Contemporary Educational Psychology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Contemporary Educational Psychology

journal homepage: www.elsevier.com/locate/cedpsych

Professional development in Self-Regulated Strategy

Development: Effects on the writing performance
of eighth grade Portuguese students q
Isabel Festas a,⇑, Albertina L. Oliveira a, José A. Rebelo a, Maria H. Damião a, Karen Harris b, Steve Graham b
a
Faculty of Psychology and of Sciences of Education, University of Coimbra, Portugal
b
Mary Lou Fulton College of Education, Arizona State University, Tempe, AZ, USA

a r t i c l e i n f o a b s t r a c t

Article history: We examined the effects of the Self-Regulated Strategy Development (SRSD) for opinion essay writing
Available online xxxx among 380 eighth grade students in six urban middle schools in a major city in Portugal. Fourteen teach-
ers in six urban middle schools in Portugal participated in the present study; 7 of these teachers partic-
Keywords: ipated in practice-based professional development (PBPD) in SRSD before implementation, and follow-up
SRSD support once instruction began. Schools were matched in pairs based on SES and teacher characteristics;
Writing a member of each pair was randomly assigned to either: (a) teacher led SRSD instruction for opinion
Professional development
essay writing; or (b) teacher implementation of the schools’ existing curriculum and language program
Self-Regulated Strategy Development
Instruction
prescriptions for opinion writing. Students in the experimental schools were taught strategies for plan-
ning and composing opinion essays once a week in 45 min sessions, over a three-month period. Multi-
level modeling for repeated measures indicated SRSD instructed students made statistically greater
gains in composition elements than the comparison students immediately after instruction and two
months later. Teachers implemented SRSD with fidelity and teachers and students rated the intervention
favorably. This study provides initial evidence for replication of the effects of PBPD and SRSD outside of
the United States. Limitations, lessons learned, and directions for future research are discussed.
Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction grammar, capitalization, punctuation, spelling, and sentence

Writing is an important skill that cuts across the school curric-
ulum and is useful for a variety of functions in daily life. Although
writing is important and challenging to learn, in Portugal as in the
United States (e.g., Gilbert & Graham, 2010; Harris, Graham,
Brindle, & Sandmel, 2009), its teaching has been neglected. In Por-
tugal, recent reform of the language arts curriculum resulted in
new guidelines and standards for language arts instruction
(Ministério da Educação e Ciência/Ministry of Education and
Science, 2009, 2012). These guidelines recognize the importance
of writing, including it as a priority area of instruction. They
require not only the development of writing skills related to
q Although the importance of writing has been recognized in the
This research was supported by European FEDER funding through COMPETE:
(Operational Program for Competitiveness Factors) FCOMP-01-0124-FEDER-
022660 and by national funding through FCT (Fundação para a Ciência e
Tecnologia/Science and Technology Foundation) under the project PTDC/CPE-CED/
102010/2008.
⇑ Corresponding author. Address: Faculty of Psychology and of Sciences of
Education, Rua do Colégio Novo, Apartado 6153, 3001-802 Coimbra, Portugal. Fax:
+351 239 851465.
E-mail address: ifestas@fpce.uc.pt (I. Festas).
construction, but also development of writing processes related
to the organization of the text, including planning and revision.
Writing across multiple genres (e.g., narrative, expository, infor-
mative, opinion essay, argumentative) is also emphasized. The
Standards for Elementary and Middle Grade Levels (2012), for
instance, require the instruction on specific attributes for different
genre texts, (e.g., premise, reasons, elaborations, and conclusion for
opinion essay). These Standards also require development of high
quality writing products and the evaluation of writing, but neither
specific instructional approaches nor time dedicated to writing
instruction are prescribed by the Portuguese curriculum. Teachers
are free to choose the teaching methods they use in their
classrooms.
Portuguese curriculum, teachers have not been trained to teach
writing strategies (Almeida, 2012; Almeida & Simão, 2007) and
students have difficulty planning and revising their writing. As in
the United States, (National Center of Educational Statistics,
2012), national data in Portugal indicates Portuguese students
experience severe problems mastering writing (Gabinete de
Informação e Avaliação do Sistema Educativo/Office of

http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
0361-476X/Ó 2014 Elsevier Inc. All rights reserved.

Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
2 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx
Information and of Evaluation of Educational System, 2005; Report
of Gabinete de Avaliação Educacional/Office of Educational
Evaluation 2011, 2012). Difficulties with writing persist at least
until university level, as research indicates many Portuguese uni-
versity students do not plan their writing (Carvalho & Pimenta,
2005) and revise only superficial text features such as spelling
and punctuation (Festas, Damião, & Martins, 2010).
Thus, improving writing abilities and developing effective
instructional procedures to help overcome problems with learning
to write are national priorities in Portugal. The use of evidence-
based practices in schools is critical to achieving this goal (Cook,
Smith, & Tankersley, 2012). Supported by rigorous studies and
research, evidence-based practices are a useful means for improv-
ing teaching and their application has been recommended by gov-
ernment policies in many countries (Cook et al., 2012). This is also
the case in Portugal, where evidence-based practices are empha-
sized in the standards-based reform movement and recommended
by current government policy (Ministério da Educação e Ciência/
Ministry of Education and Science, 2012).
In the writing domain, one of the most effective evidence based
methods for writing instruction is Self-Regulated Strategy Devel-
opment (SRSD) (Baker, Chard, Ketterlin-Geller, Apichatabutra, &
Doabler, 2009; Graham & Perin, 2007; Institute of Education
Sciences, 2012; National Center on Intensive Intervention, 2013).
SRSD is appropriate to the needs of Portuguese pupils and the
demands of the current Portuguese language arts curriculum and
Standards. In fact, as we address next, SRSD includes the teaching
of writing processes and self-regulation strategies, features which
have been neglected in Portuguese schools and that are essential
components of proficient writing.
1.1. Self-Regulated Strategy Development (SRSD)
Developed by Karen Harris and pioneered by Harris and Gra-
ham more than 30 years ago (Harris, 1982; Harris & Graham,
1992, 1996; Harris, Graham, Mason, & Friedlander, 2008), SRSD is
an approach that combines the teaching of writing processes
(including planning, drafting, composing, revising and evaluating);
instruction in writing strategies; and development of self-regula-
tion strategies, including goal-setting, self-assessment (self-moni-
toring and self-recording), self-instruction, and self-
reinforcement. At the same time, SRSD helps students develop
the knowledge and skills needed to use these strategies and purpo-
sively develops self-efficacy for writing, attributions to strategy
knowledge and effort, and motivation for writing (Harris et al.,
2009). SRSD is a complex, multicomponent intervention based on
integrating multiple theories and lines of research which have
been detailed elsewhere (Harris & Graham, 2009; Harris et al.,
2009). Of particular importance to the present study, research indi-
cates this approach is effective when teaching typically developing
writers in a wide range of grade levels, from elementary to high
school (Graham, Harris, & McKeown, 2013).
SRSD promotes writing development through the explicit, situ-
ated, scaffolded instruction of genre-based and general writing
strategies and self-regulation strategies. Specific writing strategies
for multiple genres, such as story, personal narrative, expository,
opinion, and persuasive essays have been developed (Harris
et al., 2008). Such strategies for writing and self-regulation are
developed in six recursive, interactive, individualized instructional
stages with gradual release of responsibility for writing to stu-
dents: (1) develop and activate background knowledge; (2) discuss
and describe the strategies to be learned; (3) model the strategies;
(4) memorize the strategies; (5) support the strategies; and (6)
independent performance (Harris et al., 2008). Instruction pro-
ceeds based on students’ progress; students are given the time
they need to make these strategies their own. Procedures for
Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
maintaining what has been learned and determining how to use
this knowledge across writing tasks are integrated throughout
the stages of instruction.
SRSD has proven to be a powerful instructional approach. Its
application and effectiveness have been investigated in over 100
studies (Graham et al., 2013), and a number of meta-analyses have
examined its impact on students’ writing. Some of these meta-
analyses focused on strategy instruction in writing in general,
including SRSD studies as part of the analysis (Graham, 2006b;
Graham & Harris, 2003), whereas other reviews were broader in
scope and examined a broader range of writing treatments, includ-
ing strategy instruction and SRSD (Graham, McKeown, Kiuhara, &
Harris, 2012; Graham & Perin, 2007; Rogers & Graham, 2008).
These meta-analyses included studies conducted with students
with learning disabilities as well as with poor, average or good
writers in both special and regular school settings (Graham,
2006b; Graham & Harris, 2003; Rogers & Graham, 2008) or only
students from regular school classrooms (Graham et al., 2012;
Graham & Perin, 2007). Some of these reviews focused just on
the elementary-levels (Graham et al., 2012), others on elementary
and middle school pupils (Graham & Harris, 2003) and still others
on pupils from elementary to 12th grade (Graham, 2006b; Graham
et al., 2013; Graham & Perin, 2007; Rogers & Graham, 2008).
Across these meta-analyses, SRSD was found to be a highly
effective instructional practice, and it yielded better results than
other writing instructional methods, including other methods for
teaching writing strategies. Large effect sizes (ES) – above .80 –
were found in true and quasi-experimental studies (Graham,
2006b; Graham & Harris, 2003; Graham et al., 2013, 2012;
Graham & Perin, 2007), and a high percentage of non-overlapping
data (PND) – almost above 90% – was obtained in single-subject
design studies (Graham, 2006b; Graham et al., 2013; Rogers &
Graham, 2008).
1.2. The present study
While previous research has demonstrated that SRSD is a pow-
erful tool for improving students’ writing, the present study was
designed to address limitations in the data base on SRSD. First,
most of the true- and quasi-experiments that have tested the effec-
tiveness of SRSD in writing have involved children in the elemen-
tary grades (cf. Graham et al., 2013), and students have typically
received SRSD instruction in small groups or one-on-one.
Furthermore, and also of particular importance to the present
study, instruction in nearly all of the published studies on SRSD
was delivered by trained graduate assistants (Harris et al., 2009).
Only three published studies have involved general education tea-
cher implementation of SRSD in the middle school classroom (De
La Paz, 2005; De La Paz & Graham, 2002; Wong, Hoskyn, Jay,
Ellis, & Watson, 2008). These studies, however, involved only 2–4
classroom teachers, and little information was provided as to
how teachers were prepared to use SRSD in their classrooms.
Only one published study was found that focused on profes-
sional development for implementation of SRSD classwide in the
general education classroom (Harris et al., 2012). This randomized
controlled study involved 20 s and third grade teachers who partic-
ipated in practice-based professional development (PBPD, cf. Ball &
Cohen, 1999; Grossman & McDonald, 2008) in SRSD for either story
or opinion essay writing (each genre served as the control condi-
tion for the other genre). PBPD focuses on teacher development
of knowledge, understanding, and skills regarding an effective edu-
cational practice before they use it, with support once classroom
use begins (cf. Ball & Cohen, 1999). PBPD rejects traditional
approaches to professional development that are short-term and
top down, do not allow teachers to actively engage in the practices
 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx 3

they are learning, and do not provide much support during imple- prescriptions for opinion writing. SRSD teachers in this study
mentation, which traditionally is primarily done in isolation. participated in PBPD before implementing SRSD, and received
Congruent with sociocultural and social cognitive theories that support and feedback throughout classroom instruction.
stress the importance of meaningful learning in situated contexts, We anticipated that SRSD would have a positive impact on Por-
PBPD has six critical characteristics: (a) collective participation of tuguese students’ opinion essays in terms of structural elements
teachers within the same school with similar needs; (b) basing and writing output, and that SRSD instruction in this study would
professional development around the characteristics, strengths, result in a significant increase in the number of opinion essay ele-
and needs of the students in these teachers’ current classrooms; ments in students’ writing as well as a significant increase in the
(c) attention to content knowledge needs of teachers, including overall length of student essays. Such results have been found in
pedagogical content knowledge; (d) opportunities for active learn- other studies involving the opinion essay writing strategies taught
ing and practice of the new methods being learned, including in this study (e.g., Mason, Kubina, & Taft, 2009). We expected that
opportunities to see examples of these methods being used and these effects would be evident immediately after the intervention
to analyze the work; (e) use of materials and other artifacts during and after a two-month interval (such maintenance effects have
professional development that are identical to those to be used in been found in other studies; Graham et al., 2013). We further
the classroom, and (f) feedback on performance while learning, and expected that following PBPD, teachers would implement SRSD
before using these methods in the classroom, so that understand- with fidelity based on the study by Harris et al. (2012). Finally,
ings and skills critical in implementation are developed. In the we predicted that students and their teachers would be positive
Harris et al. (2012) study, teachers who received PBPD imple- about the effects of the intervention, indicating social validity in
mented SRSD instruction that resulted in significant and meaning- Portugal, based on our anticipation that students’ opinion essay
ful changes in student writing outcomes for both story and opinion writing would improve.
essay writing. Teachers implemented SRSD with fidelity, and SRSD
was viewed as socially valid by teachers and students.
2. Methods
Finally, the few published studies of teacher implementation of
SRSD have all been conducted in the United States, as have
2.1. Setting
the majority of studies on SRSD. Thus, there is a need to test the
generalizability of SRSD results across populations outside of
The study took place in six urban schools in a major city in Por-
the United States. Similarly, PBPD has not been tested outside of
tugal. Three of the schools were middle schools; the other three
the United States. If PBPD for SRSD’s promise is to be actualized,
combined middle and high school grades.
it is important to test its effectiveness in different contexts. Aggre-
As usual in Portugal, all of the schools involved in the study fol-
gated, consistent findings across a variety of studies in a variety of
lowed the national curriculum, as previously described. In an
contexts are necessary to argue that an intervention is reliable and
attempt to increase the likelihood that the experimental and com-
generalizable (Robinson, Levin, Schraw, Patall, & Hunt, in press).
parison groups were equivalent, pairs of schools were matched for
Recommendations for educational practice must be based upon
socioeconomic status prior to the intervention (Fraenkel & Wallen,
such aggregated finding and research syntheses, thus replication
2003). Data were collected from the School Educational Project,
and extension of previous research is critical to the field.
which reports students’ socioeconomic status; SES is defined by
Despite the difficulties Portuguese students experience learning
the level of schooling completed by students’ parents and parents’
to write, evidence-based practices in writing have rarely been
professions. School 1 (middle school with 838 students) was
tested in this country. Our study extended previous SRSD research
matched with School 2 (middle school with 690 students). Accord-
by examining if SRSD instruction was effective when it was imple-
ing to the School Educational Project data, the majority of students
mented by Portuguese teachers who received PBPD for SRSD, and
in both schools were from a medium to high socioeconomic class.
conducted classwide with Portuguese middle school students.
School 3 (middle/high school with 991 students) was matched
We examined the effects of SRSD instruction, implemented by
with School 4 (middle school with 560). According to the School
classroom teachers following PBPD, on the opinion essay writing
Educational Project, both schools had a more heterogeneous popu-
performance of eighth grade Portuguese students. Previously, we
lation from low, medium and high socioeconomic groups. School 5
had adapted SRSD lessons and materials to the Portuguese lan-
(middle/high school with 950 students) was matched with School
guage and to Portugal’s educational context (Rebelo et al., 2013).
6 (middle/high school with 700 students). According to the School
Eighth grade was selected as the target grade level because such
Educational Project, these students had medium/high socioeco-
instruction should help these students prepare for the national
nomic status.
exam that is taken in the ninth grade. Writing opinion essays is
The schools were also matched on teacher characteristics
an important part of the school curriculum in eighth- and ninth-
including gender, experience, and preparation. All of the teachers
grade and is targeted in writing tests and the Portuguese national
completed a questionnaire to elicit information about: gender,
exam. Criteria used in the national exam include evaluation of
age, licensing, writing training received during teacher preparation
opinion essays’ structural elements (premise, reasons and conclu-
and professional career, teaching procedures, and teaching experi-
sions) (Gabinete de Avaliação Educacional/Educational Evaluation
ence. One of each pair of schools was randomly assigned to the
Office, 2013), which are an important part of SRSD for opinion
experimental condition and the other to the comparison group.
essays. Furthermore, the language arts teachers in the participating
The necessary consents were obtained from the Portugal Ministry
schools requested a focus on opinion essay writing.
of Education, the Director of each school, from the teachers
Fourteen teachers in six urban middle schools in Portugal par-
involved in study, the students’ parents, and the pupils enrolled
ticipated in the present study; 7 of these teachers participated in
in the experiment before the study started.
PBPD for SRSD. The six schools were matched in pairs based on
socioeconomic status and homogeneity/heterogeneity of students
and on teacher characteristics including gender, experience, and 2.2. Teachers
preparation. One member of each pair was randomly assigned to
one of two conditions: (a) classroom teacher implemented SRSD Fourteen teachers in the six schools took part, seven each from
instruction for opinion essay writing; or (b) teacher implementa- the experimental and comparison schools. Teachers in the experi-
tion of the schools’ existing curriculum and language program mental and comparison conditions had very similar characteristics.

Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
4 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx
All teachers were female and all taught Portuguese language arts.
Additionally, all of the teachers had an undergraduate degree
obtained from a university in Portugal and only one teacher (in
the comparison group) had a master’s degree (no difference was
found concerning this variable: Fisher’s Exact Test = 1, p = .50).
All of the teachers held a teaching license and all had served as
teachers for a long period of time, ranging between 16 and 34 years
in the experimental group (M = 27; SD = 6.56) and between 23 and
39 years in the comparison group (M = 29.33; SD = 5.54). No statis-
tically significant difference was found here between the two
groups (t(12) = .686, p = 0.51]. Both groups included teachers
who taught more than one section of language arts and teachers
who taught only one section: in the experimental group there
was 1 teacher with 3 sections, 4 teachers with 2 sections, and 2
teachers with 1 section; in the comparison group there was 1 tea-
cher with 3 sections, 3 teachers with 2 sections, and 3 teachers
with 1 section. Class size ranged from 10 to 23 students in the
experimental group and from 9 to 24 students in the comparison
group.
In Portugal, there are no national guidelines for teacher prepa-
ration programs regarding preparation for teaching writing.
Research, however, indicates that preparation to teach writing is
poor (Pereira, 2001). None of the teachers had received either spe-
cific preparation in teaching writing in their preparation programs
or professional development on writing instruction during their
careers. In the present study, teachers in the comparison group fol-
lowed the Portuguese language arts curriculum, while teachers in
the experimental group followed the SRSD model for writing
instruction in opinion essays.
2.3. Students
A total of 380 students (214 in the experimental group and 166
in the comparison group) participated in this study. Initially, there
were 507 students in the 14 eighth grade classrooms, with 285 stu-
dents in the experimental classrooms (132 girls, 153 boys) and 222
in the comparison classrooms (118 girls and 104 boys). Only 436
students (some students changed schools and others were not
present during pretesting), however, participated in the pre-test,
which was administered as part of the schools’ regular assessment
procedures; of these, consent and assent to participate was
obtained for 380 students. Members of the research team collected
the following information regarding students from school records:
age, gender, grade retentions, grades in language arts and other
subjects, and special education status; occupation and educational
attainment of parents were also collected from these records.
The age of the experimental group ranged from 11.92 to
15.42 years (M = 13.33; SD = .44) and the age of the comparison
participants ranged between 12.67 and 16.42 years (M = 13.56;
SD = .65). An independent t-test revealed a statistically significant
difference in age [t(265.7) = 3.85, p < .001]. Concerning gender,
103 (48.1%) of the experimental group were male and 111
(51.9%) were female, while in the comparison schools the figures
were 72 (43.4%) male and 94 (56.6%) female. The proportion of
males and females in the two groups did not differ statistically
(p = .36).
In the experimental group, 21 students (10%) had repeated one
grade or more, whereas 18 students (11.1%) in the comparison
group had repeated one or more grades. The number of students
repeating a grade in the two groups did not differ statistically
(p = .72). The mean for the number of grade repetitions in the first
group was .09 (SD = .37), ranging from 0 to 3, and for the compar-
ison schools the mean was .20 (SD = .60) with the same range.
There were four students (1.9%) with special needs in the experi-
mental group and seven in the comparison group (4.3%). These
numbers for the two groups did not differ statistically (p = .20).
Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
The students from the experimental and comparison groups
were also compared on their grades in language arts and overall
school achievement (obtained at the end of the previous school
year). For language arts, grades ranged from 2 to 5 for both groups
(in the Portuguese educational system up to the ninth level, grades
range from a low of one to a high of five), with M = 3.66 (SD = .82)
for the experimental schools and M = 3.68 (SD = .82) for the com-
parison schools (p = 0.75). The general school achievement of the
experimental group was M = 3.82 (SD = .77), and for the compari-
son group it was M = 3.69 (SD = .79), with the scores in both groups
also ranging between 2 and 5. There was no statistical difference
between the grades of the two groups (all p’s > 0.18). Finally,
regarding the level of schooling completed by students’ parents
(ranging from less than compulsory education up to the Ph.D.),
no statistical differences were found for fathers (p = .11) or moth-
ers (p = .25) between experimental and comparison groups.
2.4. Practice-based professional development for SRSD
Before classroom instruction began, SRSD teachers participated
in 14 h of professional development (across two days) in SRSD
instructional practices. Professional development followed the
PBPD model used by Harris et al. (2012). Teachers received note-
books with the guidelines and materials needed to implement all
activities and lessons for opinion essays in their own classrooms.
SRSD instruction was modeled, practiced, and discussed during
professional development. Teachers and the research team also
discussed how to adapt SRSD program and differentiate instruction
to meet student needs in their classrooms (for more detail, see
Harris et al., 2012).
After PBPD, SRSD teachers met with research assistants weekly,
after school for about an hour, to address any questions or
concerns teachers had regarding SRSD instruction and how future
lessons might need to be adjusted to meet teacher and student
needs. The majority of teachers’ questions centered specific SRSD
instructional activities (e.g., how to model the use of the self-
regulation procedures) and on ways to differentiate instruction
for students experiencing difficulty mastering the material taught.
2.5. SRSD instruction
After the opinion writing pre-test was administered, teachers
delivered SRSD instruction. The experimental teachers taught SRSD
for opinion essays for one, 45 min class once a week, from October
until January. Portuguese language arts classes are assigned
225 min a week and school directors can organize the class time
as they wish. Class time is typically divided into 5 blocks of
45 min each. Teachers and researchers agreed that all writing
activities would be conducted only in this one class a week to
ensure that this was the only writing instruction students received
during this study. In the other 180 min allocated to language arts
curriculum, teachers focused on components such as oral and writ-
ten comprehension, reading, and language knowledge in accor-
dance with national guidelines.
The materials for teaching opinion essay writing via SRSD were
taken from Harris et al. (2008), including six lessons from Chapter
8 (lessons 1, 2, 3, 4, and 5 for older students) and one lesson from
Chapter 9 (Lesson 1). We also used other SRSD material from the
Harris et al. book, including mnemonic charts, graphic organizer,
essays rockets graph, and cue cards. We adapted this material so
that it was in Portuguese and appropriate for grade 8 classrooms
in Portugal (Rebelo et al., 2013); we created materials that fol-
lowed and respected the objectives and structure of SRSD but that
would make sense to Portuguese teachers and students.
The mnemonics for the SRSD strategies, in particular, required
careful attention because a literal translation was not possible.
 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx
We tried to find acronyms that had the same significance as the
original mnemonics and matched a Portuguese word. The general
planning strategy involving three steps, represented by the mne-
monic POW (Pick my idea, i.e., pick an initial idea of what to write
about; Organize my notes, i.e., write a plan using a graphic orga-
nizer; Write and say more, i.e., continue modify and upgrade the
plan during writing), was translated to the Portuguese acronym
PODE (Can in English). PODE signifies: Pega (Pick, in English), Orga-
niza (Organize, in English), Dizer e Escrever (Say and Write, in
English).
The genre-specific planning strategy for opinion essays to help
students to carry out the second step of POW (Organize my notes),
TREE (Topic sentence, Tell what you believe; Reasons, three or
more, and elaborate on or say more about each one; Ending, Wrap
it up right; and Examine, do I have all my parts?), was translated to
the Portuguese mnemonic TRAVE (Beam in English). TRAVE, the
Portuguese mnemonic, represents: Tema (Topic sentence, in Eng-
lish), Razões (Reasons, in English), Acabar (Ending, in English), VE
(Ver para Examinar) (See to Examine, in English).
In order to verify that SRSD was implemented with fidelity (i.e.,
SRSD was delivered as intend), we developed lesson checklists.
These checklists were used to monitor which steps of the lesson
were completed by the teachers and should be filled by research
assistants and by teachers. They contained a description of each
activity in a lesson and a place to check if the activity was com-
pleted or not completed.
Teachers carried out the six stages of the SRSD model, recur-
sively as appropriate, with time spent on stage depending on the
needs and rate of progress of their students. In the first stage,
Develop Background Knowledge, the writing and self-regulation
strategies (the general writing strategy POW/PODE, the genre spe-
cific writing strategy TREE/TRAVE, and self-regulation strategies
including goal setting, self-instructions, self-assessment, and self-
reinforcement) were introduced and discussed. At this stage, stu-
dents acquired the knowledge, understanding, and vocabulary
needed to apply the general and genre-specific writing strategies
as well as the self-regulation strategies.
During the Discuss It stage, teachers and students discussed the
importance and utility of using the writing and self-regulation
strategies. Next, teachers modeled how to use POW/PODE and
TREE/TRAVE to write an opinion essay, showing students how to
apply them (Model It stage). Simultaneously, they modeled the
use of the self-regulation strategies, showing and explaining out
loud such things as how to set goals, how and when self-instruc-
tions might be used, how to self-assess, and how to self-reinforce.
If the students had not yet memorized the mnemonics (PODE and
TRAVE) by this point, further practice was provided (Memorize It
stage).
During the Support It stage, teachers supported students
through collaborative writing experiences by helping them to
write opinion essays using POW/PODE and TREE/TRAVE, while
self-regulating the writing task. Gradual release of control led to
the last stage, Independence Performance, where students could
use the strategies to write opinion essays without the teachers’
help.
2.6. Writing instruction in the comparison classrooms
To document how writing was taught in the comparison class-
rooms, we administered a 64 item Writing Activities Questionnaire
focused on writing instructional practices - those recommended in
the eighth grade Portuguese curriculum, and those common in
SRSD1. For each of the 64 items presented on the scale, teachers
1
This scale is available from the authors.
Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
5
indicated how often it occurred in their classrooms across the school
year using a 5-point scale: 0 (never), 1 (up to 30 min a week), 2 (from
31 min to 60 min a week), 3 (from 61 min to 90 min a week), or 4 (more
than 90 min a week). Teachers also were asked if they had received
inservice or preservice instruction in teaching writing. If a teacher
answered Yes, they were asked to describe this instruction.
Finally, teachers responded to an open ended question regard-
ing any other features of their writing instruction they could
report.
Responses to the Writing Activities Questionnaire indicated
that none of the teachers had received preservice or inservice
instruction in teaching writing. The teachers generally followed
the Portuguese language arts curriculum for grade 8 students,
teaching grammar, capitalization, punctuation, spelling, sentence
construction, organization of text, planning, and revision. Teachers
reported spending more time each week on teaching grammar
(one teacher reported spending more than 90 min a week, and
the other six more than 60 min) and punctuation (two teachers
reported spending more than 90 min a week, four from 61 min to
90 min, and one from 31 to 60 min) than on the development of
planning (five reported spending up to 30 min a week and two
from 31 to 60 min) and revision (one teacher answered never,
two reported spending up to 30 min a week, and the other four
reported spending from 31 to 60 min a week).
Teachers reported teaching a variety of text genres, especially
those recommended for this grade level, i.e., narrative, opinion
essay, argumentative, and descriptive. The majority of teachers
reported spending between 61 and 90 min per week teaching text
genre, although all reported spending 31–60 min teaching opinion
essay writing. Narrative writing received the most attention (four
teachers reported spending more than 90 min and three reported
60–90 min).
Regarding elements of instruction included in SRSD, teachers
reported teaching only some features of the planning strategy
POW/PODE (most reported up to 30 min per week) telling students
they should organize their writing but not teaching any strategies
for doing so. Regarding TREE/TRAVE, only two teachers indicated
teaching all of its components. Two teachers reported teaching
only premise and conclusion, one teacher only premise and rea-
sons, one teacher only premise, and the last teacher only reasons.
All of the teachers reported spending less than 30 min per week
on these aspects of opinion essay writing. The two teachers who
reported teaching all TREE/TRAVE components only explained
them to students. None of the teachers reported following the
stages of instruction in the SRSD model. Modeling was referred
as a general method to teach writing, but not specifically reported
for teaching writing opinion essays. Comparison group teachers
did not report development of self-regulations strategies in their
educational practices, including goal setting, self-instructions,
self-assessment, and self-reinforcement. For example, they
reported never teaching students to set goals and break them into
sub-steps (goal setting), ‘‘talk to themselves’’ as they write in order
to guide, organize and structure their writing (self-instruction),
determine whether or not they included all the parts on an opinion
essay (self-assessment), or reinforces themselves (self-
reinforcement).
The times teachers reported regarding components of writing
instruction, however, must be interpreted in light of their
responses to the questionnaire’ open ended question. All of the
teachers explained that they included in these times the total time
they spent on these elements across all of the language arts curric-
ulum components (oral and written comprehension, reading, writ-
ing, language knowledge). Regarding text genre, for instance,
teachers explained they focused on genre not only in teaching
writing, but also when teaching oral comprehension and reading.
The majority of the teachers reported spending up to 30 min a
6 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx
week teaching writing, while some teachers acknowledged that
they did not teach writing every week. Thus, the comparison group
teachers reported generally spending less time a week specifically
teaching writing than the SRSD teachers.
2.7. Assessments
2.7.1. Fidelity of SRSD instruction
Lesson checklists were developed that included all activities
and elements for each lesson. A research assistant observed 25%
of instructional sessions, spread across instruction; this percentage
of lessons observed is higher than the 20% of sessions of school-
based intervention recommended for observation in establishing
fidelity (Gresham, Gansle, Noell, Cohen, & Rosenblum, 1993;
Hulleman & Cordray, 2009). Teachers told the observer where they
would start and added any changes they had made in that lesson
for their students before the lesson began; observers checked off
lesson elements as they were completed. Treatment fidelity for
each lesson was computed by dividing the number of lesson ele-
ments taught by the total number of elements possible and multi-
plying the outcome by 100%. Teachers were given a copy of the
checklist for each lesson, and were asked to check off each step
as it was completed each time they taught. These teacher check-
lists were collected from the teachers at the end of instruction.
2.7.2. Writing assessments
Pre-test, posttest, and maintenance (two months after post-
intervention probes) writing probes were used to evaluate the
effects of SRSD instruction on students’ opinion essay writing. The
probes were administered by students’ classroom teachers, and
the SRSD instruction and comparison groups were assessed at the
same time. Students were given up to 45 min, the full class period,
to complete their paper. Students wrote their opinion essays in
response to writing prompts related to the use of technology (e.g.,
Would it be possible to live without a computer today?) teachers rec-
ommended as suitable and interesting for eighth grade students.
Prior to scoring, each opinion essay was typed into a Word doc-
ument. Identifying information was removed and spelling, capital-
ization, and punctuation were corrected to avoid any influence
from surface level features, such as legibility and spelling errors,
on examiners’ judgments about writing performance (Graham,
1999). Each opinion essay was scored for number of words and
for number of structural elements. Structural elements included
those taught to students in the experimental group: premise, rea-
sons, explanations (why an author believed a particular premise or
why they refuted a counter premise), conclusion, and elaborations
(additional information on or examples of a premise, reason, or
conclusion). For premise and conclusion, ‘‘0’’ was allocated if the
element was absent and ‘‘1’’ if the element was present. For rea-
sons and elaborations, 1 point was awarded for each unique reason
and elaboration included.
Structural elements were assessed for three reasons. First, the
planning strategy students were taught encouraged them to use
these structural markers to generate ideas and think about their
topic (e.g., students were encouraged to generate and consider
multiple reasons to support their premise). Thus, if SRSD instruc-
tion was effective, students’ writing should evidence an increase
in these basic structural elements. Second, as noted earlier,
national assessments in Portugal stress the assessments of these
elements in the evaluation of opinion essays (Gabinete de
Avaliação Educacional/Educational Evaluation Office, 2013). Third,
these structural elements represent the basic building blocks of a
good opinion essay and define this genre (see Scardamalia,
Bereiter, & Goleman, 1982).
Our second writing measure, number of words, is a common
outcome measure in writing intervention research (see for
Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
example Graham et al., 2013), as it is difficult to write a convincing
argument or a good paper without a reasonable number of ideas
and elaborations. While longer text is not always better text, it
often is (Page & Petersen, 1995), as correlations between number
of words and text quality are high (see Bangert-Drowns, 1993;
Morphy & Graham, 2012). It should be noted, though, that text
can be improved when students eliminate extraneous or repetitive
material. While we anticipated that SRSD students would produce
longer text essays in the current study, it was possible that instruc-
tion would result in more succinct text, as the planning process
and self-evaluation processes built into instruction may eliminate
unneeded material. Consequently, we viewed number of words
written as an important outcome variable.
All opinion essays were scored by a trained research assistant;
one-third of the papers were independently scored by a second
trained rater. Inter-rater reliability for the three evaluation points
(pre-test, posttest and maintenance), respectively, was .88, .87
and .77 (M = .84, Cohen’s kappa). Each essay was also scored for
number of words written using the word count tool in Word.
2.7.3. Social validity
Social validity was assessed immediately after instruction.
Teachers completed the Teachers’ Intervention Rating Profile and
students the Students’ Intervention Rating Profile we constructed.
Each rating profile presented a series of statements about the social
validity of the SRSD intervention that teachers and students
responded to using a 5-point Likert-type scale (strongly disagree = 1
to strongly agree = 5). The scale that the teachers completed
included 15 items (e.g., SRSD instruction helped students to write
better opinion essays; After SRSD instruction, students included more
opinion essays’ structural elements in their writings; After the inter-
vention, students showed a greater interest in writing; With SRSD
instruction students became more self-regulated). The student scale
contained 10 items (e.g., SRSD activities were interesting; With SRSD
instruction I learned to write better; With SRSD instruction I learned to
organize my ideas when I am writing; I liked to participate in this
instruction programme). Internal consistency of the Teachers’ and
Students’ Rating Profiles, measured by the Cronbach’ alpha, were
.86 and .88, respectively.
3. Results
We examined whether: SRSD instruction was implemented as
intended (i.e., treatment fidelity), students in SRSD instruction out-
performed students in the comparison condition on structural ele-
ments and number of words written, and teachers and students
viewed the SRSD instruction positively. Means and standard devi-
ations for structural elements and number of words are presented
in Table 1.
For both writing measures, multilevel modeling for repeated
measures was conducted to examine the impact of SRSD in com-
parison to the control condition across pre-test (0), posttest (1),
Table 1
Performance of student participants in opinion essay writing and number of words.
Student performance Time Groups
Experimental Control
(n = 214) (n = 166)
M (SD) M (SD)
Opinion essay Pre 3.56 (3.85) 3.90 (3.45)
Post 6.07 (3.70) 3.62 (2.96)
Follow-up 4.86 (3.20) 2.65 (2.34)
Word count Pre 229.06 (93.30) 178.07 (84.37)
Post 176.47 (71.13) 148.89 (83.90)
Follow-up 167.00 (69.62) 155.41 (95.28)
 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx 7

and maintenance (2) (RStudio, 2013). Each of these analyses pro-
ceeded in the fashion as described below.
For level-1 variables (time of testing and the outcome variable),
we (1) determined the interclass correlation (ICC) looking at
changes over time of testing; (2) examined if scores for the out-
come measure randomly varied among individuals; (3) investi-
gated the form of the relationship between time of testing (i.e.,
pre-test, posttest, and maintenance) and the outcome measure
(i.e., Do scores generally increase, decrease, or show some other
type of relationship with time?); (4) verified whether the relation-
ship between time and the outcome measure was constant among
students or whether it varied on a student-by-student basis; and
(5) examined if the model fit improved by incorporating an autore-
gressive structure with serial correlations and heterogeneity in the
error structure.
We then conducted the analyses by adding the level-2 variables
in order to try to explain the random variation in mean test results
(intercepts) and in time of testing results slope. First, we examined
the impact of adding treatment (SRSD versus comparison) to the
model. Then we examined the effects of treatment, gender, and
age. As noted earlier, there was a small, but statistically significant,
difference between students’ age in the two groups. In addition, we
included gender in the analyses, as girls are typically better writers
than boys at these ages (Graham, 2006a).
For both structural elements and number of words, we ran a
model where the level-2 variables included school along with
treatment, gender, and age. However, the model with only treat-
ment, gender, and age provided a better fit to the data than the
model where school was also included (structural elements:
AIC = 5897.60; BIC = 5973.02; LogLik = 2933.80; p < 0.0001; number
of words: AIC = 12726.25; BIC = 12806.7; LogLik = 6347.126;
p < 0.05). The model where school was added to treatment, gender,
and age did not converge, even though a model with school as the
only level-2 variable was statistically significant. As a result, we
only report the results for the model that included treatment, gen-
der, and age.
3.1. Treatment fidelity
The SRSD writing intervention was implemented with accept-
able fidelity. Observations of 25% of the instructional sessions
indicated that teachers completed 77.86% (SD = 11.61%; range
65–100%) of the writing activities. The lower score occurred in only
one class session where the teacher was dealing with behavioral
issues. In addition, SRSD teachers were asked to check each activity
of a lesson as they completed it, using the same checklists. These
checklists were collected at the end of instruction; teachers
reported completing 82.38% (SD = 19.96%; ranged between 44
and 100) of the lesson elements. Finally, at the end of instruction
all teachers confirmed that they did not teach writing in any other
language arts classes with their students and that they did not use
anything new for writing instruction other than SRSD.
3.2. Structural elements
Analysis of level-1 variables (time of testing and number of
structural elements) resulted in an ICC coefficient value of 14.3%,
indicating that this amount of variance in students’ scores was
explained by properties of the students. Examination of fixed
effects (p < 0.001), slope variability (p < 0.464), autocorrelations
(/ = 0.139) and variances of response changes over time (showing
that a model that allowed for decreases in variance fit the data
better than a model that only included the quadratic effect:
AIC = 6018.80; BIC = 6049.01; LogLik = 3003.40; p < 0.0001)
revealed that: (a) students randomly varied little in mean levels
of structural elements in comparison to how they vary through
time; (b) there was a quadratic relationship between time of test-
ing and structural elements; (c) the strength of the quadratic rela-
tionship did not randomly vary among students and (d) there was
significant variance heterogeneity in the data.
When the level-2 variable of treatment alone was added to try
to explain random variation in mean test results (intercepts) and in
time of testing results slope, we found that the expected test score
for students in the SRSD group was on average 1.74 structural ele-
ments higher than the expected score for students in the compar-
ison group (p < 0.001). When the level-2 variables of sex, and age
were subsequently added to the analysis (along with treatment;
see Table 2), a statistically significant interaction between treat-
ment and time of testing was obtained at both posttest and main-
tenance (as expected), with SRSD students making statistically
greater gains from pre-test to posttest (p < 0.001) as well as pre-
test to maintenance (p < 0.001) than comparison students. While
the interaction between gender and treatment was not statistically
significant, girls in the study included on average 0.51 more struc-
tural elements than boys (p = 0.025). In addition, with all three
level-2 variables included in the analyses, SRSD students included
1.4 more structural elements in their opinion essays than compar-
ison students.
3.3. Number of words
Analysis of level-1 variables (time of testing and number of
words) resulted in an ICC coefficient value of 56.4%, indicating that
this amount of variance in students’ scores was explained by prop-
erties of the students. Examination of fixed effects (p < 0.001),
slope variability (p < 0.0001), autocorrelations (/ = 0.7) and

Table 2
Multilevel regression modeling analyses of the relationship between treatment, age, and gender and time of testing for structural elements in opinion essays.

b SE DF T P
Model with Treatment, Age, and Gender (Intercept) 3.657 0.197 752 18.560 0.000
Treatment 1.400 0.231 376 6.055 0.000
Age 0.383 0.209 376 1.836 0.067
Gender 0.508 0.266 376 2.253 0.025
Timea 22.650 5.398 752 4.196 0.000
Timeb 10.941 5.311 752 2.060 0.040
Treatment  Timea 36.725 6.333 752 5.799 0.000
Treatment  Timeb 23.890 6.231 752 3.834 0.000
Age  Time (0–1) 9.759 5.712 752 1.715 0.087
Age  Time (0–2) 1.835 5.620 752 0.327 0.744
Gender  Time (0–1) 9.236 6.183 752 1.494 0.136
Gender  Time (0–2) 10.999 6.083 752 1.808 0.071

Note: Treatment = SRSD versus control group; 0 = pre-test; 1 = posttest; 2 = maintenance.

a
Time = Linear effect of time;
b
Time = Quadratic effect of time.

Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
8 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx

Table 3
Multilevel regression modeling analyses of the relationship between treatment, age, and gender and time of testing for number of words in opinion essays.

b SE DF T P
Model with Treatment, Age, and Gender (Intercept) 178.379 6.19 752 28.815 0.000
Treatment 28.047 7.263 376 3.862 0.000
Age 15.959 6.55 376 2.437 0.015
Gender 36.931 7.090 376 5.221 0.000
Timea 351.954 98.59 752 3.570 0.000
Timeb 302.809 78.203 752 3.872 0.000
Treatment  Timea 542.889 115.66 752 4.694 0.693
Treatment  Timeb 72.047 91.748 752 0.785 0.756
Age  Time (0–1) 41.22 104.31 752 0.395 0.000
Age  Time (0–2) 25.765 82.749 752 0.311 0.433
Gender  Time (0–1) 74.604 112.91 752 0.661 0.509
Gender  Time (0–2) 66.591 89.571 752 0.743 0.457

Note: Treatment = SRSD versus control group; 0 = pre-test; 1 = posttest; 2 = maintenance.

a
Time = Linear effect of time.
b
Time = Quadratic effect of time.

variances (AIC = 12859.25; BIC = 12899.54; LogLik = 6421.623;
p = 0.065) revealed that: (a) students varied more in their mean
levels of number of words in comparison to how they varied
through time; (b) there was a quadratic relationship between time
of testing and number of words; (c) the strength of the quadratic
relationship varied randomly among students; (d) the variance
heterogeneity in the data was not modeled.
When we added the level-2 variable of treatment to try to
explain random variation in mean test results (intercepts) and in
time of testing results slope, we found that the expected score
for students in the SRSD group dropped an average of 27.8 words
more than the expected number of words for students in the com-
parison group (p < 0.001). When the level-2 variables of sex and
age were subsequently added to the model (along with treatment;
see Table 3), a statistically significant interaction between treat-
ment and time of testing was obtained (p < 0.001), with students
in the SRSD condition showing a greater drop in number of words
written from pre-test to posttest than students in the comparison
condition. Such a statistically significant difference, however, was
not evident from pre-test to maintenance (p = 0.756). It is impor-
tant to note that the average length of essays decreased from
pre-test to posttest to maintenance for students in both groups
(see Table 1). While the interaction between gender and treatment
and age and treatment were not statistically significant, girls in the
study wrote on average 36.9 more words than boys (p < 0.001), and
students who were one year older wrote 16 words less than their
younger counterparts (p = 0.015). Moreover, with all three level-2
variables included in the analyses, SRSD students wrote 28 words
less than comparison students.
3.4. Social validity
Responses to the Teachers’ Intervention Rating Profile and the
Students’ Intervention Rating Profile were positive. Teacher
responses (on a 5-point Likert-type scale from strongly disagree = 1
to strongly agree = 5) ranged from 4 to 5 (M = 4.58; SD = .41). Stu-
dents’ perceptions of treatment acceptability were also positive
(M = 3.80; SD = .51) with item scores ranting from 2.07 to 4.93.
4. Discussion
Given the complexity of writing, it seems reasonable to expect
special care needs to be taken when teaching it. Nevertheless,
the teaching of writing is neglected in Portugal (Pardal & Festas,
2011) and the United States (Gilbert & Graham, 2010). We pro-
vided intensive professional development in SRSD for opinion
essay writing to teachers of eight grade students following the
PBPD model (Ball & Cohen, 1999; Grossman & McDonald, 2008;
Harris et al., 2012). Students who received SRSD instruction in
opinion essay writing, as compared to students who received typ-
ical classroom instruction following Portuguese curriculum guide-
lines, showed meaningful improvements in the inclusion of basic
structural elements (premise, reasons, conclusion, and elabora-
tions) in their papers. These findings are consistent with earlier
research in the United States, where similar SRSD professional
development was provided (e.g., Harris et al., in press, 2012).
SRSD did not increase the length of students’ compositions.
While the length of SRSD instructed students’ essays decreased,
after SRSD instruction these students included most or all of an
opinion essay’s critical elements in their compositions. Their com-
positions became more organized and inappropriate text was elim-
inated, which may explain the lack of an effect on length. Previous
SRSD research includes mixed results on length; several studies
have found that after SRSD instruction students wrote better, but
not longer, compositions (Harris et al., 2009, 2012). Length, or writ-
ing more, was not a goal for students in this study (as it was not in
previous studies); this goal might be addressed in future research
once initial competency in a genre is established.
SRSD was implemented with fidelity and both teachers and stu-
dents reported strong social validity. Teachers believed that SRSD
had a positive impact on the writing of their students; students
were positive about the instruction they received and found it very
interesting. Treatment fidelity and social validity results are very
important to the continuation and expansion of research on SRSD
and to enabling its future implementation in Portugal and other
countries.
Importantly, this study demonstrated that PBPD for SRSD was
effective when applied in a different cultural context, that of Portu-
guese schools. It is important to note, however, that we adapted
the SRSD lessons used in this study so that they would fit within
the Portuguese educational context. The process of adapting pro-
grams to other cultures, and researching their effects, needs
greater attention if evidence-based practices are to be generalized
and used globally. This study provides initial evidence for the gen-
eralizability of PBPD for SRSD outside of the United States.
4.1. Examining the effectiveness of SRSD
Teachers in the comparison group did use some methods in
their classrooms common to those used in SRSD instruction (see
Section 2.6). Thus, examining the differences in instruction
between the SRSD and the comparison teachers can help build
understanding as to why SRSD was effective in this study. We
begin by noting one additional critical factor –teacher preparation
for teaching writing. As noted previously, none of the teachers in
this study reported receiving specific preservice or inservice

Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx
preparation in teaching writing. PBPD for SRSD supports teachers
in developing deep understanding of the general and genre specific
writing strategies to be taught, the changes in performance that
can be expected in their students, and evidence-based methods
for developing these strategies. In the United States, while most
teachers report some preparation in teaching writing, most prepa-
ration occurs after initial certification and many teachers report
that they are not adequately prepared to teach writing (Brindle,
2012; Graham & Harris, in press). Teaching writing is complex
and challenging; teaching it skillfully requires considerable knowl-
edge about what to teach, when to teach it, and how to teach it.
PBPD for SRSD helps meet these needs.
In addition, there were many meaningful differences between
the SRSD instruction and business as usual in the comparison
classrooms in this study. Comparison teachers in the present study,
as compared to those who received PBPD for SRSD, spent more
time teaching grammar and punctuation than they did teaching
planning; did not devote as much time to teaching opinion essay
writing; told students to organize their writing but did not offer
or teach them any strategies for doing so; most commonly
reported teaching students only 1–2 of the common elements of
opinion essays; explained the few opinion essay elements they
shared with students but did not model or support their inclusion
in student writing; did not teach students any strategies for
self-regulating the writing process; and did not model the writing
process using self-regulation, general, and genre-specific writing
strategies. We believe that all of these factors play a role in the
results found in this study.
One factor important to address here is whether or not these
same results in terms of improvement of elements would occur
without all of the stages and supports available in SRSD. While
more research is needed here, several studies indicate that for most
writers, and certainly for writers who are not making adequate
progress in writing, all of the stages and components are beneficial
(cf. Graham et al., 2013). These studies show that after the first
three stages of SRSD instruction (see Section 1.1; the genre ele-
ments have been deeply examined and found in model essays,
the mnemonics and importance of each part of the genre strategy
have been discussed and students indicate their understanding of
them, and the teacher has cognitively modeled the writing process
using the self-regulation and writing strategies), little progress
occurs in many students’ inclusion of the genre elements in their
own writing. Thus, Harris and her colleagues have argued that
‘‘P.E.E.ing (Post, Explain, Even Model) in the classroom’’ is not
enough for most students to deal with the complex factors
involved in writing. More meaningful progress is typically seen
in SRSD instruction only with the gradual release of control that
occurs across stages 5–6.
SRSD is a complex, multi-component instruction approach (see
Section 1.1) developed for use across complex academic learning,
not just for writing. Harris and her colleagues have traced the
development of SRSD elsewhere more completely than we can
here (cf. Harris, 1982; Harris & Graham, 1992, 1996, 2009; Harris
et al., 2009), but all of the components and characteristics of SRSD
instruction have been developed and continue to be fine-tuned
based on multiple theories that have produced extensive empirical
support for each characteristic and component.
Clearly, components analyses of both SRSD and PBPD for SRSD
are needed. Graham et al. (2013) identified five studies that com-
pared SRSD with explicit development of self-regulation (the usual
model) to SRSD with these components removed. They found that
the added value of teaching self-regulations procedures specifically
in the SRSD model was 0.48 standard deviations. No other compo-
nents analyses studies of SRD were found; additional research is
needed to determine what components of SRSD instruction are
responsible for differing gains in students’ writing performance,
Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
9
attitudes toward writing, writing self-efficacy, and so on. As Harris
and her colleagues have argued, however, their experience indi-
cates that each student brings a unique, complex set of personal
characteristics and abilities to learning to write, thus necessitating
the combination of the multiple components and characteristics of
SRSD instruction when teaching multiple students (cf. Harris &
Graham, 2009).
4.2. Lessons learned, limitations, future research and conclusions
In the current study, we successfully addressed many of the dif-
ficulties of doing school-based intervention research, but others
must be addressed in future research. We also learned several
important lessons to share with other researchers. First, we were
successful in revising SRSD instruction for Portuguese teachers
and students. We adapted the SRSD lessons used in this study to
the Portuguese educational and linguistic context. Adapting SRSD
material, especially mnemonics, required carefully attention, as
we have seen. Critical to our success was collaborating with a Por-
tuguese language arts’ teacher; he played a key role in the adapta-
tion of the original mnemonics and materials for SRSD to
Portuguese. This kind of collaboration will be critical to investigat-
ing evidence-based practices developed in one country to the cul-
tural and linguistic contexts of other countries. Similarly, our
research team included members who had done extensive research
on SRSD and PBPD for SRSD in the United States. This collaboration
helped ensure the integrity of the PBPD and SRSD approaches. In
order to test the generalizability of school-based interventions
such as SRSD outside of their original countries, it is very important
that such collaboration, or at least dialogue, be established.
Another important lesson we learned was to work to find ways
to increase the rate of parental consent and student assent in
school-based research. Of the 436 students who took the pre-test,
we received consent and assent for 380 students. In future studies,
researchers might achieve even higher levels of participation by
reaching out to parents to inform them further of the evidence base
for the intervention we were studying. This has the added benefit
of informing parents about the importance of school-based inter-
vention research.
One of the most important lessons we learned concerns
resources needed to conduct school-based research. School-based
research is complex and demanding, requiring numerous
resources, and thus, funding. It is important to engage in discussion
with national government and organizational leaders to establish
the importance of such research and develop adequate funding
for these efforts. Such funding is not readily available in many
countries. Because our resources were limited, we faced challenges
that resulted in limitations in our study.
In the present study, we lacked the funds to holistically score
students’ essays at each time point. Qualitative, holistic scoring is
a common procedure for scoring writing quality in intervention
studies (Graham & Perin, 2007). Trained raters read each paper
to obtain a general impression of overall writing quality at that
grade level, attending to features such as ideation, organization,
sentence structure, aptness of word choice, and grammar. Training
raters, achieving reliability, and scoring large numbers of essays is
expensive. Future research on the generalizability of PBPD for SRSD
should include assessment of effects on writing quality as well as
structural elements. In addition, future research should investigate
the effects of PBPD for SRSD on national writing exams. Because
criteria used in scoring the Portuguese national exam include eval-
uation of opinion essays’ structural elements (premise, reasons,
elaborations, and conclusions), PBPD for SRSD in opinion essay
writing could impact results on national (or state) writing exams
targeting opinion essay writing.
10 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx
Another limitation of this study is the lack of observations of
instruction in comparison classes, which we did not have the
resources to conduct. Although we had the teachers’ answers to
the Writing Activities Questionnaire, future research needs to
include observations of typical instruction to rather than rely
solely on teachers’ self-reports. In the present study, it appears
some teachers in the business as usual comparison group spent
less time directly teaching writing during the language arts blocks
than did the teachers in the experimental group. While comparing
school-based interventions to business as usual in schools is
important, future studies might seek to control time spent specif-
ically teaching writing. Research, however, indicates that spending
more time teaching many of the elements of writing comparison
condition teachers focused on (such as grammar and punctuation)
typically does not result in better compositions (Graham & Perin,
2007).
Further, we were not able to investigate an alternative, evi-
dence-based writing intervention in comparison to either SRSD
or typical instruction. Future research should examine the differing
effects of different approaches to writing instruction. Another pos-
sibility in future research into the generalizability of PBPD for SRSD
outside of the United States is to compare the effects of SRSD
instruction in different writing genres (e.g., one condition receives
SRSD for narrative writing and the other receives SRSD for opinion
essay writing; both groups are tested in both genres), as done by
Harris et al. (2012) with students in the United States. This is pos-
sible as research indicates little to no improvement in either of
these genres given instruction in the other genre. This kind of
design has the added advantage of allowing all teachers and stu-
dents involved in school-based research to participate in evi-
dence-based practices.
Similarly, we were not able to investigate effects of PBPD for
SRSD among teachers on variables such as knowledge, efficacy
for teaching writing, attitudes toward writing instruction, and so
on. In future research, it is important to procure funding to assess
multiple outcomes among teachers in addition to fidelity of
instruction (Harris et al., 2012). Further, in this study, as in
Harris et al. (2012), some students did not make as much progress
as their teachers thought they could have given more differenti-
ated instruction. Further research is needed to determine how best
to support teachers, during and after PBPD, in differentiating SRSD
instruction for students with varying writing abilities in their
classrooms. Establishing the importance of allocating more
instructional time for writing should also be addressed in future
research. Given the positive results obtained in the present study,
we are hopeful that in future Portuguese schools and their direc-
tors will allow allocation of more time to this kind of writing
instruction and research, allowing us to conduct studies further
examining our approach to professional development.
In conclusion, this is the first study of PBPD for SRSD in opinion
essay writing carried out in Portugal, and the first to demonstrate
that SRSD implemented by classroom teachers following PBPD can
be effective in teaching opinion essay writing to students outside
of the United States. Our experiences conducting this large scale
school-based intervention study resulted in important lessons to
share with future researchers. In addition, further research on scal-
ing-up and cross-cultural generalizability is needed and warranted
to explore the effects of PBPD for SRSD in other countries, cultures,
populations, and genres of writing.
References
Almeida, T. (2012). Desafios ao desenvolvimento profissional: Do trabalho cooperativo
ao nível da escola a um grupo sobre a escrita [Challenges to professional
development: From school cooperative work to a writing group work]. Doctoral
thesis not published. Lisboa: Faculdade de Psicologia e de Ciências da Educação
da Universidade de Lisboa.
Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
Almeida, T., & Simão, A. V. (2007). Concepções de professores sobre o processo de
composição escrita [Teachers’ conceptions about composition writing
processes]. In A. V. Simão, A. L. Silva, & I. Sá (Eds.), Auto-regulação da
aprendizagem [Self-regulation learning] (pp. 41–61). Lisboa: EDUCA.
Baker, S. K., Chard, D. J., Ketterlin-Geller, L. R., Apichatabutra, C., & Doabler, C.
(2009). Teaching writing to at-risk students: The quality of evidence for Self-
Regulated Strategy Development. Exceptional Children, 75, 303–318.
Ball, D. L., & Cohen, D. K. (1999). Developing practice, developing practitioners:
Toward a practice-based theory of professional education. In L. Darling-
Hammond & G. Sykes (Eds.), Teaching as a learning profession: Handbook for
policy and practice (pp. 3–31). San Francisco: Jossey-Bass.
Bangert-Drowns, R. (1993). The word processor as an instructional tool: A meta-
analysis of word processing in writing instruction. Review of Educational
Research, 63, 69–93.
Brindle, M. (2012). Examining relationships among teachers’ preparation, efficacy, and
writing practices. Unpublished dissertation. Vanderbilt University.
Carvalho, J. B., & Pimenta, J. (2005). Escrever para aprender, escrever para exprimir o
aprendido [Write to learn, write to express what was learned]. In B. Silva, & L.
Almeida (Orgs.), Actas do Congresso Galaico-Português de Psicopedagogia
[Proceedings of Psychopedagogy Galaico-Portuguese Congress] (pp. 1877–1885).
Braga: Centro de Estudos em Educação e Psicologia da Universidade do Minho.
Cook, B. G., Smith, G. J., & Tankersley, M. (2012). Evidence-based practices in
education. In K. Harris, S. Graham, T. Urdan, C. B. McCormick, G. M. Sinatra, & J.
Sweller (Eds.). APA educational psychology handbook: Theories, constructs, and
critical issues (Vol. 1, pp. 495–527). Washington, DC, US: American Psychological
Association.
De La Paz, S. (2005). Teaching historical reasoning and argumentative writing in
culturally and academically diverse middle school classrooms. Journal of
Educational Psychology, 97, 139–158.
De La Paz, S., & Graham, S. (2002). Explicitly teaching strategies, skills, and
knowledge: Writing instruction in middle school classrooms. Journal of
Educational Psychology, 94, 291–304.
Festas, I., Damião, H., & Martins, I. (2010, April). Estratégias de escrita de textos em
alunos universitários [Text writing strategies on university students]. In M.
Simões (Chair), Experiências de Ensino e Aprendizagem na Universidade: Retenção
de conhecimentos, contratos e estratégias de aprendizagem, reorganização
curricular, job shadowing, avaliação [Teaching and learning experiences at the
university: knowledge acquisition, learning contracts and strategies, curriculum
organization, job shadowing, and evaluation]. Symposium conducted at the
meeting Conference Learning and Teaching in Higher Education. Évora
(Portugal).
Fraenkel, J. R., & Wallen, N. E. (2003). How to design and evaluate research in
education (5ª ed.). New York: McGraw-Hill.
Gabinete de Avaliação Educacional [Office of Educational Evaluation] (2011). Provas
de aferição 2.° ciclo – Língua Portuguesa [2.nd cycle exams- Portuguese Language].
Lisboa: Ministério da Educação.
Gabinete de Avaliação Educacional [Office of Educational Evaluation] (2012). Provas
de aferição 1.° ciclo – Língua Portuguesa [2.nd cycle exams- Portuguese Language].
Lisboa: Ministério da Educação.
Gabinete de Avaliação Educacional [Office of Educational Evaluation] (2013). Provas
finaise exames nacionais – Ensino básico e secundário [Final probes and national
exams – Elementary and secondary levels]. <http://www.gave.min-edu.pt/np3/
np3/451.html>.
Gabinete de Informação e Avaliação do Sistema Educativo [Office of Information
and of Evaluation of Educational System] (2005). Estatísticas da Educação 04/05
[04/05 Educational statistics]. Lisboa: Ministério da Educação.
Gilbert, J., & Graham, S. (2010). Teaching writing to elementary students in grades
4–6: A national survey. Elementary School Journal, 110, 494–518.
Graham, S. (1999). Handwriting and spelling instruction for students with learning
disabilities: A review. Learning Disability Quarterly, 22, 79–98.
Graham, S. (2006a). Writing. In P. Alexander & P. Winne (Eds.), Handbook of
Educational Psychology (pp. 457–478). Mahwah, NJ: LEA.
Graham, S., & Harris, K.R. (in press). Common Core State Standards and writing: An
introduction. Elementary School Journal.
Graham, S., Harris, K. R., & McKeown, D. (2013). The writing of students with LD and
a meta-analysis of SRSD writing intervention studies: Redux. In L. Swanson, K.
R. Harris, & S. Graham (Eds.), Handbook of Learning Disabilities (2nd ed.,
pp. 405–438). NY: Guilford Press.
Graham, S., & Harris, K. (2003). Students with learning disabilities and the process
of writing: A meta-analysis of RSD studies. In L. Swanson, K. Harris, & S. Graham
(Eds.), Handbook of research on learning disabilities (pp. 323–344). New York, NY:
Guilford Press.
Graham, S. (2006b). Strategy instruction and the teaching of writing: A meta-
analysis. In C. A. MacArthur, S. Graham, & J. Fritzgerald (Eds.), Handbook of
writing research (pp. 187–207). New York, NY: Guilford Press.
Graham, S., McKeown, D., Kiuhara, S., & Harris, K. (2012). A meta-analysis of writing
instruction for students in the elementary grades. Journal of Educational
Psychology, 104, 879–896. http://dx.doi.org/10.1037/a0029185.
Graham, S., & Perin, D. (2007). A meta-analysis of writing instruction for adolescent
students. Journal of Educational Psychology, 99, 445–476.
Gresham, F. M., Gansle, K. A., Noell, G. H., Cohen, S., & Rosenblum, S. (1993).
Treatment integrity of school-based behavioral intervention studies: 1980–
1990. School Psychology Review, 22, 254–273.
Grossman, P., & McDonald, M. (2008). Back to the future: Directions for research in
teaching and teacher education. American Educational Research Journal, 45,
184–205.
 I. Festas et al. / Contemporary Educational Psychology xxx (2014) xxx–xxx
Harris, K. (1982). Cognitive-behavior modification: Application with exceptional
students. Focus on Exceptional Children, 15, 1–16.
Harris, K., Lane, K., Driscoll, S., Graham, S., Wilson, K., Sandmel, K., et al. (2012). Tier
1, teacher-implemented self-regulated strategy development for students with
and without behavioral challenges. Elementary School Journal, 113, 160–191.
http://dx.doi.org/10.1086/667403.
Harris, K., & Graham, S. (1996). Making the writing process work: Strategies for
composition and self-regulation (2nd ed.). Cambridge, MA: Brookline Books.
Harris, K., & Graham, S. (2009). Self-regulated strategy development in writing:
Premises, evolution, and the future. British Journal of Educational Psychology, 6,
113–135.
Harris, K., Graham, S., Brindle, M., & Sandmel, K. (2009). Metacognition and
children’s writing. In D. Hacker, J. Dunlosky, & A. Graesser (Eds.), Handbook of
metacognition in education (pp. 131–153). Mahwah, NJ: Erlbaum.
Harris, K., Graham, S., Mason, L., & Friedlander, B. (2008). Powerful writing strategies
for all students. Baltimore: Paul H. Brookes Publishing Co.
Harris, K., & Graham, S. (1992). Self-regulated strategy development: A part of the
writing process. In M. Pressley, K. Harris, & J. Guthrie (Eds.), Promoting academic
competence and literacy in school (pp. 277–309). New York: Academic Press.
Harris, K., Lane, K., Graham, S., Driscoll, S., Sandmel, K., Brindle, M., et al. (2012).
Practice-based professional development for self-regulated strategies
development in writing: A randomized controlled study. Journal of Teacher
Education, 63(2), 103–119. http://dx.doi.org/10.117/0022487111429005.
Hulleman, C. S., & Cordray, D. (2009). Moving from the lab to the field: The role of
fidelity and achieved relative intervention strength. Journal of Research on
Intervention Effectiveness, 2(1), 88–110.
Institute of Education Sciences (2012). IES practice guide: Teaching elementary school
students to be effective writers. NCEE 2012-4058, <http://ies.ed.gov/ncee/wwc/
pdf/practice_guides/writing_pg_062612.pdf>.
Mason, L. H., Kubina, R., & Taft, R. (2009). Developing quick writing skills of middle
school students with disabilities. Journal of Special Education, 44, 205–220.
Ministério da Educação e Ciência [Ministry of Education and Science] (2009).
Programa de Português do ensino básico [Portuguese language arts curriculum for
elementary and middle levels]. <http://www.dgidc.min-edu.pt/index.php?s=
noticias&noticia=396>.
Ministério da Educação e Ciência [Ministry of Education and Science] (2012). Metas
curriculares do ensino básico [Standards for elementary and middle grade levels].
<http://www.dgidc.min-edu.pt/index.php?s=noticias&noticia=396>.
Please cite this article in press as: Festas, I., et al. Professional development in Self-Regulated Strategy Development: Effects on the writing performance of
eighth grade Portuguese students. Contemporary Educational Psychology (2014), http://dx.doi.org/10.1016/j.cedpsych.2014.05.004
11
Morphy, P., & Graham, S. (2012). Word processing programs and weaker writers/
readers: A meta-analysis of research findings. Reading and Writing: An
Interdisciplinary Journal, 25, 641–678.
National Center for Education Statistics (2012). The nation’s report card: Writing
2011. Washington, D.C.: Institute of Educational Sciences, U.S. Department of
Education.
National Center on Intensive Intervention (2013). Academic interventions tool
chart. <http://www.intensiveintervention.org/chart/instructional-intervention-
tools>.
Page, E., & Petersen, N. (1995). The computer moves into essay grading: Updating
the ancient test. Phi Delta Kappan, 76, 561–566.
Pardal, E., & Festas, I. (2011, September). A competência de escrita em Manuais do
10.° ano de Português [Writing competence on 10th level textebooks]. In I.
Festas (Chair), Composição de textos [Text composition]. Symposium conducted at
the meeting XI Congreso Internacional Galego-Portugés de Psicopedagoxia, Coruna
(Spain).
Pereira, L. A. (2001). A formação de professores para o ensino da escrita [Teacher
training to the teaching of writing]. Cadernos de Formação de Professores
[Notebooks of Teacher Training], 2, 35–49.
Rebelo, J., Sousa, C., Inácio, M., Vaz, J., Festas, I., & Oliveira, A. (2013). O Programa de
Escrita SRSD e a sua adaptação para um estudo em escolas de Coimbra. Revista
Portuguesa de Pedagogia, 47(1), 31–51.
Robinson, D. H., Levin, J. R., Schraw, G., Patall, E. A., & Hunt, E. B. (2013). On going
(way) beyond one’s data: A proposal to restrict recommendations for practice in
primary educational research journals. Educational Psychology Review, 25(2),
291–302. http://dx.doi.org/10.1007/s10648-013-9223-5.
Rogers, L. A., & Graham, S. (2008). A meta-analysis of single subject design writing
intervention research. Journal of Educational Psychology, 100, 879–906. http://
dx.doi.org/10.1037/0022-0663.100.4.879.
RStudio (2013). RStudio: Integrated development environment for R (Version 0.98.490)
[Computer software]. Boston, MA. Retrieved December 22, from http://
www.rstudio.org/.
Scardamalia, M., Bereiter, C., & Goleman, H. (1982). The role of production factors in
writing ability. In M. Nystrand (Ed.), What writers know: The language, process,
and structure of written discourse (pp. 175–210). San Diego, CA: Academic Press.
Wong, B. Y. L., Hoskyn, M., Jay, D., Ellis, P., & Watson, K. (2008). The comparative
efficacy of two approaches to teaching sixth graders opinion essay writing.
Contemporary Educational Psychology, 33, 757–784.