document students’ progress on their IEP goals and objectives. However, one teacher reported that recording data every time she observed an instance of a student’s target behavior was “time-consuming.” Furthermore, she noted that it was “somewhat overwhelming.” However, all teachers reported that their participation was a valuable experience.

Discussion

The purpose of this investigation was to examine latency effects on the accuracy of data recorded by three teachers on the target behaviors of the three students. Data on each student’s target behaviors were recorded at three different time periods: immediately after the target behavior occurred, at the end of the school day, and at the start of the following school day. Results are interpreted to indicate that the data recorded immediately after a student’s target behavior occurred were more accurate than any other time period for all teachers. Likewise, each teacher had a higher percentage of agreement with the researcher for data recorded during the time period immediately after the student’s target behavior occurred than at any other time period.

In addition, data recorded by each teacher had a mean agreement of 97% or more for the time period immediately after a student’s behavior occurred. However, each teacher had a mean agreement between 37% and 74% for the time period at the end of the school day after a student’s behavior occurred and between 14% and 75% for the time period at the start of the following day after a student’s behavior occurred. There are several factors possibly influencing the accuracy of the data recorded by teachers during the time periods at the end of the school day and at the start of the following day. First, we focused on a population of students with mild disabilities in this study as opposed to Logan (1991) who included students with profound disabilities. Students with mild disabilities engage in more behaviors making accuracy more challenging if teachers wait to record data. This may be because teachers are being attentive to the various behaviors that their students are displaying, affecting the accuracy of the data recorded on the specific behavior being targeted. Second, teachers in this study were required to teach in their classrooms as they normally would. This included working on activities with small groups of students, engaging students in individual group work, or working individually with students. Teachers’ participation in these activities may have affected the accuracy of the data they recorded, particularly because their attention was shared between the activities with students and the behavior being targeted. In addition, the length and number of activities teachers were engaged in prior to data recording may have affected data accuracy.

Clearly, immediacy in data recording was unmatched. The findings indicated that each teacher experienced more success when they recorded data immediately after students’ target behavior occurred. These data were more accurate and reliable during this time period because teachers did not have to retain the information for an extended period and were required to record the information immediately after it was observed. Furthermore, this allowed teachers to provide an accurate representation of the number of times each student exhibited a target behavior. This same strategy could also be used when collecting data on a student’s academic progress, a student’s IEP goals and objectives, or the effectiveness of instructional practices.

Although results are interpreted to support the need for educators to record data immediately after the behavior occurs, there was some variability in the teacher’s IOA results. Neelu demonstrated a higher percentage of agreement for the time period at the end of the day than the time period at the start of the following day. This may be explained by the time of day at which she observed her student’s behavior (12:45–1:15 p.m.). Thus, she experienced the shortest time lapse between behavior observation and the documentation of observed behavior at the end of the day. It is unknown why Suzie and Karol obtained a higher percentage of agreement for the time period at the start of the following day. One would expect data collected at the end of the day to be more accurate than data recorded at the start of the following day. However, this investigation did not produce such results. In the future, researchers should consider further examining the cause of the variability between those two time periods.

Our results can be used to support the assumption that data recorded immediately after a student’s behavior occurs will be more accurate than data collected later in the school day or the following school day. Furthermore, it supports researchers suggesting direct measurement as an effective means of evaluating instructional effectiveness and increasing student performance (Cooke et al., 1991; Gable et al., 1997; Horner et al., 2001; Munger et al., 1988). Replications of the present investigation are needed to confirm these results.

Limitations, Implications, and Future Research

Teachers in the present study engaged in classroom activities and recorded data on their students as they typically would on a given day. However, they also focused data recording on their target student’s predetermined behavior that occurred during the specific instructional period. For example, Suzie taught and recorded performance data for various students on calendar-related activities during the morning instruction period. Concomitantly, she also observed her target student’s behaviors (e.g., his number of requests, not following directions during calendar, not following instructions after initial directive). Depending on the intervention schedule, Suzie would record data on one of these behaviors immediately, at the end of the day or at the beginning of the next day. Although paraeducators were available to assist with data recording on calendar-related skills, Suzie also recorded these data while leading the instruction. Thus, one limitation possibly affecting results may be attributed to the teacher’s roles. The number and type of instructional activities in which the teacher was engaged may have contributed to the accuracy or inaccuracy of data recorded on each student’s target behavior. As such, future investigators might examine how variations of these variables may influence the accuracy of teacher data.

One limitation affecting generalizability may be attributed to the number and diversity of participants included in the current investigation. We involved only three special education teachers who taught in the same school, and taught students with mild and moderate exceptional learning needs. To increase generalizability, future investigators may consider the data-collection accuracy of teachers serving students who represent different disability populations as well as general education teachers who teach students with and without disabilities in inclusive settings.

Another limitation affecting the generalizability of results was the use of event recording as a means to document students’ behavior. The findings from this study are generalizable only to the event recording method of collecting data. To increase generalizability, researchers may want to consider conducting this investigation using alternative data recording methods such as duration or latency recording. In the future, investigators may seek to examine the use of the event recording method to increase or accelerate students’ target behaviors.

A final limitation was the use of the video camera to record student behavior. Although the video camera was placed in each student’s classroom 2 weeks prior to data collection to reduce novelty effects, the presence of the video camera potentially could have affected the students’ performance causing an increase or decrease in their behavior. For instance, Neelu’s (Teacher 2) student Pete was observed instigating his peers excessively on many occasions prior to the implementation of the intervention; however, these behaviors remained low throughout the study. This result could be attributed to the presence of the video camera in the classroom.

The evidence provided from our investigation, in addition to the dissertation investigation conducted by Logan (1991), suggests that collecting data immediately after a behavior occurred, or soon after, was more effective because data were likely to be more accurate and more reliable. These accurate and reliable data will provide educators with a better understanding of student progress and achievement. For educators, a primary concern should be student academic progress. To ensure students are gaining the skills and knowledge necessary to help them achieve academically, accurate documentation of their progress (e.g., data collected immediately after the behavior/performance occurs) is necessary. This will ensure that educators are able to base instructional decision on accurate information.

Researchers should consider examining the use of data recording with different time periods (e.g., 10 min after a behavior occurs, 30 min after a behavior occurs, and 1 hr after a behavior occurs) to further indicate, precisely, how much time can lapse between student performance and teachers data collection before the data are no longer valid. Researchers also should consider evaluating the strength of the data teachers record and how it influences their instructional decisions; specifically, whether instructional decisions are being made with substantiated evidence (e.g., accurate and reliable data). In the future, researchers also should consider examining the impact of teachers’ data-driven instructional decisions on student achievement as well as examining the accuracy of students recording their own performance data.

Footnotes

Appendix

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Alberto

P. A.

Troutman

A. C.

(2012). Applied behavior analysis for teachers (9th ed.). Upper Saddle River, NJ: Prentice Hall.

Babkie

A. M.

Provost

M. C.

(2004). Teachers as researchers. Intervention in School and Clinic, 39, 260–268.

Barlow

D. H.

Hersen

(1984). Single case experimental designs (2nd ed.). Elmsford, NY: Pergamon.

Boardman

A. G.

Arguelles

M. E.

Vaughn

Hughes

M. T.

Klingner

(2005). Special education teachers views of research-based practices. Journal of Special Education, 39, 168–180.

Brown

Snell

M. E.

Lehr

(2006). Meaningful assessment. In Snell

M. E.

Brown

(Eds.), Instruction of students with severe disabilities (6th ed.) (pp. 173-206). Upper Saddle River, NJ: Pearson.

Burns

M. K.

Ysseldyke

J. E.

(2009). Reported prevalence of evidence-based instructional practices in special education. Journal of Special Education, 43, 3–11.

Cooke

Heward

W. L.

Test

D. W.

Spooner

Courson

F. H.

(1991). Student performance data in the classroom: Measurement and evaluation of student progress. Teacher Education and Special Education, 14, 155–161.

Farlow

L. J.

Snell

M. E.

(1989). Teacher use of student performance data to make instructional decisions: Practices in programs for students with moderate to profound disabilities. Journal of the Association for Persons With Severe Handicaps, 14, 13–22.

Fuchs

L. S.

(1986). Monitoring progress among mildly handicapped pupils: Review of current practice and research. Remedial and Special Education, 7, 5–12.

10.

Gable

R. A.

Arllen

N. L.

Evans

W. H.

Whinnery

K. M.

(1997). Strategies for evaluating collaborative mainstream instruction: “Let the data be our guide.” Preventing School Failure: Alternative Education for Children and Youth, 41, 153–158.

11.

Horner

R. H.

Sugai

Todd

A. W.

(2001). “Data” need not be a four-letter word: Using data to improve schoolwide discipline. Beyond Behavior, 11, 20–26.

12.

Individuals With Disabilities Education Improvement Act, 20 U.S.C. § 1400 et seq. (2004).

13.

Jones

M. L.

(2009). A study of novice special educators’ view of evidence-based practices. Teacher Education and Special Education, 32, 101–120.

14.

Kazdin

A. E.

(1982). Single case research designs: Methods for clinical and applied settings. New York, NY: Oxford University Press.

15.

Lewis-Palmer

Sugai

Larson

(1999). Using data to guide decisions about program implementation and effectiveness. Effective School Practices, 17(4), 47–53.

16.

Logan

(1991). An analysis of the accuracy of student performance data recorded by teachers under three conditions (Unpublished doctoral dissertation). Georgia State University, Atlanta.

17.

McDuffie

K. A.

Scruggs

T. E.

(2008). The contributions of qualitative research to discussions of evidence-based practice in special education. Intervention in School and Clinic, 44, 91–97.

18.

Munger

Snell

M. E.

Loyd

B. H.

(1988). How often do you need to collect student performance data? A study of the effects of frequency of probe data collection and graph characteristics on teachers’ visual inference. Retrieved from ERIC database. (ED310560)

19.

Parson

B. S.

Baer

D. M.

(1986). The graphic analysis of data. In Poling

Fuqua

R. W.

(Eds.), Research methods in applied behavior analysis (pp. 157–184). New York, NY: Plenum.

20.

Rathgen

(2006). In the voice of teachers: The promise and challenge of participating in classroom-based research for teachers’ professional learning. Teaching and Teacher Education, 22, 580–591.

21.

Sandall

S. R.

Schwartz

I. S.

Lacroix

(2004). Interventionists’ perspectives about data collection in integrated early childhood classrooms. Journal of Early Intervention, 26, 161–174.

22.

Simonsen

Shaw

S. F.

Faggella-Luby

Sugai

Coyne

M. D.

Rhein

. . . Alfano

(2010). A schoolwide model for service delivery: Redefining special educators as interventionists. Remedial and Special Education, 31, 17–23.

23.

U. S. Department of Education. (2002). Evidence-based education. Paper presented at the Student Achievement and School Accountability Conference, October, 2002. Retrieved from http://www.ed.gov/nclb/methods/whatworks/eb/edlite-slide003.html

24.

Walton

W. T.

(1985). Educators’ responses to methods of collecting, storing, and analyzing behavioral data. Journal of Special Education Technology, 7, 50–55.

25.

White

O. R.

Haring

N. G.

(1980). Exceptional teaching. Columbus, OH: Merrill.

26.

White

O. R.

Liberty

K. A.

(1976). Behavioral assessment and precise educational measurement. In Haring

N. G.

Schiefelbusch

R. L.

(Eds.), Teaching special children (pp. 31–71). New York, NY: McGraw-Hill.

27.

Wolery

Bailey

D. B.

Sugai

G. M.

(1988). Effective teaching: Principles and procedures of applied behavior analysis with exceptional students. Boston, MA: Allyn & Bacon.

28.

Ysseldyke

J. E.

Nania

P. A.

Thurlow

M. L.

(1985). Instructional decision-making practices of teachers of preschool handicapped children: Research Report No. 3. Retrieved from ERIC database. (ED269952)