Post on 21-Apr-2023
University of Groningen
SEN and the art of teachingvan der Kamp, Antoinette Jacqueline
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.
Document VersionPublisher's PDF, also known as Version of record
Publication date:2018
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):van der Kamp, A. J. (2018). SEN and the art of teaching: The effect of systematic academic instruction onthe academic and behavioural problems of students with EBD in special education. RijksuniversiteitGroningen.
CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-amendment.
Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.
Download date: 27-07-2022
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 17PDF page: 17PDF page: 17PDF page: 17
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 18PDF page: 18PDF page: 18PDF page: 18
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 19PDF page: 19PDF page: 19PDF page: 19
19
2.1 Introduction
In many countries, the aim of current primary education policy is to increase performance
in literacy and mathematics. The No Child Left Behind Act (NCLB) of 2001 (USA) is a good
illustration of this aim. In order to hold State Education Agencies accountable for students’
measurable academic progress, NCLB drew up performance standards based on adequate
yearly progress. In the Netherlands, because of alarming PISA results, we see something
similar in primary education, i.e. ‘Schooling for Tomorrow’ (Ministry of Education, Culture
and Science, 2010), a reference guide emphasising what literacy and numeracy skills all
students should have at various points in their school career. Consequently, schools are
encouraged to focus more on academic outcomes for both typically developing pupils as
well as those with severe behavioural problems. The underlying assumption is that,
despite their problems, most of the latter group of pupils do have the cognitive ability to
learn. However, obtaining higher scores for academic learning seems to be a serious
challenge for students with emotional or behavioural difficulties (EBD), with research
showing that they score considerably below typically developing students (Siperstein,
Wiley, & Forness, 2011; Ledoux, Roeleveld, van Langen, & Smeets, 2012) and even below
those with other disabilities (Zigmond, 2006).
The relationship between behavioural problems and academic learning is complex.
Gest and Gest (2005) propose two possible functional links between behavioural and
academic problems. One possibility is that behavioural problems undermine academic
skill development because of reduced learning opportunities, for instance as a result of
limited time-on-task, lost instruction time because teachers pay more attention to
controlling behaviour (Pianta & Hamre, 2009), or even disciplinary removal from the
learning environment. Conversely, it is possible that academic problems cause or
exacerbate behavioural problems because students feel overwhelmed and consequently
‘flee’ their assigned learning tasks. This can be caused by absence of prerequisite skills,
weaknesses in executive skills (Dawson & Guare, 2004) or because the curriculum does
not suit their needs. Since learning takes place in the zone of proximal development
(Vygotsky, 1978), students are by definition challenged to do something beyond their
knowledge or gained skills and this can be threatening for them. Thus, the function of
disruptive behaviour could be the avoidance of work (Penno, Frank, & Wacker, 2000;
Gunter, Coutinho, & Cade, 2002; Moore, Anderson, & Kumar, 2005). It is even plausible
that students who have experienced much academic failure in their school careers develop
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 20PDF page: 20PDF page: 20PDF page: 20
20
problem behaviour that ‘helps’ them avoid academic settings (Scott, Nelson, & Liaupsin,
2001). Briefly, it seems to be the ‘chicken or the egg’ dilemma, and various authors refer to
a cycle of problem behaviour and academic failure (Beck, Burns, & Lau, 2009), which can
even start in kindergarten (Park & Scott, 2009).
Regardless of the origin of the academic and behavioural problems of these
students, teachers have to regulate problem behaviour and improve academic results. This
dual role of teacher vs. disciplinarian may not only cause feelings of incompetence among
teachers (Lane, Wehby, & Barton-Arwood, 2005), it can even lead to burnout (Friedman,
2003). Over the last 20 years, teachers, especially those new to the field, have reported
behavioural problems as one of their greatest challenges (Darling-Hammond, 2003). Too
often, a school’s response to behaviour that undermines short-term goals like class
teaching is quick and intrusive, like giving timeout, suspension or even exclusion from
school (Banks & Zionts, 2009). So, many teachers feel a lack of expertise to handle
students with behavioural problems and report having a limited repertoire of approaches,
interventions and management strategies in dealing with them (Jones & Chronis-Tuscano,
2008).
For decades, the approach towards learning and behavioural problems of students
with EBD has primarily focused on behavioural or emotional interventions (Levy &
Vaughn, 2002). Teachers often considered good behaviour and well-being more
important, or even a condition for performance and this was underlined in their teaching
approach (van der Wolf & van Beukering, 2009). Wehby et al. (2003) refer to an almost
exclusive focus on behavioural problems. So, adapting the curriculum has never been the
first approach teachers use in order to handle intrusive social and behavioural problems
and was scarcely mentioned as a tactic for such problems (Gunter, Denny, & Venn, 2000;
Anderson, Kutash, & Duchnowski, 2001; Wehby, et al., 2003; van der Wolf & van
Beukering, 2009). Over the past decade, however, partly as a result of the aforementioned
education policy, researchers and teachers have become increasingly aware of the
importance of academic intervention for students with EBD (Wehby et al., 2003; Lane et
al., 2005). Diverse reviews (i.e. Ruhl & Berlinghoff, 1992; Coleman & Vaughn, 2000;
Mooney, Epstein, Reid, & Nelson, 2003; Ryan, Reid, & Epstein, 2004; Vannest, Temple-
Harvey, & Mason, 2009) show that an increasing number of studies on teaching academic
skills to students with EBD have emerged in the last decade. Together these studies
provide a growing and comprehensive overview of research on this subject.
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 21PDF page: 21PDF page: 21PDF page: 21
21
Given the growing focus on academic learning of students with EBD and the
supposed relationship between behavioural problems and academic functioning, it is
remarkable that only one review study explicitly addresses the influence of an academic
approach on the behaviour of these students (Sutherland & Wehby, 2001). The latter
conclude that if students have increased opportunities to actively respond to academic
requests, their academic outcomes and task engagement increased along with a decrease
in inappropriate and disruptive behaviour. Since the effects of remediation of academic
deficits on socio behavioural outcomes seem to be promising, further studies are
necessary. (Lane, O’Shaughnessy, Lambros, Gresham, & Beebe-Frankenberger, 2001).
Kauffman (2010) puts it even stronger: “We need to approach behavioural disorders as
instructional problems” (p. 180). Scheerens also undescribed that the idea that teachers
should focus on behaviour prior to dealing with learning development could be replaced
by the idea of learning instruction as the basis of the prevention, improvement and
treatment of behavioural problems (2007).
Conversely, other authors, based on research, claim that students with EBD show
lower levels of on-task behaviour despite curricular or materials modifications intended to
improve this (Mastropieri et al., 2009). Some interventions, grounded in attachment
theory, offer students a combined approach of affection, structure and teaching with a
positive effect on emotional and behavioural functioning and academic attainment
(Reynolds, MacKay, & Kearney, 2009). However, the mere effect of teaching academic
skills on behaviour remains vague and is definitely worth studying.
This chapter, therefore, seeks to address the following two questions. Firstly, what
research has been done concerning the effect of teaching academic skills on the behaviour
of students with EBD. Secondly, is it possible to provide a valid answer on whether
academic learning positively effects the behaviour of students with EBD. Although we are
aware of the limitations of untangling the instruction into a social-emotional and academic
focus, we find it important to study the simple effect of academic interventions on
behavioural problems. Precisely because they are so closely linked, it is important to know
whether a single adjustment in curriculum or instruction can have an effect on behaviour.
This being the case, teachers would be more attentive to the effect that instruction or
curriculum has on the behavioural problems of their students. To put it another way, in
order to know the effect of an academic approach on behaviour, it is necessary to isolate it
as key ingredient of educating students with EBD.
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 22PDF page: 22PDF page: 22PDF page: 22
22
Consequently, in this chapter, we focus on research having teaching academic skills
to students with EBD as the independent variable and the effect on behaviour as
dependent variable. The independent variable entails academic instruction: teaching,
presenting and modelling the essential skills and knowledge assessed by state or national
standards, supplemented by teaching support to students during instructional time with
minimal or no direct instruction to students, i.e. guided and independent practice (Vannest
et al., 2011). Interventions conducted as part of general schoolwide programmes (e.g.
‘PBS’, ‘nurture groups’ or ‘Success for all’) or carried out in psychiatric facilities were
excluded as were interventions that not exclusively focused on teaching academic skills.
The dual intervention of these approaches on behaviour as well as academic learning
makes it hard to distinguish the effect of the latter from the impact of behavioural
interventions.
Although we focus on students with EBD, it is important to note that definitions of
EBD, are contested, contain several disorders and conceptualisations of EBD, and differ
between continents and countries. Additionally, disruptive behaviour can be seen as a
characteristic of the individual (e.g. associated with genetic or biological conditions), but it
can also be context based (a response to particular situations), must be understood in the
context of culture (Evans, Harden, & Thomas, 2004; Chakraborti-Ghosh, Mofield, &
Orellana, 2010) and often coexist with other disabilities. Generally, students with EBD
cannot make full use of the educational opportunities offered to them and often need a
special form of education. They may become apparent through withdrawn, passive,
aggressive or self-injurious tendencies, often having difficulties in forming and
maintaining positive relationships.
The dependent variables comprise all forms of behaviour perceived as problematic
by teachers, or inappropriate behaviour that adversely influences students’ academic
progress and ability to achieve (Wehby et al., 2003), including externalising and
internalising behavioural problems.
2.2 Method
Studies for inclusion in this review were found via several means. First, an electronic
database (EBSCO-complete) was searched for articles published between 2000 and 2012
by using the following keywords: behaviour (social emotional behavioural
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 23PDF page: 23PDF page: 23PDF page: 23
23
disorder/difficulty/disturbance OR externalising behaviour OR disrupting behaviour OR
behavioural problems) AND education (teach* OR instruct* OR intervention OR education
OR classroom OR pedagogy) AND academics (academic outcome OR reading OR math OR
achievement). The search included all possible combinations of the three topics. An initial
scan of the results, based on title and abstract, looked at whether the study involved an
academic intervention given to students with EBD and whether its effect on their
behaviour was given. This search resulted in identifying 442 possible articles. Abstracts
often gave a vague description, so, in the next stage, we looked at the method and result
paragraphs of these found studies to identify those with the intended (in)dependent
variables. Based on this more detailed analysis we were able to remove 310 articles from
the selection. Mostly because the academic interventions were more or less explicitly
combined with a behavioural intervention. The nature of the behavioural problems often
remained unclear, as did the way in which the effect of the intervention on behaviour was
measured. The remaining 132 studies were read more thoroughly to see if they met all of
the following criteria:
(1) Subjects in the studies (aged 5–18) had emotional and/or behavioural disorders,
expressed in terms of problem behaviour or off-task behaviour. Where formal diagnoses
were absent, only participants receiving special education services for emotional or
behavioural problems were included.
(2) The subjects received academic interventions conducted in educational settings by
teachers, support staff or research associates.
(3) The studies measured the effect of the academic intervention on problem behaviour,
reporting quantitative outcome measures.
(4) The studies were published in a peer reviewed international journal.
This search resulted in 22 articles. To avoid omission errors, a manual search was
carried out of the issues of four journals that publish a great number of articles covering
this particular field: the Journal of Emotional and Behavioural Disorders, Behavioural
Disorders, Journal of Research in Special Educational Needs and European Journal of
Special Needs Education (for the 2000–2012 period). Examining tables of content,
abstracts and references pages resulted in including eight extra articles. Consequently, the
review yielded 30 empirical studies on academic interventions for students with EBD,
measuring also their effect on behaviour.
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 24PDF page: 24PDF page: 24PDF page: 24
24
The studies found fell into two types: those using a single subject design and those
using a group design (see Table 2.1a/b). The academic and instructional focus of the
intervention, the sample size, and the academic and behavioural outcomes were noted for
each study. In case that graphical representations were present, percentage Non-
overlapping Data (PND) between baseline and successive intervention phases were used
as effect size, and were calculated by identifying the highest data point in baseline and
determining the percentage of data points exceeding this level during the intervention
phase. PND was chosen because it can be easily conducted and is applicable to any form of
single-subject design (Kavale, Mathur., Forness, Quinn, & Rutherford Jr, 2000). A PND
between 91 and 100 is considered a highly effective intervention; between 71 and 90 is
considered a moderately effective intervention; between 51 and 70 is considered mildly
effective intervention; and between 0 and 50 is considered non-effective intervention
(Scruggs & Mastropieri, 1998). When a study did not present a graphical representation of
the results, we used the ES as shown by the authors of the study. An ES above 0.80 is
considered to be highly effective, between 0.40 and 0.80 to be moderately and between
0.20 and 0.40 to be small (Cohen, 1988).
A summary of the authors, academic field, instructional focus, intervention
strategies used, target behaviour, number of participants (N), diagnosis and effect on
academic and behavioural outcomes was drawn up and is presented by design (Single
Subject design vs. Group design) in Table 2.1a/b.
2.3 Results
2.3.1 Research concerning the effect of teaching academic skills on the
behaviour of students with EBD Teaching academics, as independent variable, and a key selection criterion for this review,
mostly involved reading (37%), maths (33%) and, to a lesser extent, language, writing,
science (23%). The interventions focused on instruction (43%), task (17%), seatwork
(7%) or a combination (33%) using diverse existing methods, some of which were
evidence based (Peer-Assisted Learning Strategies-PALS, Teach Your Child-TYC, Great
Leaps-GL). In the majority of studies, research assistants, (reading) specialists or school
assistants supported the interventions. Concerning the participants, most students were
diagnosed with emotional disturbance or EBD grounded in state and federal guidelines.
Other students showed high levels of inappropriate classroom behaviour or were at risk
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 25PDF page: 25PDF page: 25PDF page: 25
25
for EBD. A number of students were diagnosed with ADHD, CD, ODD or autism. The
descriptions of problem behaviour were various, ranging from shirt biting and chair
tipping to verbal threats to another person and physical attacks. Origin and context of the
problem behaviour often remained unclear. Across the single-subject studies (87% of the
total studies found) there were 120 pupils, with between 1 and 15 (mean 4.6, SD 3.0)
participating per study. Seventy percent of them attended self-contained schools or
classrooms especially for students with EBD. The number of participants across the group
design studies (13%) was 167 (mean 42, SD 35), all attended general education.
The behavioural outcomes, as dependent variable, and the second key selection
criterion of this review, involved on-task behaviour (on/off task behaviour and
engagement, 45%), problem behaviour (disruptive, destructive, (in)appropriate behaviour
and compliance, 31%) or a combination of both (24%). Most studies used observation to
measure behaviour; one study used the TRF (Teacher Report Form, Achenbach, 1991) as
outcome and one study used the number of office referrals. Fifty-seven percent of the
reviewed studies presented academic outcomes as well as behavioural outcomes as
dependent variable. The academic outcomes were mainly reported in terms of correct
answers (per minute), measured with standard tests. Data from the single-subject studies
were generally presented by mean score comparisons across phases, whether or not
combined with SD, and often supplemented with a graphical representation. Only in four
studies effect scores were calculated. Concerning the variety of studies, it was hard to
estimate the relationship between academic outcomes, on-task behaviour and disruptive
behaviour. At first glance, the effect on on-task behaviour seemed to be larger than the
effect on disruptive behaviour. Moreover, in some studies students were removed from
the research group because of their disruptive behaviour.
Regarding the quality indicators for single subject design (Horner et al., 2005,
174), we saw numerous flaws in the description of participants and settings. For instance,
only six studies described the process for selecting participants with replicable precision.
Also, in numerous studies, the scores of the dependent variable fluctuated widely during
baseline phase, making it difficult to judge the outcomes of the intervention phase.
Numerous studies revealed weak or no functional analyses to examine the relationship
between the research subject and problem behaviour of the students. The relation
between educational approaches and behavioural problems often remained unclear.
Design controls for common threats to internal validity (e.g. of rival hypotheses) were
limited. In all but one study the (research) assistants played a major role in implementing
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 26PDF page: 26PDF page: 26PDF page: 26
26
the independent variable, which hampers the ecological validity of the studies. Because of
the limited replications of the diverse studies, external validity of the studies was also
hard to establish.
2.3.2 The effects of academic learning on the behaviour of students
with EBD The questionable quality of the studies, combined with the fact that these contained a wide
variety of (in)dependent variables measured with mainly small research groups with a
diversity of types and needs of EBD and diverse expressions of problem behaviour (the
‘apples and oranges’ problem as stated by Kavale et al., 2000), made it hard to interpret
the results and to draw overall conclusions. Nevertheless, we think it is worthwhile to
describe these studies as a first tentative step to answering our second research question.
Studies with a similar instructional focus were combined for the benefit of the discussion
of results. Successively we discuss opportunity to respond (OTR), peer tutoring, various
types of instructional methods, task modification and seatwork. The latter two are
discussed explicitly, because they take up a large part of the instructional process and
seem to be a source of behavioural problems for students with EBD (Moore et al., 2005;
Umbreit, Ferro, Liaupsin, & Lane, 2007; Hayling, Cook, Gresham, State, & Kern, 2008).
Sutherland, Alder, and Gunter (2003), Haydon et al. (2010), Tincani and Crozier
(2008) and Benner, Ralston and Feuerborn (2012) studied approaches to increase OTR.
The outcomes demonstrated that increasing the quantity of interactions between teacher
and student resulted in (a) more responses that were correct and (b) increased on-task-
behaviour. The studies also revealed, though less obviously, lower levels of disruptive
behaviour. The degree to which instruction provided students with regular and
predictable natural reinforcement (i.e. success) seemed to be an important factor in
keeping students on task.
Four articles were found on peer tutoring as a form of instruction. They all used
and researched PALS, a peer-mediated intervention with a strong evidence base
(Sutherland & Snyder, 2007). Wehby et al. (2003) examined the effects of PALS, combined
with a modified version of Open Court Reading, while in a similar study Barton-Arwood,
Wehby, and Falk (2005) assessed the effects of PALS combined with Horizons Fast Track
A-B Reading Program. Both studies revealed moderate gains in reading achievement and a
slight increase in attendance. An impact on problem behaviour, however, was not
observed (Wehby et al., 2003), or observed but questionable (Barton-Arwood, Wehby, and
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 27PDF page: 27PDF page: 27PDF page: 27
27
Falk, 2005). Sutherland and Snyder (2007) also conducted a study using PALS, and
carefully compiled the dyads in order to maximise the effect of peer tutoring. They also
combined this approach with self-graphing, a component of self-evaluation in which
students graph their own data in a visual representation of their performance over time.
Results indicated that during the intervention phase, most students made progress on
words read correctly per minute. Furthermore, the disruptive behaviour (where present)
decreased and the active responding of all students increased. The researchers found an
apparent link between active responding and disruptive behaviour. This effect however
decreased during follow-up. The above studies were conducted in classrooms by teachers
assisted by researchers, research assistants (RAs) or school-site personnel and therefore
did not correspond with everyday reality. So, in their study, Lane, Little, Redding-Rhodes,
Phillips, and Welsh (2007) focused on whether teachers were able to use PALS in a general
education setting without support from either researchers or on-site support staff. Results
revealed lasting improvements in reading fluency for all students, but less improvement
on academic engaged time. In a descriptive study, Conroy, Asmus, Boyd, Ladwig, and
Sellers (2007) compared academic peer-to-peer interactions to adult directed activities.
They concluded that peer-to-peer interactions even produced higher rates of disruption in
comparison to adult-directed activities. Sutherland and Snyder (2007) considered the
social skills deficits of students with EBD as a threat to the efficacy of peer-to-peer
intervention because some students may not interact well with others. Consequently,
PALS used in academic interventions does not seem to have an immediate positive effect
on problem behaviour. It appears to be important to arrange groups carefully.
Six studies on specific forms of instruction were conducted by McComas, Hoch,
Paone, and El-Roy (2000), Scot and Shearer-Lingo (2002), Tyler-Wood, Perez Cereijo, and
Pemberton (2004), Lee, Sugai, and Horner (1999), Lane et al. (2001) and Lane et al.
(2002). These studies involved various types of instructional methods namely the so-
called three-step approach, effective reading instruction and component skills instruction.
Effects on problem and on-task behaviour were moderate to large. Successful outcomes
appeared to be related to making academic tasks less complicated by giving instruction at
the right level for the students, maintaining direct teacher–student interaction, and
providing opportunities to practice and respond. The degree to which instruction
provided students with regular and predictable natural reinforcement (i.e. success) was
again asserted as possibly the most important intervention. Lee et al. (1999) found an
apparent link between reduced problem behaviour and strategies based on accurate
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 28PDF page: 28PDF page: 28PDF page: 28
28
knowledge of the maintenance of problem behaviour. Lingo, Jolivette, and Barton-Arwood
(2009) used visual cards to introduce a clear goal followed by visual feedback, which also
improved students’ appropriate behaviour. Lane et al. (2002) concluded that the increased
efficacy in early reading skills enables students to participate in reading activities during
literacy. According to them, this confirmed previous studies showing that improved early
literacy skills are associated with lasting decreases in disruptive classroom behaviour. One
study (Rafferty, 2012) added a tactile prompting device for self-monitoring during
instruction. All these studies also linked teacher–student interactions and positive
feedback to decreased problem behaviour.
Seven studies concentrated on task modification as an intervention when teaching
academic subjects to students with EBD. The various approaches were about difficulty of
task (Lee et al., 1999; Beck et al., 2009), size or function (Skinner et al., 2002; Miller et al.,
2003; Moore et al., 2005), freedom of choice (McComas et al., 2000; Kern, Bambara, & Fogt
,2002) and pre-teaching, visual adaptations combined with physical movement between
tasks (Pang & Zhang, 2011). The studies supported the hypothesis that off-task or problem
behaviour is maintained by escaping from difficult or unattractive tasks. The various
outcomes showed that providing students with an appropriate task (e.g. through choice
making, altering task duration or making tasks understandable by pre-teaching or
repeating) probably prevented problem behaviour and increased on-task behaviour.
Functional behavioural assessment seemed to be an effective approach to define such
tasks for students. In particular, the understanding of the task appears to be important in
keeping students with EBD occupied and diminishing problem behaviour. Shortening the
assignment or adding, a function to a task had a minimal effect.
Concerning seatwork, we found a variety of study design types. Four single-subject
studies concerned interventions that provided students with a problem-solving worksheet
(Alter, Brown, & Pyle, 2011), choice of task (Jolivette, Wehby, Canale, & Massey, 2001),
several breaks during seatwork (Moore et al., 2005), or self-monitoring (Levendoski &
Cartledge, 2000). These studies showed a small to large effect on on-task behaviour.
External controlling, i.e. through teachers giving breaks or through stimulating, self-
monitoring beeps, seemed to be most effective. On-task behaviour, however, failed when
the intervention (the beeps) was gradually removed. Four other studies compared
seatwork to different forms of instruction (Conroy et al., 2007; Baker, Clark, Maier, &
Viger, 2008; Hart, Massetti, Fabiano, Pariseau, & Pelham, 2011). Baker et al. (2008)
concluded that during what they called the most demanding instructional context,
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 29PDF page: 29PDF page: 29PDF page: 29
29
individual seatwork, students with behavioural problems were only academically engaged
half the time. They found that the context of seatwork was considerably less structured
and that the structure provided was mainly in the form of reactive discipline, which
elicited problem behaviour instead of decreasing it. The conclusion of all these studies is
that whole-group instruction and independent seatwork correlated most strongly with
problem behaviour: considerable more disruptive behaviour arises in these settings than
during small-group or one-to-one instruction.
2.4 Discussion
Since traditional behavioural approaches towards students with EBD have not always had
the desired effect on academic outcomes, we see a tendency to teach academics as a mean
to improve education results while decreasing the problem behaviour of students with
EBD. This approach fits seamlessly into the recent policy of focusing on academic
outcomes (e.g. NBCL in the USA and Schooling for Tomorrow in the Netherlands). The
mere effect of teaching academic skills on behaviour remains vague and therefore a reason
for conducting a review.
Given the number of reviews and studies on teaching academic subjects to
students with EBD, knowledge on the topic is expanding. Yet, many teachers feel
incompetent to handle the current dual ambition of teaching, i.e. improve both
behavioural and academic outcomes. Consequently, it is interesting to focus on
approaches with a positive effect on both. From this perspective, it is disappointing that
many studies on specific approaches of teaching academics to students with EBD limit
themselves to only measuring academic outcomes, while the effect on behaviour is
definitely worth studying. Thus, this review deals with studies on teaching academic skills,
focusing on both academic and behavioural outcomes.
In our research of studies on the effect of teaching academics on the behaviour of
students with EBD, we found 30 articles that met our criteria. The interventions and
circumstances in which they took place varied, however, considerably. Except for studies
with the same (co)-author, the instructional methods were different, were used in
different stages of the instructional process and in different settings. Although
externalising behaviour as off-task behaviour and noncompliance dominated, the context
of the problem behaviour or type of EBD often remained unclear. Moreover, most studies
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 30PDF page: 30PDF page: 30PDF page: 30
30
were single-subject ones that did not meet all the quality indicators for single-subject
design (Horner et al., 2005). Given this ‘apples and oranges’ situation (Kavale et al., 2000)
and poor design, the studies found do not provide a solid basis for firm conclusions. Thus,
we can state that research on the effect of academic learning on behaviour has not been
undertaken systematically and that it is too early for well-founded statements on the
subject. Conversely, Kavale et al. (2000) state that it is precisely diverse studies that need
to be integrated to achieve convincing generalisation and practical simplicity. So, in order
to study the effects of academic learning on the behaviour of students with EBD, we
combined different studies using a couple of denominators to make it possible to draw
some overall cautious conclusions about the effect of opportunity to respond (OTR), peer
tutoring, various types of instructional methods, task modification and seatwork on the
behaviour of students with EBD. Focusing on these subjects can enhance our
understanding of the effect of teaching academics on behaviour.
The studies concerning OTR showed that reciprocal interaction with an
instructional and academic purpose increased on-task behaviour and decreased disruptive
behaviour. These results correspond to the results of a review on this subject conducted
by Sutherland and Wehby (2001). The latter, however, also found that providing students
with the necessary OTR seems to be challenging for teachers and that the rate of OTR in
classrooms for students with EBD appeared to be alarmingly low. Peer tutoring or
cooperative learning (PALS) could be a possible alternative to teacher instruction, but the
results of this review proved this approach to be less effective: academic interaction
between peers appeared to have a smaller positive effect on behaviour than teacher–
student interaction. However, apart from one study (Sutherland & Snyder, 2007), the
construction of the peer groups in these studies remained unclear. Since the nature and
quality of the interaction is particularly important for students with EBD, it is vital to take
this into account when interpreting these outcomes.
Concerning instruction, we noted that, although instructional methods can be
diverse, provided they are at the appropriate level, they appeared to be an important
condition for prompting on-task behaviour. The same applied to the academic tasks
offered: when students with EBD understood what was expected of them and
comprehended the material, less problem behaviour was observed. This requires a
carefully constructed curriculum in a supportive setting. The latter also applies to
seatwork. It appeared that students with EBD, during this stage of the instruction process,
still needed support to stay on-task.
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 31PDF page: 31PDF page: 31PDF page: 31
Tabl
e 2.
1a. S
yste
mat
ic su
mm
ary
of si
ngle
subj
ect d
esig
n st
udie
s.
Auth
ors
Acad
emic
fie
ld
Focu
s In
terv
enti
on
Dep
ende
nt v
aria
ble
N D
iagn
osis
O
utco
mes
on
acad
emic
s O
utco
mes
on
beha
viou
r Su
ther
land
, Ald
er
and
Gunt
er (2
003)
M
aths
In
str.
OTR
– te
ache
r pra
ise
Corr
ect r
espo
nses
4
ED
PND
= 7
0%
Dis
rupt
ive
beha
viou
r PN
D =
5%
On
- tas
k be
havi
our
PND
= 8
2%
Hay
don
et a
l. (2
010)
La
ngua
ge
arts
In
str.
OTR
- Ind
ivid
ual v
s. ch
oral
resp
ondi
ng
Dis
rupt
ive/
Off-
task
be
havi
our/
Activ
e re
spon
ding
6 EB
D (a
.r)
(SSB
D)
PND
=
70/5
0/62
%
OTR
- Cho
ral v
s. m
ixed
resp
ondi
ng
PND
=
67/2
3/18
OT
R - I
ndiv
idua
l vs.
mix
ed re
spon
ding
PN
D =
83
/64/
38%
Ti
ncan
i and
Cr
ozie
r (20
08)
Oral
la
ngua
ge
Inst
r. Br
ief v
s. ex
tend
ed w
aitin
g tim
e Co
rrec
t aca
dem
ic re
spon
ses
2 N
o/au
tism
PN
D =
11%
D
isru
ptiv
e be
havi
our
PND
= 4
0%
Benn
er, R
alst
on,
and
Feue
rbor
n (2
012)
Proc
essi
ng
spee
d In
str.
Proc
essi
ng sp
eed
10
EBD
ES
= 0
.57
Beha
viou
r (TR
F)
ES =
0.8
7
Weh
by, F
alk,
et a
l. (2
003)
Re
adin
g In
str.
Corr
ect w
ords
per
min
ute
8 ED
PN
D =
40%
At
tent
ion
PND
= 0
%
Inap
prop
riat
e be
havi
our
PND
= 5
%
Bart
on-A
rwoo
d,
Weh
by, a
nd F
alk
(200
5)
Read
ing
Inst
r. En
gage
men
t 6
n.a.
PND
= 5
9 %
In
appr
opri
ate
beha
viou
r (n
egat
ive
talk
, agg
ress
ion)
PN
D =
5%
Suth
erla
nd a
nd
Snyd
er (2
007)
Re
adin
g In
str.
Wor
ds re
ad c
orre
ctly
per
m
inut
e
4 ED
PN
D =
0%
Dis
rupt
ive
beha
viou
r PN
D =
0%
A
ctiv
e re
spon
ding
PN
D =
32,
5%
Lane
et a
l. (2
007)
Re
adin
g In
str.
Corr
ect l
ette
r sou
nds p
er
min
ute
7 BD
(a.r)
PN
D =
51%
Acad
emic
eng
agem
ent
PND
= 3
%
McC
omas
et a
l. (2
000)
D
iver
se
Task
D
isru
ptiv
e be
havi
our/
com
plia
nce
1 Au
tism
75
/25%
1 10
0/30
%
Inst
r.
Lang
uage
for t
hink
ing
Open
cour
t rea
ding
PA
LS
Hor
izon
s fas
t tra
ck
PALS
PALS
Self-
grap
hing
PALS
Choi
ce o
f tas
k se
quen
ce v
s. no
cho
ice
Repe
ated
task
s vs
. no
repe
ated
task
s
No
Inst
ruct
ion
vs.
inst
ruct
ion
on
1 10
0/10
0%
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 32PDF page: 32PDF page: 32PDF page: 32
Scot
t and
She
arer
-Li
ngo
(200
2)
Read
ing
fluen
cy
Inst
r. Te
ach
your
child
/gre
at le
aps
Wor
ds re
ad p
er m
inut
e 3
n.a.
PND
=
44%
/PN
D =
98
%
On-t
ask
beha
viou
r PN
D =
60
%/P
ND
=
100%
Ty
ler-
Woo
d et
al.
(200
4)
Lang
uage
an
d m
aths
Ta
sk
Curr
icul
um b
ased
mea
sure
men
t N
umbe
r of o
ffice
refe
rral
s 15
BD
ES
= 0
.77
Lee,
Sug
ai, a
nd
Hor
ner (
1999
) M
aths
Ta
sk
Diff
icul
t vs.
easy
task
s Co
rrec
t ans
wer
s (N
= 2
) 2
EBD
PN
D =
100
%
Occu
rren
ce o
f pro
blem
be
havi
our (
N =
1)
ADH
D
PND
= 6
6%
Off-t
ask
beha
viou
r (N
= 2
) PN
D =
65%
In
str.
Corr
ect a
nsw
ers
PND
= 1
00%
Oc
curr
ence
of p
robl
em
beha
viou
r PN
D =
72%
Off-t
ask
beha
viou
r PN
D =
56%
La
ne e
t al.
(200
1)
Read
ing
Inst
r. Ph
onol
ogic
al a
war
enes
s tra
inin
g fo
r re
adin
g (P
ATR)
Co
rrec
t wor
ds p
er m
inut
e 7
CP+H
IA
(a.r.
) PN
D =
66%
To
tal d
isru
ptiv
e be
havi
ours
PN
D =
6%
La
ne e
t al.
(200
2)
Read
ing
Inst
r. Jo
hn S
hefe
lbin
e’s
Phon
ics C
hapt
er
Book
s Co
rrec
t wor
ds p
er m
inut
e 5
n.a.
ES =
0.8
6 To
tal d
isru
ptiv
e be
havi
ours
(T
DB)
in th
e cl
assr
oom
ES
= −
1.04
Neg
ativ
e so
cial
inte
ract
ions
–
play
grou
nd
ES =
−0.
63
Ling
o, Jo
livet
te,
and
Bart
on-
Arw
ood
(200
9)
Read
ing
Inst
r. Ve
rbal
feed
back
Ap
prop
riat
e be
havi
our
1 n.
a.PN
D =
0%
Ve
rbal
& v
isua
l fee
dbac
k PN
D =
80%
Raffe
rty
(201
2)
Read
ing
Inst
r. Se
lf-m
onito
ring
On
-tas
k be
havi
our
4 EB
D
PND
= 9
6%
Beck
, Bur
ns, a
nd
Lau
(200
9)
Read
ing
Task
/ins
tr.
Pre-
teac
hing
unk
now
n re
adin
g ite
ms
On-t
ask
beha
viou
r 2
BD
PND
= 9
4%
Mill
er e
t al.
(200
3)
Wri
ting
Task
/ins
tr.
Prov
idin
g a
func
tion
for w
ritt
en
assi
gnm
ent/
mod
ellin
g Co
rrec
t wri
ting
resp
onse
s 3
EBD
PN
D =
0/5
3%
Stud
y 1
On-t
ask
beha
viou
r PN
D=9
/0%
Id
em
Mat
hs
Task
Sh
orte
ned
mat
hs a
ssig
nmen
ts
Corr
ect r
espo
nses
per
m
inut
e PN
D =
4%
PN
D =
29%
On-t
ask
beha
viou
r St
udy
2 M
oore
, And
erso
n,
Mat
hs
Task
/sea
twor
k Re
duct
ion
in ta
sk d
urat
ion
duri
ng
Off-t
ask
beha
viou
r 1
Na.
PN
D =
100
%
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 33PDF page: 33PDF page: 33PDF page: 33
Mat
hs
Seat
wor
k On
-tas
k be
havi
our
4 ED
PN
D =
47%
Scie
nce
Task
En
gage
men
t 6
SED
+ PN
D =
90%
D
estr
uctiv
e be
havi
our
PND
= 0
%
Read
ing
Task
/ins
tr.
Inte
rspe
rsin
g ad
ditio
nal b
rief
pr
oble
ms
Choi
ces a
nd h
igh-
inte
rest
act
iviti
es
Visu
al c
lues
and
sepa
rate
inst
ruct
ions
in
ters
pers
ed w
ith p
hysi
cal
mov
emen
ts +
Pre
-tea
chin
g
Corr
ect a
nsw
ers
3 AD
HD
(a.r.
) PN
D =
67%
Off-t
ask
beha
viou
r PN
D =
95%
M
aths
In
str./
seat
wor
k W
ord
prob
lem
-bas
ed le
arni
ng
Prob
lem
s sol
ved
corr
ectly
3
EBD
PN
D =
81%
PN
D =
36%
Ti
me
on ta
sk
Mat
hs
Inst
r./se
atw
ork
Mul
tiste
p pr
oble
m-s
olvi
ng st
rate
gy +
re
info
rcem
ent
Prob
lem
s sol
ved
corr
ectly
3
EBD
PN
D =
85%
PN
D =
24%
Ti
me
on ta
sk
Mat
hs
Seat
wor
k M
ultip
le c
hoic
e-m
akin
g op
port
uniti
es
duri
ng se
atw
ork
Atte
mpt
ed ta
sk p
robl
ems
3 EB
D
PND
= 3
9%
Prob
lem
s cor
rect
PN
D =
54%
Ta
sk E
ngag
emen
t PN
D =
57%
Of
f-tas
k be
havi
our
PND
= 6
2%
Dis
rupt
ion
PND
= 0
%
Skin
ner e
t al.
(200
2)
Kern
, Bam
bara
, an
d Fo
gt (2
002)
Pa
ng a
nd Z
hang
(2
011)
Alte
r, Br
own,
and
Py
le (2
011)
Al
ter (
2012
)
Joliv
ette
et a
l. (2
001)
Leve
ndos
ki a
nd
Cart
ledg
e (2
000)
M
aths
Se
atw
ork
Self-
mon
itori
ng
Mat
hs p
robl
ems c
ompl
eted
co
rrec
tly
4 SE
D
PND
= 6
0%
Tim
e on
task
PN
D =
100
%
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 34PDF page: 34PDF page: 34PDF page: 34
Tabl
e 2.
1b.
Syst
emat
ic su
mm
ary
of g
roup
des
ign
stud
ies.
Auth
ors
Acad
emic
fie
ld
Focu
s In
terv
entio
n D
epen
dent
var
iabl
e N
D
iagn
osis
Or
der o
f effe
ctiv
enes
s
Bake
r et a
l. (2
008)
N
A In
str./
seat
wor
k D
irec
t (w
hole
cla
ss)
On-t
ask
beha
viou
r 39
n.
a.
DI <
IS <
SGI
< IT
In
stru
ctio
n Sm
all G
roup
Inst
ruct
ion
Indi
vidu
al S
eatw
ork
Inte
ract
ive
Teac
hing
H
aylin
g et
al.
(200
8)
NA
Inst
r./se
atw
ork
Who
le C
lass
Inst
ruct
ion
Dis
rupt
ive
beha
viou
r 90
EB
D+
Smal
l Gro
up In
stru
ctio
n En
gage
men
t IS
< C
L< W
CI, S
GI, O
OI
Coop
erat
ive
Lear
ning
D
estr
uctiv
e be
havi
our
IS <
OOI
< W
CI, S
GI, C
L In
depe
nden
t Sea
twor
k OO
I < IS
W <
WCL
, SGI
, CL
One-
on-O
ne In
stru
ctio
n Co
nroy
et a
l. (2
007)
M
aths
, Sc
ienc
e In
stru
ctio
nal s
ettin
g D
isru
ptiv
e be
havi
our
5 AS
D
Wri
ting,
Re
adin
g (I
ndep
ende
nt S
eatw
ork
vs. G
roup
Wor
k)
IS =
GW
Inst
ruct
iona
l act
ivity
(Adu
lt vs
. Chi
ld)
Ch <
Ad
Har
t et a
l. (2
011)
Re
adin
g In
str./
seat
wor
k Sm
all-g
roup
Inst
ruct
ion
On-t
ask
beha
viou
r /w
ork
prod
uctiv
ity
33
ADH
D
IS <
SG
Who
le-g
roup
Inst
ruct
ion
WG
< SG
In
depe
nden
t Sea
twor
k
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 35PDF page: 35PDF page: 35PDF page: 35
35
Based on the outcomes, we can conclude that systematic research on this
subject is still in its infancy. Nevertheless, the findings of this chapter suggest that
academic learning may have a positive effect on behavioural outcomes in the
classroom if teachers positively affect the educational experience of students with
EBD (Tyler-Wood et al., 2004). Given the fact that one of the major functions of
unacceptable behaviour in classrooms is escaping from difficult tasks (Gunter &
Coutinho, 1997), successful academic learning seems to reduce the incidence of
behaviour problems in the classroom. Therefore, to avoid frustration and all its
implications, it is imperative for teachers to enhance the curriculum in line with
the special educational needs of such students. This does not imply a less
demanding curriculum (Wehby et al., 2003), but a more carefully constructed one.
However, it was impossible to ascertain exactly the background, cause and
/or impact of the behavioural problems of each student in the reviewed studies.
Moreover, since comorbidity often occurs, it is impossible to indicate if the most
appropriate approaches were chosen for students with a specific disorder and/or
behavioural problem. The majority of the reviewed studies did not make entirely
clear whether the researchers sufficiently met the specific educational needs of the
students. Although this is inherent to these types of short-term case studies, this
important aspect of teaching academics to students with special educational needs
remains uncovered. In some cases, this could even be a cause of disappointing
results (McComas et al., 2000). Given the diversity of types and needs of students
with EBD, knowing and anticipating on these needs seems to be the first step to
success. Herein lies the key to teaching academics: the curriculum continuously
has to be adjusted to the needs of the student in an ongoing process. This approach
fits the recent policy of data-driven teaching. However, we need to realise that
working systematically only makes sense if teachers have the opportunity, skills
and autonomy to adapt their approach to the specific needs of students with EBD.
In reality, teachers often lack the knowledge and skills to optimise the curriculum
(Jones & Chronis-Tuscano, 2008), so further work is required to accomplish this.
Several limitations to this review need to be acknowledged. The
aforementioned variety in research, the small internal and external validity and
524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorp524629-L-bw-vdWorpProcessed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018Processed on: 3-10-2018 PDF page: 36PDF page: 36PDF page: 36PDF page: 36
36
restricted sample sizes make it hard to draw overall conclusions concerning
students with EBD. Additionally, some researchers speak of idiosyncratic
outcomes: all students labelled with EBD react in their own personal way and
cannot be treated as if they were the same. Knowing that the target student
described in one setting may be quite different from the target student in another
setting makes it hard to repeat the studies: an important step in the documentation
of an evidence-based practice. Secondly, although PND is a widely used method to
calculate effect sizes, has certain disadvantages. One or two deviating scores
during baseline can have a major influence on the calculated effect size. Given the
observed fluctuating scores during the baseline phases, some of the displayed
effects do not fully match the effect as apparently shown when eyeballing the
graphs, or as described by the researchers. Thirdly, researchers and research-
assistants played a major role in performing the intervention. This makes it hard to
translate the outcomes of these kinds of studies to daily practise, especially
considering the above limitations of teachers.
In conclusion, although we were unable to give an unambiguous answer to
our research question, the added value of this descriptive analysis is the
augmented awareness that only a single intervention with respect to the
curriculum or instruction may possibly have an effect on problem behaviour.
Confronted with problem behaviour, rather than initially focusing on behavioural
approaches, teachers may do well to reconsider the requested task and their
instructional approach. It also elucidates that more research needs to be
undertaken before the relationship between academic instruction as provided by
teachers and its impact on behaviour of their students with EBD is more clearly
understood. The next two chapters focus on analysing this impact.