src="../../images/uparrow.gif" alt="Arrow Up" align="right" border="0" height="13"
width="20" />

Phase-I

Method

Participants

The participants in this study included 78 teachers (76 women and 2 men, mean
age 44.7 years) from three different elementary schools from a large metropolitan
area. Two of the schools were considered mid-level SES schools ( N =21
& 21) and one low SES ( N =36) based upon their placement within their
county for receiving free or reduced school lunches. Sixty-eight of the teachers
were White, seven were Black, and one Asian. Two teachers did not provide their
race. The teachers were all regular education teachers with an average of 15.75 (
SD =10.18) years of teaching experience. A breakdown of the number of
teachers representing each grade level was as follows: kindergarten N =11,
first grade N =11, second grade N =12, third grade N
=18, fourth grade N =10, fifth grade N =13, and three who did not
provide their grade level.

Materials

Participants were asked to complete a survey that was comprised of three parts.
The first part was demographic information. The second part was an eight-item
inventory comprised of two factors developed by Carol Dweck (see Dweck, 1999). Four
items measured “theories of intelligence” and four items measured
“the kind of person” one is (see Appendix). The items measured the
extent to which a person adopts an entity-based theory versus an
incrementally-based theory. Higher scores on each scale are based upon entity
theories and lower scores on incremental theories. The third part of the survey
included three different fictional scenarios, developed by the authors, that
described a child with hyperactivity (see Appendix). The scenarios were
systematically varied according to the number of ADHD characteristics exemplified
by the child. According to the DSM-IV-TR (American Psychiatric Association, 2000) a
child must meet three requirements to be considered for ADHD diagnosis. They
include 1) six out of nine symptoms for hyperactivity-impulsivity have to be
present for at least 6 months at a maladaptive level 2) some maladaptive symptoms
were present before the age of seven and 3) some symptoms present in two or more
settings (ex. at school and home). According to these qualifications Scenario 1 is
the only scenario presenting enough evidence to suggest the child may have ADHD. In
Scenario 1 the child displays the following symptoms: fidgeting, difficulty waiting
turn, talks excessively, blurts out answers before questions complete, interrupts
others, and leaves seat. In addition, the child's symptoms were present before the
age of 7 and are present in more than one setting. In Scenario 2 the child only
displays the symptoms of fidgeting and squirming. It is unknown if the symptom
occurred before the age of 7, or if the symptom occurs in other settings. In
Scenario 3 the symptoms include: difficulty waiting turn, leaves seat, interrupts
others, and talks incessantly. It is unknown if these symptoms occurred before the
age of 7, or if the symptoms occur in other settings. Each scenario was followed by
eight options that included possible actions the child's teacher could take to deal
with the child's hyperactive behavior. The directions were as follows:
“Please read the following scenario and then rank order ALL of the options
that follow, in the order that you feel is the most appropriate, from 1 (most
likely) to 8 (least likely). Please do not add any additional options or change
existing options when ranking your preferences.” Four of the options were
behavioral options (e.g., set up a system for reinforcing the child's appropriate
classroom behavior) and four were clinical options (e.g., refer the child for ADHD
diagnosis). The options remained the same for each scenario.

Procedure

Survey packets were delivered to each of the schools and picked up one week
later. Each teacher at the three schools received a packet and were asked to
complete the survey independently within one week. An overall response rate of 72%
was obtained from the three schools (21/30; 21/32; 36/46).

Results

In this section descriptive statistics will be reported followed by
correlational and inferential statistics. Composite scores were used in the
analysis of the rankings from each scenario that included one score for an overall
behavioral ranking and one for an overall clinical ranking. The two dimensions of
the Dweck scale were analyzed separately. Coefficient alpha reliability indices
indicated that both the intelligence scale (alpha=.947) and kind of person scale
(alpha=.894) showed a high degree of internal consistency.

Descriptive Statistics

Descriptive statistics for Phase I of the study are shown in Table 1.

The means reported for each scenario include a composite score from both the
behavioral intervention options and the clinical intervention options. Lower scores
indicate a higher preference for that type of intervention. The participants
overwhelmingly rated behavioral interventions higher than clinical interventions
across all three scenarios. Separate mean ranks for each of the eight intervention
options are shown in Table 2. Using a behavior monitoring plan was the highest
rated option across all three scenarios while suggesting that the child take
Ritalin was the lowest rated option across the three scenarios.

Correlational Statistics

Correlational statistics are provided in Table 3. Teachers with more experience
tended to report that basic attributes about a person can change as evidenced by
the positive correlation between experience and kind of person. An inconsistent
pattern of correlations developed between beliefs about intelligence and responses
to the scenarios. Significant correlations were found in scenario two for both the
behavioral and clinical ratings revealing that teachers choosing more behavioral
options tended to support a more malleable view of intelligence. For scenario 3 the
tendency to select behavioral interventions was correlated with the malleable view
of intelligence but the tendency to select clinical interventions did not correlate
significantly with a static view of intelligence. No significant correlations were
found between intelligence and behavior management strategies for scenario one. In
addition, no significant correlations were revealed between the kind of person
variable and strategy selection on any of the scenarios.

Inferential statistics

In order to address our first two research questions we conducted t-tests and
repeated measures ANOVA procedures to examine differences in intervention
preference for each scenario and the level of consistency in the ratings across the
scenarios. A comparison of the preference for behavioral interventions versus
clinical interventions revealed significant differences in favor of the behavioral
interventions across all three scenarios (Scenario 1: t (66)=56.62,
p <.001; Scenario 2: t (66)=46.03, p <.001;
Scenario 3: t (67)=33.31, p <.001). Repeated measures ANOVA
revealed no differentiation between scenarios in the ratings provided by the
teachers for either behavioral interventions, F (2, 132)=.514, p
=.589, or clinical interventions, F (2,132)=.996, p =.369.
Therefore, manipulation of ADHD-like symptoms had no effect upon teachers'
selection of interventions.

src="../../images/uparrow.gif" alt="Arrow Up" align="right" border="0" height="13"
width="20" />

Phase-II

Pre-service teachers served as participants in the second phase of the study
that employed somewhat different procedures. An attempt was made in this phase to
avoid influencing the participants' response by first asking for open-ended
response for each scenario before ranking the intervention options. In addition, an
attempt was also made to get at what interventions the pre-service teachers viewed
as ultimately having the most impact on the behavior rather than what sequence the
interventions should be employed, therefore the instructions were changed for the
ranking process. The following research questions were investigated in this second
phase:

• What types of interventions do pre-service teachers recommend when
presented with scenarios of students who exhibit hyperactive behavior?

• Do pre-service teachers choose different interventions when working
with children who meet the diagnostic qualifications for ADHD versus children who
do not?

• Do pre-service teachers who score high on Dweck's entity scale tend
to select clinical interventions for hyperactive students?

In the second phase we hypothesized that pre-service teachers would show a
similar preference for behavioral interventions that the elementary-school teachers
showed in the first phase. Similarly, we hypothesized that they would differentiate
between the scenarios in the interventions that they selected. With regard to our
third question, we expected to see teachers with more entity-based beliefs to
select clinical options more frequently than teachers with incrementally-based
beliefs in both their open-ended response and in their ranking preferences from the
intervention options.

Method

Participants

The participants in this study included 93 pre-service teachers (73 women and 20
men, mean age 24.4 years) from three different sections of an educational
psychology course at a medium-sized university in the South. The course was taken
during the junior or senior year after admission to the teacher education program
and at the beginning stages of their practicum experiences. Seventy-nine of the
teachers were White, 13 were Black, and one Hispanic. The students came from a mix
of specialty areas within the College of Education . Participants were offered
extra credit for their participation.

Materials

Participants were asked to complete a survey that was comprised of four parts.
The first part was demographic information. The second part was the eight-item
inventory by Dweck used in Phase I. The third part of the survey included the same
three fictional scenarios that were used in Phase I but with different requirements
for participant responses.

Procedure

The participants first completed the demographic sheet and Dweck's entity scale
before reading the scenarios. After the participants read each scenario they were
asked to provide an open-ended response to the question, “If this student
were in your classroom, what would be your initial response in dealing with their
behavior. Please list only one response.” Upon completion of their open-ended
responses for each scenario the participants were shown a Powerpoint slide asking
them to rank order the same options used in Phase I. Specifically, the instructions
asked them to, “ rank order ALL of the options listed below, in the order you
feel ultimately will be the most successful in changing the students' behavior,
from 1 (most successful) to 8 (least successful). The wording of this statement was
intended to emphasize what variables the participant felt would actually be most
likely to correct the misbehavior rather than asking what sequence of actions they
felt they should employ. Participants then rank ordered the options for each of the
three scenarios.

Results

In this section descriptive statistics will be reported followed by
correlational and inferential statistics. Composite scores were used in the
analysis of the rankings from each scenario that included one score for an overall
behavioral ranking and one for an overall clinical ranking. The two dimensions of
the Dweck scale were analyzed separately. Coefficient alpha reliability indices
indicated that the intelligence scale (alpha=.902) and kind of person scale
(alpha=.849) showed a high degree of internal consistency.

Descriptive Statistics

Descriptive statistics for Phase II of the study are shown in Table 4. Once
again, the participants overwhelmingly rated behavioral interventions higher than
clinical interventions across all three scenarios. Approximately ten percent of the
total responses to the open-ended question were first reviewed to examine trends in
the responses and to develop an initial list of 13 categories. Both authors then
independently coded all of the responses within these categories. These categories
were eventually collapsed into the four categories shown in Table 4 due to overlap
between the categories.

Any discrepancies were discussed until coming to agreement. Initial coding
discrepancies occurred in less than four percent of the cases based upon the final
categories. Behavioral responses included teacher/student interactions, changing
the child's location in the classroom, adjusting classroom activities or tasks, or
developing a behavior chart or reward system. Clinical responses included
suggesting the child take Ritalin, referring the child to be tested for ADHD, or
referring the child to a school counselor or other special services such as special
education. Parental responses included anything related to parental involvement or
a parental conference. In Table 4 the means are broken down by open-ended responses
for each scenario. Open-ended behavioral responses comprised 62.4% of the total
responses for Scenario 1, 19.4% of the responses were clinical, 15.1% were
parental, and 3.2% suggested using some form of punishment. Open-ended behavioral
responses comprised 73.1% of the total responses for Scenario 2, 17.2% of the
responses were clinical, 7.5% were parental, and 2.2% suggested using some form of
punishment. Open-ended behavioral responses comprised 52.7% of the total responses
for Scenario 3, 21.5% of the responses were clinical, 12.9% were parental, and
12.9% suggested using some form of punishment. The scenario that included the least
amount of ADHD-like indicators (Scenario 2) had the highest percent of behavioral
intervention responses. Scenario 3 had the lowest percent of behavioral
intervention responses even though it was in the middle with regard to ADHD-like
indicators. It appears likely that the participants perceived the student in this
scenario to be the least controllable based upon the description given in the text
even though evidence of hyperactive behavior was not described from multiple
contexts or before the age of seven. Mean ranks for each of the eight intervention
options are provided in Table 5 below.

The two most preferred options across scenarios by students were utilizing a
behavior monitoring plan and reinforcing appropriate behavior. The two least
preferred options across scenarios included referring the child for special
education services and suggesting that the child take Ritalin. Not surprisingly,
there was consistency between the open-ended responses and the ratings for the
intervention options. As in the first phase of the study, there was no
differentiation between scenarios in the ratings provided by the teachers.
Paired-samples t-tests were conducted and revealed no significant differences
between means for any of the scenarios. Therefore, manipulation of ADHD-like
symptoms had no effect upon teachers' selection of interventions.

Correlational Statistics

Correlational statistics for Phase II variables are provided in Table 6. The
kind of person variable was related to responses on the third scenario. No
significant correlations were found though between the kind of person variable and
intervention options for scenarios two or three. Students rating behavioral
interventions higher also tended to hold incremental views on a person's
attributes. The intelligence variable did not show a relationship with any of the
scenario variables.

Inferential statistics

In order to address our first two research questions we conducted t-tests and
repeated measures ANOVA procedures to examine differences in intervention
preference for each scenario and the level of consistency in the ratings across the
scenarios. A comparison of the preference for behavioral interventions versus
clinical interventions revealed significant differences in favor of the behavioral
interventions across all three scenarios (Scenario 1: t (91)=-10.43,
p <.001; Scenario 2: t (91)=-9.98, p <.001;
Scenario 3: t (91)=-7.66, p <.001. Repeated measures ANOVA
revealed no differentiation between scenarios in the ratings provided by the
teachers for either behavioral interventions, F (2, 182)=1.25, p
=.289, or clinical interventions, F (2,182)=1.15, p =.319.
Therefore, manipulation of ADHD-like symptoms had no effect upon teachers'
selection of interventions. These results found with pre-service teachers
replicated those found with the teachers in Phase I.

src="../../images/uparrow.gif" alt="Arrow Up" align="right" border="0" height="13"
width="20" />

Discussion

This study sought to examine what teachers believe to be the most effective
interventions when working with students who exhibit hyperactive behavior that
disrupts the classroom. We investigated the extent to which 1) teachers chose
behavioral versus clinical solutions when working with hyperactive children; 2)
whether teachers differentiate between scenarios that have varying levels of
ADHD-like behavior as reflected by their choice of different interventions; and 3)
whether implicit beliefs such as entity/incremental theories show relationships
with the choice of interventions.

The elementary school teachers in Phase I of the study and pre-service teachers
in Phase II of the study overwhelmingly chose behavioral rather than clinical
interventions as appropriate strategies to deal with the disruptive child. This is
an encouraging finding from a pedagogical standpoint. The teachers in Phase I
ranked the behavioral strategies as the “most appropriate” options and
pre-service teachers in Phase II considered these strategies to be
“ultimately the most successful” of the available options. The four
most highly rated intervention options for the scenarios in both phases of the
study were all behaviorally-based strategies. This indicates that teachers, in
general, have efficacy for making positive environmental changes in the classroom
that will curb misbehavior with hyperactive students. This finding also dispels the
notion that teachers, at least from this sample, see a “quick fix”
clinical or medical model option as the most appropriate means of curbing
hyperactive behavior. Finally, these results show alignment with studies suggesting
the need for more balanced approaches to working with hyperactive behavior (DuPaul
& Eckert, 1997; Reid & Maag, 1998).

The teachers in this sample did not, however, differentiate in their suggested
interventions across the scenarios despite having different levels of ADHD-like
symptoms present. This finding may be due to the lack of knowledge that teachers
have about the technical qualifications related to the diagnosis of ADHD. It might
also be that teachers would maintain their selection of strategies regardless of
the diagnosis and/or ADHD-like symptoms displayed by the student. Teachers may
consider their actions as independent of the clinical diagnosis choosing instead to
focus on interventions intended to alter classroom behavior.

Surprisingly, Dweck's entity/incremental scales did not show any consistent
pattern of relationships with the choice of intervention selected by the teachers.
We predicted that teachers adopting an incrementally-based perspective on
intelligence and person attributes would tend to choose a greater proportion of
behavioral to clinical strategies than their entity-based counterparts. This
hypothesis was only partially supported through correlational findings. In the
first phase of the study the teachers did show this pattern for the second and
third scenarios. No such pattern was revealed for the intelligence variable. In the
second phase of the study this hypothesized pattern was found only for the kind of
person variable on the third scenario. These findings appear to only partially
replicate findings from other studies in education conducted by Dweck and her
colleagues (Dweck, 1999). It is possible that the teachers, including those with
entity viewpoints, answered in a way that they perceived to be socially acceptable.
It could also be that personal experiences in the field of education have led them
to feel obligated, to some extent, to respond in a certain way (i.e., present a
behavioral strategy). It is also possible that the instrument is not sensitive
enough to reveal any meaningful relationships that may be present. Hyperactivity
may possess enough domain-specific variance associated with it as a unique facet of
behavior to warrant more specific items in the inventories.

Finally, an interesting finding in the first phase of the study was that
teachers with more experience tended to report a more malleable view about the kind
of person one is. It is possible that it is the more experienced teachers possess a
greater reservoir of experiences seeing changes in student behavior and how
individual attributes can be changed in an educational context. Also, this finding
may indicate the need for teacher education programs to place a greater curricular
emphasis on incremental beliefs by emphasizing evidence-based research that shows
the positive benefits of specific behavioral strategies and interventions.

Limitations of the Study

The overall findings of this study are limited somewhat by the nature of the
task that the teachers completed. First, the scenarios were of fictional students
rather than actual students in a classroom. We cannot expect to fully approximate
the intricacies of working with real-life students through case-based scenarios. In
real-life contexts the teachers would most likely make decisions regarding students
with hyperactive issues or ADHD-like symptoms in conjunction with a team of other
teachers and administrators. This study hoped to measure the general reactions of
teachers when posed simulated behaviors. Second, due to the nature of the study and
time constraints the scenarios were limited in their length and amount of detail.
Ideally, one would prefer to allow each teacher to get a rich description of each
student with the opportunity to observe actual behavior. Future studies could make
important gains by examining the interactions of teachers with students in their
natural contexts over extended periods of time.

Future Studies

This study leads to numerous questions for future research. One avenue might
involve delving more fully into the tendency for teachers to adopt differential
classroom strategies when working with ADHD versus non-ADHD students. In this study
teachers did not suggest different interventions even when the ADHD symptoms varied
by scenario but one variation to that approach might be to indicate to teachers
that the student in the scenario has been professionally diagnosed as having ADHD.
It would be possible to systematically vary this information to see if this leads
to a different selection of interventions. If findings were to reveal that
educators maintain consistent classroom strategies across ADHD students and
non-ADHD students alike there would exist a schism between their beliefs and
summary findings and recommendations by the National Institute of Health (1998). In
their summary statements they report that a combined intervention program of
behavioral treatments with medication have added little to strictly medical
interventions alone. This examination of intervention selection might be furthered
even more with a comparison of special education and regular education
teachers.

Finally, the exploration of implicit beliefs systems and their relationship
interventions and strategies for working with disruptive students is still in its
infancy. This study was a first attempt to look at the basic relationships between
entity/incremental theories and intervention strategies. From these preliminary
findings it appears that measurement instruments need to be developed with a
greater focus on the domain specific aspects of working with disruptive students.
Furthermore, greater understanding might be gained by the observation of teachers
holding various beliefs in their interactions with students in the classroom over
extended periods of time.

src="../../images/uparrow.gif" alt="Arrow Up" align="right" border="0" height="13"
width="20" />

Author

John Nietfeld is an Assistant Professor of
Educational Psychology at North Carolina State University . His research interests
include metacognition, motivation, and adult reading.Correspondence should be sent
to John Nietfeld, North Carolina State University, Educational Psychology, 602D Poe
Hall, Raleigh, NC 27695; john_nietfeld@ncsu.edu; phone (919) 513-7444; fax (919)
513-1687 .

Angie Hunt is a graduate student in Counseling &
Educational Psychology at the State University of West Georgia. Her interests are
in teacher perceptions and training related to behavior management.