Effects of teacher-implemented explicit

writing instruction on the writing

self-efficacy and writing performance of
5th grade students

Érick Falardeau1, Frédéric Guay1, Pascal Dubois2, & Daisy Pelletier1

1
Université Laval | Canada
2
McGill University | Canada

Abstract: Meta-analyses indicate that explicit writing instruction (EWI) is an effective method
for improving student writing self-efficacy and writing performance. EWI relies on explicit
instruction of writing strategies through modeling, scaffolding and self-regulation. Most
EWI-based interventions have been conducted by researchers, generally with subgroups of
students or on a one-on-one basis, and very few have been conducted in other languages
than English. Our quasi-experimental study aims to address these limits by testing EWI’s
effects when teachers themselves intervene using peer feedback during the writing of
opinion letters. We used practice-based professional development to teach teachers how to
use EWI, and compared two experimental conditions (EWI with and without peer feedback)
to a control group (Business as Usual). A total of 483 French-speaking 5th grade students
participated in the study. Results from repeated measure analyses showed that, with or
without peer feedback, the EWI intervention produced better writing performance and
higher self-efficacy compared to the control group. We discuss the role of EWI for writing
performance and self-efficacy.

Keywords: Explicit writing instruction, writing performance, self-efficacy, argumentative

writing, practice-based professional development

Falardeau, E., Guay, F., Dubois, P., & Pelletier, D. (2024). Effects of teacher-implemented
explicit writing instruction on the writing self-efficacy and writing performance of 5th grade
students. Journal of Writing Research, 16(1), 1-38. DOI: https://doi.org/10.17239/jowr-
2024.16.01.01
Contact: Érick Falardeau, Faculté des sciences de l’éducation, Université Laval, Pavillon
Jeanne-Lapointe, 2320, rue des Bibliothèques, Québec G1V 0A6 | Canada –
[email protected].
Copyright: This article is published under Creative Commons Attribution-Noncommercial-
No Derivative Works 3.0 Unported license.
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 2

In the United States, 75% of 8th and 12th graders do not meet, or only partially meet,
national standards in writing (National Center for Education Statistics, 2012). In the
Canadian province of Quebec, French-speaking students likewise experience
writing difficulties that become more pronounced as they progress through grade
levels. Between 55% and 65% of secondary school students struggle with the
provincial writing exam administered in 8th grade, which focuses on primary school
curriculum content (1st to 6th grades; Ministère de l’Éducation, du Loisir et du
Sport, 2013).
Struggling writers see writing solely as a way of producing content. They plan
their text as their ideas arise and make only surface revisions (e.g., grammar and
spelling). They also spend less time planning, writing and revising their texts than
proficient writers. Struggling writers perceive writing as a linear process, limited to
producing a text based on their knowledge of the topic (Gillespie & Graham, 2014).
Explicit writing instruction (EWI) has been shown to be an effective way to improve
students’ writing skills. This method relies on making writing strategies visible or
explicit to struggling students, generally through modeling, scaffolding and self-
regulation of learning strategies. Below, we provide details on EWI, an approach
that mobilizes students’ self-regulation and self-efficacy.

1. The Writing Process and the Role of Self-Regulation in Learning to Write

Writing is a difficult task as writers must constantly keep their audience in mind
without interacting directly with readers (Graham & Harris, 2018). Some types of text
accentuate this difficulty. One such is the argumentative text, the focus of this study.
Writers are unable to fully anticipate how readers will react to their arguments. By
constructing counterarguments, however, the writer must enter in a dialogue with
readers, anticipating their responses and objections (Prata et al., 2019).
Learning to write also involves self-regulation, by which writers can control their
writing activity. Self-regulation is a cognitive or metacognitive process that aims to
ensure the control and adjustment of cognitive, affective, and social activities
contributing to the transformation of the learners’ knowledge and skills into a
coherent text (Mottier-Lopez, 2012). Self-regulation involves identifying goals and
steps for achieving them and selecting strategies for success (Hattie and Timperley,
2007). From a Vygotskian perspective, it is a process of internalizing skills and
knowledge through interactions with others (Cartier & Mottier-Lopez, 2017).

2. The Role of Self-Efficacy in Writing Development

Self-efficacy plays a key role in students’ writing self-regulation (MacArthur &
Graham, 2016; Zimmerman & Risemberg, 1997; De Smedt et al., 2020). Self-efficacy
refers to an individual’s beliefs about their abilities to achieve goals (Bandura, 1977).
Students with high writing self-efficacy feel in control during the writing process.
3 | JOURNAL OF WRITING RESEARCH

They believe in their abilities to produce a text that meets grade-level expectations
(MacArthur & Graham, 2016) and are more successful in writing (Pajares, 2003). In
contrast, students who accumulate failures in writing and who master few learning
strategies do not feel confident in their ability to produce texts that meet
expectations (Bruning & Kauffman, 2016). The link between self-efficacy and
students’ writing performance is worth studying.

3. Explicit Instruction to Enhance Writing Self-Efficacy

Because self-efficacy is important for developing writing skills (Graham et al., 2017),
it is valuable to uncover strategies for fostering students’ perceptions of self-
efficacy. Recent meta-analyses suggest that explicit writing instruction (EWI) is one
of the instructional methods with the strongest effect sizes on self-efficacy,
regardless of students’ competence, gender or school level (Rogers & Graham, 2008;
Harris et al., 2012; Graham et al., 2013, 2016; Gillespie & Graham, 2014; Graham &
Harris, 2017; De Smedt & Van Keer, 2018b). EWI teaches writing strategies for
planning (generating and selecting ideas, creating a plan), drafting (choosing
vocabulary, constructing sentences), revising (ensuring the progression of ideas,
checking the accuracy of vocabulary and spelling) and correcting. Efficient EWI
models include five practices: 1) showing how the strategies work, 2) modeling, 3)
guiding students’ practices by focusing on scaffolding and promoting student
collaboration for better verbalization of strategies, 4) fostering independent
practice and 5) transferring the strategies to writing tasks.
There are different models of EWI. One of the most efficient is Self-Regulated
Strategy Development (SRSD), which prioritizes activation of background
knowledge, discussion, modelling, memorization, support and independent
practicing of the writing strategies presented above and incorporates four self-
regulatory strategies (i.e., goal setting, self-monitoring, self-talk and self-
reinforcement; Graham & Harris, 2005) into EWI. Another important EWI model
similar to SRSD is that of De Smedt and Van Keer (2018b), who kept the focus on
self-regulatory strategies. However, instead of using SRSD’s mnemonic acronyms
(such as POW or TREEi) for memorization, they use strategy cards that explain to
students how to use writing and self-regulatory strategies. Teachers then support
students in learning these strategies through guided practice. De Smedt and Van
Keer emphasize questioning the students on their use of a strategy in order to
develop their self-regulation and adjust teaching as needed.

4. Peer Feedback in Explicit Instruction

Peer feedback appears to enhance the effect of explicit instruction, as explicit
instruction coupled with peer support provides students with more opportunities
to discuss strategies, assess their mastery of them and build their self-regulation
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 4

skills. Students verbalize their own approach to writing by pointing out strengths
and weaknesses in their peers’ writing using their own success criteria (Panadero et
al., 2018). Practicing argumentation through dialogue, therefore, promotes
internalization of argumentative skills. Creating groups of two or three students to
defend their opinions and listen to those of their peers supports their
understanding of the dialogical nature of argumentative discourse.
In our project, we used peer feedback during all steps of the writing process to
develop students’ self-regulation. Students authored their own texts rather than
writing in groups, but used the targeted strategies by co-planning and by revising
peers’ texts. Discussion of others’ texts promotes the internalization of strategies
that students will use in their own writing. Teachers play an important role in this
feedback loop (Zimmerman & Risemberg, 1997), as they guide students working in
teams toward effective forms of feedback that focus on self-regulation (Hattie &
Timperley, 2017).
Peer feedback is another effective practice to improve students’ revision skills
(Bruning & Kauffman, 2016; Graham et al., 2015; MacArthur, 2016; Wigglesworth &
Storch, 2012). Specifically, students learning how to give feedback benefit from
better comprehension of learning strategies (Boscolo & Ascorti, 2004; Crinon, 2012;
Lundstrom & Baker, 2009). This positive effect has been observed even among
struggling learners (Crinon & Marin, 2010). Nevertheless, peer feedback’s effect on
student writing performance is not always significant (De Smedt & Van Keer, 2018a).

5. Explicit Instruction, Teacher Implementation and Practice-Based

Professional Development
Studies on teacher implementation of EWI remain rare, especially in primary
schools. Finlayson and McCrudden (2020) identified only 13 publications in which
primary teachers implemented writing interventions, mainly SRSD. Documenting
this scarcity, Palermo and Thomson (2018) report only 3 studies on SRSD models
implemented by middle school teachers. Moreover, most of the research on SRSD
that includes students with emotional and behavioral disorders (9 out of 11)
involves the researcher (McKeown et al., 2014) playing the role of the teacher.
Teacher implementation of EWI and its strategies must be tested in classroom
settings if we want to achieve long-lasting effects for student learning (Finlayson &
McCrudden, 2020).
Teachers’ interventions usually have strong effects on writing quality, even
though they may experience difficulties in implementing EWI. EWI’s effectiveness
often relies on teachers’ training and treatment fidelity. Of the studies conducted
with teachers, one compared teacher-implemented SRSD to Business as Usual
(BAU) instruction in 3rd, 4th and 5th grades (McKeown et al., 2016) and found a
limited effect on overall text quality (Cohen’s d = .15) and structure (Cohen’s d =
.24), but the treatment fidelity was low in this study. Collins et al. (2021) studied
5 | JOURNAL OF WRITING RESEARCH

teaching of expository essays in social sciences with 3rd grade teachers. They
compared an intervention using SRSD to BAU instruction and found an effect size
of 1.07 for elements of the targeted genre and .72 for holistic quality of the texts.
Palermo and Thomson (2018)’s teacher implementation study compared three
conditions: 1) SRSD + feedback via an automated system, 2) BAU + feedback via an
automated system and 3) BAU. Within the SRSD group they found an effect size of
1.18 for writing quality of the argumentative essay and one of .97 for essay elements
compared to both BAU groups. De Smedt and Van Keer (2018a) compared the
effectiveness of four teachers’ interventions in 5th and 6th grade: 1) EWI + peer
assistance (PA), 2) EWI alone, 3) BAU + PA and 4) BAU alone. EWI and PA groups
outperformed the BAU group (effect sizes of between 1.94 and 2.14 for EWI and 1.43
for the PA writing program).
Effect sizes in these studies, although strong, are generally smaller than the ones
(Cohen’s d = 1.59) reported in a meta-analysis by Graham and Harris (2017).
Nevertheless, it should be noted that most of the studies reported in this meta-
analysis are based on interventions in which the researcher intervenes directly with
a student or with subgroups of students. Rarely, the researcher intervenes with all
the students in the class. Most EWI studies using teacher-provided instruction
address this implementation issue by planning a professional development
program to coach teachers in intervening with all their students. This teacher
support before and during the intervention is part of practice-based professional
development (PBPD). Several studies showed positive results when the teachers
themselves intervened with their students in keeping with the following PBPD
principles (Festas et al., 2015; McKeown et al., 2019):

(a) collective participation of teachers within the same school with similar
needs; (b) basing professional development around the characteristics,
strengths, and needs of the students […]; (c) attention to content knowledge
needs of teachers, including pedagogical content knowledge; (d)
opportunities for active learning and practice of the new methods being
learned […]; (e) use of the materials […] that are identical to those to be used
in the classroom; and (f) feedback on performance while learning and before
using these methods in the classroom (Harris et al., 2012, p. 105).

But following PBPD’s principles does present challenges for the classroom context.
In a single case study on SRSD implementation, McKeown et al. (2019) highlighted
the difficulty of training teachers to teach SRSD to the whole class, because they do
not have time to ensure that all students have mastered the target strategies. Other
implementation challenges include interrupted lessons, uneven teacher
engagement and classroom schedule disruptions.
To explain the variation of effect sizes when teachers teach writing strategies,
McKeown et al. (2019) pointed out that differences in writing quality are difficult to
observe over a short period of time. Overall, teaching teachers to use explicit
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 6

instruction poses challenges that previous intervention studies have not always
addressed. Nonetheless, because teaching generally takes place in the classroom
with the whole class and without researcher input, effective methods should be
tested under conditions that are as similar as possible to teaching practice.

6. Our Theoretical Framework for EWI

To put EWI into practice in the Quebec culture, we adapted Graham and Harris’s
(2005) model for explicit SRSD instruction. Our interactionist approach (see De
Smedt and Van Keer, 2018a and b) relies more on students’ verbalization of the
strategy used than on memorization of the strategies as in SRSD. Much like De
Smedt and Van Keer, we used visual supports to present the procedure for using
strategies. Our model had six iterative phases: 1) showing how strategies work, 2)
modeling, 3) guiding students’ practices by focusing on scaffolding and promoting
student collaboration for better verbalization of strategies, 4) fostering
independent practice, 5) transferring the strategies to students’ writing tasks and 6)
questioning students about their use of the targeted strategy and their self-
regulation activity. Metacognitive questioning by and feedback from teachers in
these different phases are two essential practices to ensure that the intervention is
adjusted to fit students’ capacities. To this end, questioning should focus more on
the process (how students mobilize the strategy and their level of control over it)
than on the product expected. Information obtained through these questions will
allow teachers to identify subgroups of students who are experiencing difficulties
and then to support them (for a more detailed description, see Falardeau, 2021).

7. The Present Study

Few EWI studies that tested an intervention’s effectiveness have been conducted
with teachers implementing it. More research of this type is needed, as it is more
ecologically valid. It is therefore important to study teacher-led EWI intervention,
as well as to do so in the Quebec context where it has not been evaluated
rigorously. Our study is in line with other studies (e.g., Festas et al., 2015) that aimed
to validate, in different cultural contexts, the effectiveness of teacher-led EWI
interventions. Writing instruction practices that develop self-efficacy in classroom
contexts are likewise little studied (Bruning & Kauffman, 2016).
To adapt EWI for use in a French-speaking culture (Quebec, in Canada), we
worked with teachers and pedagogical advisors to develop an intervention tailored
to Quebec’s school curriculum. We also included peer feedback to see how 5th
grade students (ages 10–11) could benefit from critically reading their peers’
argumentative texts to improve their own writing performance and self-efficacy.
Given the lack of clear evidence regarding the effects of peer feedback on writing
7 | JOURNAL OF WRITING RESEARCH

performance and self-efficacy (De Smedt & Van Kleer, 2018a) in a teacher-led
intervention, further research combining peer feedback and EWI is needed.
In our study, we have focused on the opinion letter, a type of writing that in
Quebec is the subject of a provincial test at the end of 6th grade and for which there
are few teaching materials. The reason why argumentative writing is so important
in elementary school is that it is a critical skill for becoming an engaged member of
society: the student who can support a thesis with arguments backed up by
evidence will be better equipped to advance and support their point of view.
The purpose of this research, then, is to compare the writing performance and
self-efficacy of 5th grade students writing argumentative texts before and after an
intervention. The study design included three conditions. The first experimental
group combined peer feedback with EWI (EWI+PF) while the second experimental
group used only EWI without peer feedback. Teachers in the control group
conducted business-as-usual interventions (BAU). This allowed us to isolate the
impact of EWI with and without peer feedback.
Our hypotheses were as follows:

• Hypothesis 1: The writing performance and writing self-efficacy of students who

receive explicit writing strategy instructions (experimental groups 1 and 2) will
improve more than those of students in the control group (Control).

• Hypothesis 2: The writing performance and writing self-efficacy of students who

receive explicit writing instruction and practiced peer feedback (EWI+PF) will
improve more than those of students who were not trained in this intervention
(EWI only and Control).

To ensure that the observed differences between the groups could be attributed to
teachers’ implementation of writing instructional practices, we controlled for the
gender and age of the students. Gender may influence writing performance and
self-efficacy (Graham et al., 2017). Controlling for age is also important because
students’ writing skills improve unevenly with age in elementary school (Fayol,
2016). Age is also related to self-efficacy, although the complexity of the writing task
plays a greater role for this outcome (Bruning & Kauffman, 2016).

8. Method

8.1 Participants and Procedure

In our quasi-experimental study, the first group (EWI+PF) consisted of students
taught by 8 teachers from 7 schools; the second group (EWI) consisted of students
taught by 8 teachers from 3 schools; the Control group consisted of students taught
by 9 teachers from 5 schools. Data was collected at two points during the 2020–2021
school year: a pre-test in September (Time 1; T1) and a post-test (Time 2; T2) in May
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 8

or June 2021, after the intervention (teachers ran the intervention between April and
June 2021 at a time of their choosing, based on their planning for the year). All 25
teachers in the three groups completed the research project. Students wrote an
opinion letter at the post-test, one week after their teachers finished teaching the
opinion letter.
Regarding gender, 236 girls and 247 boys participated in this study,
proportionally distributed among the three groups. Appendix A presents the
characteristics of the students in the three groups. Most students were aged
between 9 and 10 (88.4%). Overall, their characteristics were similar, except that the
EWI+PF group was slightly younger than the two other groups. In all three groups,
a similar proportion of students had French as their first language (88.3% in EWI+PF;
88.8% in EWI; 85.4% in the Control group). The difference of 2.8% and 3.4% between
the Control group and the experimental groups is not significant. To measure the
socio-economic level of the schools, we used the School Socio-Economic Index
(SSEI) developed by the Quebec Ministry of Education. This index uses a scale of 1
to 10, with 1 representing the most advantaged and 10 the most disadvantaged
schools. Both experimental groups had a higher average index than the Control
group (an average SSEI of 4.67 for the Control group versus 6.33 for the EWI group
and 7.25 for EWI+PF). The Control group was therefore socio-economically more
advantaged than the two experimental groups. However, we did not take this
covariate into account in our analysis as we did not have enough schools to do so
(n=14).

8.2 Research Design

Participating schools were randomly assigned to one of the three groups. To avoid
any threat to the internal validity of the study, teachers were not randomly assigned
to the three groups. If teachers had been assigned randomly, teachers working in
the same school could have been be assigned to different groups and thus could
have discussed the content material learned in each group, potentially endangering
the validity of the study.

8.2.1 Experimental Group with Peer Feedback (EWI+PF)

In the first experimental group, teachers implemented explicit instruction with peer
feedback in each guided practice activity. Our research team, in collaboration with
5th grade teachers and elementary school consultants, developed an explicit
instructional sequence following our EWI model (presented in the theoretical
framework). The sequence contained 15 lessons of approximately 90 minutes each,
delivered over a period of 6 to 8 weeks. According to Palermo and Thomson (2018),
EWI generates positive effects after 8 to 12 lessons of 20 to 45 minutes. We chose to
teach five writing strategies related to the opinion letter: 1) analysis of the reader’s
characteristics in order to build convincing arguments and directly appeal to the
9 | JOURNAL OF WRITING RESEARCH

reader; 2) argumentative consistency between the theme of the letter, the opinion
defended, the arguments and the justifications; 3) the structure of the opinion letter
and all its required elements; 4) the use of a rich and varied vocabulary and 5) the
use of exclamatory, interrogative and imperative sentences to increase the reader’s
interest. Each of these five strategies was taught in the specific context of opinion
letters. For example, in the modeling phase, teachers showed students how they
analyzed their reader’s characteristics to plan their letter. Students practiced each
of the five strategies at least three times under different conditions of increasing
complexity: once by analyzing letters written by unknown students and modified
by the researchers to highlight strengths or weaknesses; a second time to improve
the letter the student had written in the pre-test; and a third time after the student
had written a new letter during the intervention. Students wrote their own opinion
letters and learned to plan and revise them with the help of their peers. During each
guided practice, students were divided into teams of three to discuss their use of
the strategy and to give and receive feedback. The intervention followed this
structure:

Lesson # Content Lesson # Content

1 Analysis of the reader’s 9 Feedback and analysis of the
characteristics reader’s characteristics
2 Feedback and analysis of the 10 Feedback and use of a rich
reader’s characteristics and varied vocabulary
3 Consistency of argumentation 11 Analysis of the reader’s
characteristics in a new letter
4 Use of exclamatory, 12 Structure of a new opinion
interrogative, and imperative letter
sentences
5 Structure of the opinion letter 13 Feedback and consistency of
argumentation in a new letter
6 Consistency of argumentation 14 Revision of the vocabulary in a
and planning of a new letter new letter
7 Use of a rich and varied 15 Revision of varied sentences
vocabulary in a new letter
8 Feedback and structure of the
opinion letter

Each lesson featured a script written by our research team, accompanied by a video
explaining to teachers the rationale behind how the session was organized.ii We
also made anchor charts based on those proposed by Calkins (Martinelli & Mraz,
2015/2017) to make the process of using each strategy explicit to students. To
develop their self-regulation, students were asked to verbalize with their peers how
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 10

they used the strategy. During guided practice, the teacher met with each student
individually to help them identify learning goals. As students strengthened their
mastery of the strategies and self-regulation, the teacher lessened scaffolding (De
Smedt & Van Keer, 2018a).
The instructional sequence also included lessons dedicated to the explicit
teaching of peer feedback. These lessons were always linked to a strategy to teach
students how to give feedback on specific content. Students were provided with an
anchor chart outlining how to give and receive feedback: 1) give positive feedback
on the text; 2) give feedback on a specific aspect of the text that needs to be
improved; and 3) provide concrete examples or solutions. The teacher described
the feedback strategy, modeled it and showed the whole class how to do it by
bringing a team of students into the center of the class and guiding them in giving
feedback. During the guided practice, the teacher referred to the anchor chart to
remind the students of how to give feedback. This reference to the chart was
intended to develop students’ self-regulation by encouraging them to reflect on
what they are doing as they give feedback.
Following the principles of practice-based professional development (PBPD;
McKeown et al., 2014), which prescribes two days of teacher training prior to
classroom intervention, we first met with teachers in both experimental groups for
half a day to introduce them the EWI model and the research process. We then
provided them with online training modules about the explicit teaching model we
used, accompanied by a theoretical article (Falardeau, 2021), and the theoretical
foundations for the use of peer feedback. We also asked them to read the first three
lesson scenarios, accompanied by explanatory videos (seven to eight hours of
reading and listening). Once this step was completed, teams of teachers from each
school met for half a day. This meeting with the principal investigator ensured that
there was a shared understanding by providing an opportunity to clarify all
components of the research.
Once a week, teachers in the experimental groups filmed one of their lessons.
The video was viewed by the researcher, who used an observation grid listing all
teaching practices related to our explicit teaching model. Each item was to be rated
on a 5-point Likert scale based on adherence to the lesson script, where 1 = not at
all and 5 = absolutely, with space to justify ratings based on observed interventions.
This video analysis was followed by a weekly 30-minute virtual meeting in which the
researcher encouraged good practices that adhered to the script and reviewed
items that needed improvement. After a few meetings, the eight teachers of the
EWI+PF group taught all components of the lessons.

8.2.2 Experimental Group with Individual Writing (EWI)

For the second experimental group, the explicit teaching approach and the support
offered to the teachers were identical to those of the first group, except that the
11 | JOURNAL OF WRITING RESEARCH

students never worked in collaboration with their peers. All guided practice
activities were done individually with teacher support.

8.2.3 Control Group

The teachers in the Control group taught the opinion letter in their usual way. When
they finished their teaching, we met with each teacher and asked them to list the
content and the practices they used. Most used Calkins’ writing workshops (e.g.,
Calkins & Boland Hohne, 2013/2019), selecting various activities for mini-lessons.
Some conducted sharing activities on creating arguments or had students work in
teams to strengthen their arguments. Others taught only the structure of the
opinion letter before the post-test. All of them had received the list of the five
strategies we chose to teach the opinion letter (1) analysis of the reader’s
characteristics; 2) consistency of the argumentation; 3) the structure of the opinion
letter and all the elements expected; 4) the use of a rich and varied vocabulary and
5) the use of exclamatory, interrogative and imperative sentences), but were not
given any indication of how these strategies worked or how to teach them. We gave
them this list to minimize the dropout rate, which is usually higher in a Control
group than in experimental groups (Guay et al., 2020). They did not have any access
to the training or our teaching material during their teaching of the opinion letter.
We gave them the material after the completion of the post-test.

8.3 Measures

8.3.1 Perceived Self-Efficacy in Writing

Items used to measure self-efficacy regarding writing skills were drawn from
Bruning et al. (2013). We used two of the three subcategories of their self-efficacy
scale (ideation and regulation) without alteration. We adapted the convention
category to the content we taught, because our intervention did not address
spelling or grammar. Of the 14 items added, 2 focused on reader awareness (e.g., I
can think about my reader, his/her ideas, his/her age, to convince him/her of my
opinion), 2 focused on consistency (e.g., I can find convincing reasons [arguments]
to support my opinion), 3 focused on structure (e.g., I can write an introduction
announcing my opinion and my reasons [arguments]), 3 focused on the use of
imperative, interrogative, or exclamatory sentences (e.g., I can use interrogative
sentences [questions] in my text to make my reader think), and 4 focused on peer
feedback (I can offer another student clear comments and examples to help
him/her improve his/her text). We used a Likert scale ranging from 0 (Not confident
at all) to 100 (Absolutely confident). The alpha coefficient for this measure was .92
(see Appendix B for the complete scale, translated from French).
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 12

8.3.2 Writing Measures

We measured students’ writing performance at T1 and T2. Students in all three
groups completed the same pre-test in September, choosing one of six prompts
that addressed themes rooted in their culture and environment (e.g., Your school
wants to make uniforms mandatory. Do you agree? Write to your principal to
convince him or her of your opinion or Do you think it’s a good idea to have a cell
phone at school? Write to your principal and convince him or her of your opinion).
Because the need for autonomy is a key factor in writing engagement (Guay et al.,
2020), we wanted to enhance their motivation by offering them a wide choice of
prompted writing. The writing task required no reading materials, so that the
students’ reading skills would not interfere with their writing skills. All the
instructions had the same reader, the school’s principal. A panel of researchers and
5th grade teachers validated the equivalence of difficulty levels between the six pre-
test prompts and the six post-test ones. All presented the same level of difficulty,
due to the unique identity of the intended reader and the anchoring of all themes
in the students’ culture. Students were given no indication of length or writing time
and no support was allowed.
The post-tests were conducted in May and June, one week after each teacher
had finished teaching the opinion letter. The conditions of the test were identical
to those of the pre-test, and the same reader was kept. Only the prompts were
changed (e.g., Many experts say that young people should use electronic devices
less at home. Do you agree? Write to your principal to convince him or her of your
opinion or Spring break could be two weeks instead of one. However, summer
vacation would be shortened by one week. Which would you prefer? Write to your
principal to convince him or her of your opinion).
Research assistants rated the students’ pre-test and post-test letters in all three
groups to ensure that the criteria were applied consistently. While teachers could
use the pre-test letters to enhance the structure and content of students’ letters,
neither teachers nor students saw the results of our research assistants’ (RAs’)
rating.
Bouwer et al. (2024) argue that the definition of writing adopted to assess
students’ capacities must be based on “the overall aim of the writing intervention”
(p. 202). Accordingly, the score given to the letters was based on the strategies
taught:

1. Reader awareness: analyzing the reader’s characteristics and directly speaking

to the reader in the letter (10 pts/75)

2. Consistency of the argumentation (divided into two criteria):

a. Progression of information and noncontradiction (10 pts/75)

b. Relevance and development of ideas (15 pts/75)

3. Structure and elements of the letter (10 pts/75):

13 | JOURNAL OF WRITING RESEARCH

a. Introduction, development, and conclusion

b. Progression of the information and noncontradiction

4. Use of rich and varied vocabulary (20 pts/75):

a. Vocabulary richness (adding points)

b. Vocabulary error, distracting repetition, vagueness (subtracting points)

5. Use of various types of sentences to convince the reader: interrogative,

exclamative or imperative (10 pts/75)

Criteria were weighted according to the points awarded by the Quebec Ministry of
Education for the opinion letter portion of the 6th grade provincial test. The
weighting therefore reflects the relative importance given in Quebec to the
different components of writing. Using multiple criteria for quality of
argumentation enabled us to examine students’ progression for each strategy,
rather than simply using a general score for content and another for the presence
of expected elements (structure). The weighted criteria were added to calculate the
overall score. This rating procedure is innovative compared to the general
performance score we observed in most studies of EWI studies, and enabled a more
refined analysis of student progress. For example, McKeown et al. (2019), De Smedt
et al. (2018 a-b and 2020), Graham et al. (2017), Harris et al. (2012), Graham et al. (2005)
assessed overall text quality rather than considering multiple criteria as we did.
Other studies did not even assess overall quality, only the number of words and the
presence of expected structural elements (Festas et al., 2015). Although benchmarks
are a recommended assessment method (Bouwer et al., 2023), we opted not to use
them because we wanted to evaluate each criterion in isolation to better
understand how students progressed with each strategy taught.
Before scoring the pre-test, we trained the RA to evaluate the opinion letters.
We collected 22 letters from a 5th grade test class that was not involved in the
project, and the four assistants corrected each of these letters during summer 2020,
before the experiment began. This procedure allowed us to refine the performance
indicators for each criterion. According to Van Gass et al. (2019), such training using
various performance levels is a good way to improve an RA’s comprehension of text
quality criteria.
We scored using the ratings A-B-C-D-E, with A referring to an excellent grade
(clearly above expectations), B to a very good grade, C to an acceptable grade, and
D and E indicating that the student did not meet expectations for the criterion being
assessed. We used the following scale to convert letters to numbers: A=10/10;
B=8/10; C=6/10; D=3/10 and E=0/10 (see Appendix C for an example of a descriptive
scale for awarding points based on the level of success on criterion 1). A similar
scale out of 15 total points was used for the relevance and development of ideas
criterion and out of 20 for the vocabulary criteria.
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 14

The first three classes’ work was rated by multiple RAs (2 or 4 raters; see Table 1).
We calculated inter-rater agreement for the six criteria described above based on
25 students per group. The coefficients ranged between .69 and .98, which is an
adequate level of reliability. Correlations among the six criteria within each group
(EWI+PF, EWI, Control) are presented in Appendix D and provide good support for
the metric’s test-retest fidelity and its convergent and divergent validity.

Table 1. Inter-rater correlations

Teacher #1 Teacher #4 Teacher #7
Criteria T1 T1 T2
4 raters 2 raters 4 raters
Reader awareness .84 .73 .81
Consistency (progression) .90 .69 .79
Consistency (development) .88 .84 .93
Structure and elements .98 .95 .76
Vocabulary .94 .82 .95
Use of various types of sentences .89 .80 .90

Based on these agreements, we used the following process to assess the 22 other
classes. The RA rated all letters using the indicators in our scoring guide. Each of the
25 groups was assigned to one RA, and within each group, three letters were rated
by all four RAs in validation meetings to ensure that they were all applying the
criteria consistently. To avoid any disparity in understanding and applying the
criteria, any indecision regarding scoring was brought back to the team at the
validation meetings for discussion to reach a unanimous decision. Bouwer et al.
(2024) recommend this monitoring of RA rating throughout the process to ensure
that criteria are understood and applied consistently.

8.4 Data Analysis

We performed an analysis of covariance (ANCOVA) with the PROC MIXED
procedure of SAS® (version 9.4) to test our research hypotheses. Fundamental to
each hypothesis is the group*time interaction, which we expected to be significant.
When the group*time interaction term was significant, simple effects were
calculated. The multilevel structure of the data was considered, with students
nested in their classroom. Analyses were conducted on the following dependent
variables: perceived self-efficacy, overall achievement score and the six criteria
composing this score. The covariates for each of these analyses were student
gender and student age. The analyses were all performed with multiple imputations
for handling missing values (see Table 2).
15 | JOURNAL OF WRITING RESEARCH
Table 2. Mean, Standard Error, p-value and Cohen’s d for Self-Efficacy, Assessment Criteria and Overall Score, with Multiple Imputations
T1 T2 p Cohen’s d

n Adjusted Mean (SEa) n Adjusted Mean (SE)

Self-efficacy

EWI+PFb 162 68.17 (2.02) 162 79.45 (2.01)

EWI c 157 70.56 (2.08) 157 77.74 (2.09)

Control 164 72.38 (2.11) 164 74.19 (2.07)

Interaction group x time <.001

T1: EWI+PF vs. EWI 1.00 .09

T2: EWI+PF vs. EWI 1.00 .07

d
T1: EWI+PF vs. contr .45 .16

T2: EWI+PF vs. contr .21 .20

T1: EWI vs. contr 1.00 .07

T2: EWI vs. contr .69 .14

EWI+PF: T1 vs. T2 <.001 .43

EWI: T1 vs. T2 <.001 .30

Contr: T1 vs. T2 .93 .07

Reader awareness

EWI +PF 162 5.35 (.22) 162 7.21 (.21)

EWI 157 5.28 (.21) 157 7.06 (.20)

Control 164 4.81 (.21) 164 5.71 (.21)

.002
Interaction group x time

T1: EWI+PF vs. EWI 1.00 .03

T2: EWI+PF vs. EWI 1.00 .06

T1: EWI+PF vs. contr .24 .20

T2: EWI+PF vs. contr <.001 .56

T1: EWI vs. contr .33 .18

T2: EWI vs. contr <.001 .52

EWI +PF: T1 vs. T2 <.001 .53

 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 16
T1 T2 p Cohen’s d

n Adjusted Mean (SEa) n Adjusted Mean (SE)

EWI: T1 vs. T2 <.001 .55

Contr: T1 vs. T2 <.001 .27

Consistency:
Progression of ideas
EWI+PF 162 6.37 (.16) 162 7.58 (.14)

EWI 157 6.57 (.14) 157 7.52 (.14)

Control 164 6.24 (.14) 164 7.08 (.14)

Interaction group x time .13

Consistency: Development of
ideas
EWI+PF 162 7.84 (.25) 162 9.86 (.26)

EWI 157 7.85 (.25) 157 9.75 (.25)

Control 164 7.47 (.25) 164 9.69 (.25)

Interaction group x time .64

Structure and elements

EWI+PF 162 3.20 (.30) 162 8.13 (.28)

EWI 157 2.99 (.29) 157 7.16 (.29)

Control 164 3.66 (.29) 164 6.34 (.29)

Interaction group x time <.001

T1: EWI+PF vs. EWI 1.00 .06

T2: EWI+PF vs. EWI .06 .27

T1: EWI+PF vs. contr .75 .12

T2: EWI+PF vs. contr <.001 .49

T1: EWI vs. contr .30 .18

T2: EWI vs. contr .12 .22

EWI+PF: T1 vs. T2 <.001 1.10

EWI: T1 vs. T2 <.001 .92

Contr: T1 vs. T2 <.001 .54

17 | JOURNAL OF WRITING RESEARCH

T1 T2 p Cohen’s d

n Adjusted Mean (SEa) n Adjusted Mean (SE)

Vocabulary
EWI+PF 162 12.23 (.26) 162 15.14 (.25)

EWI 157 12.84 (.24) 157 15.10 (.24)

Control 164 12.68 (.24) 164 14.48 (.24)

Interaction group x time .002

T1: EWI+PF vs. EWI .24 .19

T2: EWI+PF vs. EWI 1.00 .01

T1: EWI+PF vs. contr .57 .14

T2: EWI+PF vs. contr .15 .21

T1: EWI vs. contr 1.00 .05

T2: EWI vs. contr .21 .20

EWI+PF: T1 vs. T2 <.001 .82

EWI: T1 vs. T2 <.001 .71

Contr: T1 vs. T2 <.001 .52

Use of various types of sentences

EWI+PF 162 5.63 (.21) 162 7.50 (.20)

EWI 157 5.56 (.21) 157 7.13 (.20)

Control 164 5.27 (.21) 164 6.26 (.21)

Interaction group x time .01

T1: EWI+PF vs. EWI 1.00 .03

T2: EWI+PF vs. EWI .63 .15

T1: EWI+PF vs. contr .69 .13

T2: EWI+PF vs. contr <.001 .47

T1: EWI vs. contr 1.00 .11

T2: EWI vs contr .01 .33

EWI+PF: T1 vs. T2 <.001 .55

EWI: T1 vs. T2 <.001 .45

 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 18
T1 T2 p Cohen’s d

n Adjusted Mean (SEa) n Adjusted Mean (SE)

Contr: T1 vs. T2 <.001 .30

OVERALL SCORE

EWI+PF 162 40.62 (.84) 162 55.42 (.84)

EWI 157 41.07 (.84) 157 53.71 (.83)

Control 164 40.15 (.83) 164 49.57 (.81)

Interaction group x time <.001

T1: EWI+PF vs. EWI 1.00 .04

T2: EWI+PF vs. EWI .45 .16

T1: EWI+PF vs. contr 1.00 .04

T2: EWI+PF vs. contr <.001 .56

T1: EWI vs. contr 1.00 .09

T2: EWI vs. contr .001 .40

EWI+PF: T1 vs. T2 <.001 1.33

EWI: T1 vs. T2 <.001 1.12

Contr: T1 vs. T2 <.001 .89

Note. The p-value is adjusted by the Bonferroni correction for comparisons between groups or between times.

a SE: standard error

b
EWI+PF: Explicit Instruction + Peer Feedback group
c EWI: Explicit Instruction only
d Contr: Control group
Covariables: Student age and student gender (see Appendix E)
19 | JOURNAL OF WRITING RESEARCH

9. Results
The pattern of missing values is presented in Table S1 (see Online Supplement).
Students who responded at T1 only or were evaluated only once (T1 or T2) by their
teacher have similar scores compared to students for whom we have data at both
times. The Little’s MCAR tests were not significant for self-efficacy variables (x2(116) =
109.32; p = .66) nor for evaluation criteria (x2(12) = 10.79; p = .55). Data are thus MCAR.
Although missing values do not represent a potential bias to the precision of
statistical estimates, we used multiple imputations with 20 estimated samples
generated to keep a sufficient degree of statistical power. Consequently, analyses
were performed 20 times for each of the 20 samples generated. Once these analyses
were performed, the MIANALYSE function synthesized the 20 results into a single
result. This procedure enabled us to calculate statistical parameters such as
adjusted means and differences between groups and time with their p-value.
Results with multiple imputations are presented in Table 2 and Figures 1 to 6, while
results without multiple imputations are presented in Table S2 of the online
supplement. When the group*time interaction term was significant (p .05), we
compared the results between the groups for T1 and T2, as well as the results
between the two times for each group. Of the eight dependent variables presented
in Table 2, only the two related to consistency (progression of information and
noncontradiction; relevance and development of ideas) did not have a significant
interaction term. Our substantive interpretations below are based only on
significant interaction terms. Main effects are not interpreted. Cohen’s d was
calculated to measure the effect sizes. According to Cohen (1992), a d of 0.2
corresponds to a weak effect, 0.5 to a medium effect, and 0.8 to a strong effect.
Effects of covariates (age and gender) are presented in Appendix E.

9.1 Self-Efficacy
For self-efficacy, we observed a baseline equivalence (see Figure 1). There is no
significant difference among the three groups at T1. However, it should be noted
that students in the Control group had slightly higher scores at T1 than students of
the EWI+PF group, as evidenced by the Cohen’s d of .16. We observed an increase
between T1 and T2 for the two experimental groups, with Cohen’s d = .43 for
EWI+PF and .30 for EWI. For the Control group, the increase is small, with a Cohen’s
d of .07. Surprisingly, these within-subject differences did not translate into
significant differences among the three groups at T2. However, based on Cohen’s
d, some differences are worth noting, though small in magnitude. This is the case
for the difference between EWI+PF and the Control group (.20) and between EWI
and the Control group (.14). Based on these results, we can conclude that both
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 20

85.00

80.00

75.00

70.00

65.00

60.00
T1 T2

EI+PF EI CONTROL

Figure 1. Self-Efficacy Scores at T1 and T2 by Group.

experimental groups produced a degree of increase in students’ self-efficacy scores

that the Control group did not.

9.2 Reader Awareness

For reader awareness, we again observed a baseline equivalence (see Figure 2).
There were no significant differences among the three groups, though Cohen’s d
values tended to reveal small differences, especially between the two experimental
groups and the Control group. We observed an increase between T1 and T2 for the
two experimental groups as well as for the Control group (Cohen’s d for EWI+PF =
.53; EWI = .55; Control = .27).

7.50
7.00
6.50
6.00
5.50
5.00
4.50
4.00
T1 T2

EI+PF EI CONTROL

Figure 2. Reader Awareness at T1 and T2 by Group.

21 | JOURNAL OF WRITING RESEARCH

In contrast to self-efficacy, there are significant differences in reader awareness

between groups at T2, especially between the two experimental groups and the
Control group, with a Cohen’s d of .52 (EWI) and .56 (EWI+PF). Based on these
results, we can safely conclude that both experimental groups produced an
increase in students’ reader awareness scores that was greater than that of the
Control group.

9.3 Structure and Elements

For structure and elements, we observed a baseline equivalence with no significant
difference among the three groups at T1 (see Figure 3). We observed an increase
between T1 and T2 for the two experimental groups as well as for the Control group
(Cohen’s d for EWI+PF = 1.10; EWI = .92; Control = .54), but with greater effect sizes
for the two experimental groups. There were also meaningful differences among
the three groups at T2, especially between the EWI+PF group and the Control group
(.49). Based on these results, we can conclude that the PF component of the EWI
intervention had an added value for this dependent variable. However, it should be
noted that there was no significant difference at T2 between EWI+PF and EWI, nor
between EWI and the Control group.

8.50
8.00
7.50
7.00
6.50
6.00
5.50
5.00
4.50
4.00
3.50
3.00
2.50
T1 T2

EI+PF EI CONTROL

Figure 3. Structure and Elements at T1 and T2 by Group.

9.4 Vocabulary
For vocabulary, we again observed a baseline equivalence, with no significant
difference among the three groups at T1 (see Figure 4). We observed an increase
between T1 and T2 for the two experimental groups as well as for the Control group
(Cohen’s d for EWI+PF = .82; EWI = .71; Control = .52), but with greater effect sizes
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 22

for the two experimental groups. Surprisingly, these differences do not translate
into significant differences among the three groups at T2.

15.50
15.00
14.50
14.00
13.50
13.00
12.50
12.00
T1 T2

EI+PF EI CONTROL

Figure 4. Vocabulary at T1 and T2 by Group.

However, based on Cohen’s d, some differences are worth noting, though small in
magnitude. This is the case for the difference between EWI+PF and the Control
group (.21) and between EWI and the Control group (.20). Based on these results,
we can conclude that both experimental groups produced an increase in students’
vocabulary scores that was greater than that of the Control group.

8.00
7.50
7.00
6.50
6.00
5.50
5.00
T1 T2

EI+PF EI CONTROL

Figure 5. Use of Various Types of Sentences at T1 and T2 by Group.

23 | JOURNAL OF WRITING RESEARCH

9.5 Types of Sentences

For use of various types of sentences, we again found a baseline equivalence (see
Figure 5). There are no significant differences between the three groups at T1. We
observed an increase between T1 and T2 for the two experimental groups as well as
for the Control group (Cohen’s d for EWI+PF = .55; EWI = .45; Control = .30). There
were also significant differences between groups at T2, especially between the two
experimental groups and the Control group with a Cohen’s d of .47 (EWI) and .55
(EWI+PF). Based on these results, we can safely conclude that both experimental
groups produced an increase in student scores for the use of various types of
sentences that was greater than that of the Control group.

9.6 Overall Score

For the overall score, once again we observed a baseline equivalence (see Figure 6).
There were no significant differences between the three groups at T1. We observed
an increase between T1 and T2 for the two experimental groups as well as for the
Control group (Cohen’s d for EWI+PF = 1.33; EWI = 1.12; Control = .89). There were
also significant differences between groups at T2, more precisely between the two
experimental groups and the Control group with a Cohen’s d of .56 (EWI+PF vs.
Control) and .40 (EWI vs. Control). Based on these results, we can safely conclude
that both experimental groups produced an increase in students’ overall scores that
was greater than that of the Control group.

60.00

55.00

50.00

45.00

40.00

35.00
T1 T2

EI+PF EI CONTROL

Figure 6. Overall Score at T1 and T2 by Group.

10. Discussion
The purpose of our study was to compare 5th grade students’ writing performance
and self-efficacy before and after two interventions: explicit instruction with peer
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 24

feedback (EWI+PF) and explicit instruction without peer feedback (EWI), with a BAU
control group. Regarding overall writing performance, students’ scores improved
substantially in all three groups between Time 1 and Time 2 (large effect sizes).
However, this increase was greater in the two experimental groups, as students’
scores differed from those of the Control group at T2. Regarding the six writing
criteria assessed, we found a significant increase in students’ scores for reader
awareness, structure and use of various types of sentences for the two experimental
groups (EWI and EWI+PF). The effects were more tenuous for students’ vocabulary
scores, and we found no effect for the two criteria related to consistency. Students’
self-efficacy scores increased slightly between T1 and T2 in both experimental
groups but stagnated in the Control group. Overall, the results provide support for
the EWI intervention itself and for EWI plus peer feedback. Because there was no
statistical difference between the two experimental groups, we cannot conclude
that peer feedback had any effect on students’ writing performance and self-
efficacy, only for one criterion relating to structure and elements. However, peer
feedback did not negatively affect students’ results either.

10.1 Teacher Implementation

Although teachers in the Control group used adequate pedagogical practices,
teachers in the two experimental groups seemed more effective in fostering self-
efficacy and improving writing performance. This means that the EWI intervention
that we implemented in the classroom by training the teachers is an adequate
method for developing primary school students’ writing skills. It thus compares
favorably to other EWI interventions implemented by teachers (Harris et al., 2012;
McKeown et al., 2016, 2019; De Smedt & Van Keer, 2018a and b; Collins et al., 2021).
One innovative aspect of our study is that it was the first transposition of EWI to a
French-language school system. Our study provides evidence for EWI’s
effectiveness in the Quebec school system.

10.2 Effects of Explicit Instruction on Writing Performance Criteria

The two performance criteria for which we found no significant group by time
interaction were those that are the most complex for students to master, namely
the progression and development of ideas. To explain these findings, we argue
tentatively that two months of intervention may not be enough for students to
master these two writing abilities. The effects were more pronounced for writing
performance criteria relating to more accessible skills such as reader awareness, the
use of various types of sentences, and vocabulary. For structure and elements of
the letter, we found an even more pronounced effect when explicit instruction was
combined with peer feedback: peer discussion about the letter’s structure
apparently helps students to master this component. Our criteria enabled us to
refine our analysis of writing performance and to find, for example, that students in
25 | JOURNAL OF WRITING RESEARCH

the experimental groups outperformed those of the Control group on specific

content but not on others.
However, it is difficult to compare these results with those of other studies,
because other studies using EWI interventions did not analyze the writing’s overall
content quality with reader awareness and consistency criteria as we did. As
discussed in the Method section, past studies have used a variety of indicators of
learning progression, such as overall writing score, adherence to structure and
word count, although this last variable is not always a reliable indicator of text
quality (Graham et al., 2017). These studies do not provide information on specific
components of writing that may not be positively affected by EWI, such as
progression and development of ideas. We therefore believe that further research
should be conducted using a multidimensional evaluation grid of writing
performance to shed further light on the conditions under which EWI allows the
development of complex skills such as argumentative consistency.

10.3 Peer Feedback Effects on Writing Achievement

Our findings provide support for EWI interventions, but limited support for the
added value of peer feedback, at least as regards the variables measured. No
significant differences were found between the two experimental groups, except
on the structure criterion, even though the effect sizes between T1 and T2 were
higher for the EWI+PF group. While these results seem to contradict the literature
reviews of Panadero et al. (2018), Koster et al. (2015), Prata et al. (2019) and Graham
et al. (2015), who report that peer feedback had a significant effect on writing
performance, they are supported by several other studies on the effects of explicit
instruction combined with peer feedback. Graham et al. (2005) implemented
interventions for narrative and argumentative writing by comparing the effects of
using SRSD with or without peer support. They found no significant difference
between the experimental groups. De Smedt and Van Keer (2018a) implemented a
research design comparing interventions with or without peer feedback in 5th and
6th grade classrooms. They found no significant difference between individual
writing groups and peer-assisted groups, only significant effects for the
effectiveness of explicit instruction. They replicated their study, integrating peer
feedback into all phases of the writing process and providing more peer assistance
in teaching descriptive writing (De Smedt et al., 2020). In that study, the EWI+PF
group outperformed the EWI and Control groups.
These inconsistent results highlight the need for more research comparing
groups with and without peer assistance (including feedback) at the end of primary
school. De Smedt et al. (2020) suggest that this collaboration is effective in EWI when
collaboration is a real object of instruction. Although we made peer feedback
anchor charts for students and trained teachers (including after viewing their class
videos) on how to explicitly teach peer feedback to their students, it does not seem
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 26

to have been sufficient for effectively implementing such a complex learning

activity. The videos we saw suggested that most students did not really know what
to do with peer feedback periods unless their teacher joined their team to scaffold
questions about their peers’ work. Teachers thus played an essential catalytic role,
modeling questions and referring students to their anchor charts. This mediation
seems a worthwhile avenue for strengthening the effectiveness of peer feedback
for writing learning.

10.4 Effects of the Intervention on Self-Efficacy

Self-efficacy is a self-perception that students develop over a long period of time.
Students gradually build up their perception of their capability over the course of
their success. While many studies support the positive relationship between self-
efficacy and writing performance (MacArthur & Graham, 2016; Zimmerman &
Risemberg, 1997), it remains difficult to see this relationship emerge in interventions
conducted with students, especially in the classroom context, in a two-month span.
In our study, we found the effects on self-efficacy to be weak for both experimental
groups compared to the Control group. However, both experimental groups had
small gains in self-efficacy between T1 and T2 (Cohen’s d = .43 for EWI+PF and .30
for EWI) whereas the Control group did not (.07). Our study thus supports the
positive effect of EWI on self-efficacy. Repeated practice of the same strategies in
different tasks, always for the same type of text (opinion letter) but with increasing
complexity, reinforced students’ belief in their ability to succeed in writing.
Other research has not found similar effects. De Smedt and Van Keer (2018a)
compared experimental groups using SRSD with and without peer assistance in a
teacher-led intervention, and found the intervention to have no effect on self-
efficacy with 5th and 6th grade students. In another study with a larger sample of
students of the same age with more sustained peer assistance (De Smedt et al.,
2020), they again found mixed results for self-efficacy. Only self-efficacy for ideation
(planning and writing of ideas) showed a significant difference for the EWI+PA
group compared to the EWI and Control groups. They found no significant
difference in self-efficacy for conventions (using appropriate syntax, grammar and
spelling) or self-efficacy for self-regulation (the capacity to regulate their writing).
Even studies in which researchers intervene with students in subgroups have not
found conclusive evidence regarding the effect of explicit instruction on self-
efficacy (Graham et al., 2005). It remains to be seen under what conditions effective
practices such as EWI and peer assistance can enhance self-efficacy.

10.5 Limits of Our Research Design

While our study shows that the explicit instruction intervention we implemented in
classrooms with teachers did contribute to improving student outcomes, some
limitations should be underscored. One limitation relates to the duration of our
27 | JOURNAL OF WRITING RESEARCH

intervention. It lasted six to eight weeks, depending on the number of sessions (two
to three) that teachers conducted in a week. This duration is consistent with that
used in several other studies (De Smedt & Van Keer, 2018a; De Smedt et al., 2020;
Harris et al., 2012). However, others have implemented a shorter intervention
window lasting approximately four to five weeks, with almost four hours weekly
(McKeown et al., 2016), while others implemented longer ones: three months for
Festas et al. (2015) and Rosário et al. (2019), and up to five months for Graham et al.
(2005). The six to eight weeks that our intervention lasted is, in our opinion, the
maximum amount of time we can practically impose on 5th grade teachers in
Quebec, as they have a great deal of other content to cover during the school year.
This duration was sufficient to observe the effects of explicit instruction on writing
performance and self-efficacy, but insufficient for the peer feedback to produce
meaningful results for writing outcomes or increase self-efficacy.
A second limitation relates to the absence of a more distal post-test intervention
measurement. Some researchers (Prata et al., 2019, Festas et al., 2015 and Graham et
al., 2005) used a second post-test a few months after the intervention to measure
how skills were maintained after explicit strategy instruction. This research method
greatly improves the generalizability of findings because any effects are no longer
limited to the intervention setting. However, due to a lack of teacher availability, we
were unable to implement this delayed post-test. After the six or eight weeks of
intervention, the teachers had to teach other content in their 5th grade classes. We
are therefore unable to comment on the maintenance of the effects we obtained
with explicit instruction.
Another limitation relies on the self-efficacy questionnaire, which could have
triggered students in both experimental groups to use the mentioned strategies
during the post-test evaluation. We are not certain of this influence for 10-year-old
students, but there is a possibility that completing the self-efficacy questionnaire
could have activated the use of the strategies taught in both experimental groups.

11. Conclusion
Our study clearly shows the positive effect of our explicit writing instruction (EWI)
intervention implemented in a French-speaking culture. Our results also have a
high ecological validity, as all our EWI and peer feedback interventions were
implemented by teachers. Thus, our results converge with studies that, for decades,
have shown the effectiveness of EWI. It also shows that explicit instruction can be
adapted to and usefully implemented in various cultures and that educational
systems should consider this teaching method. Furthermore, EWI taught through
PBPD not only shows effects on students’ outcome, but also on teachers’
professional development, providing them a valid framework for delivering
effective writing instruction in their classroom (Gillespie Rouse & Kiuhara, 2017).
However, we need to further examine the effect of peer feedback, how students
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 28

use it in class, what they are capable of in terms of peer feedback and what they
specifically get out of it compared to the explicit strategy instruction. For this
purpose, quantitative experimental designs are not sufficient. We need to conduct
research using mixed-method sequential explanatory designs that will not only
shed light on the relationship between variables under study but will also provide
detailed explanations of how students use the tools taught and, above all, the
benefits they get in terms of their learning and motivation.

Author’s note
This research was made possible through funding from SSHRC’s Insight Grants
(#435-2019-0059).

References
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological
Review, 84, 191–215. https://doi.org/10.1037/0033-295X.84.2.191
Boscolo, P., & Ascorti, K. (2004). Effects of collaborative revision on children’s ability to write
understandable narrative texts. In L. Allal, L. Chanquoy, & P. Largy (Eds.), Revision cognitive
and instructional processes (pp. 157–170). Springer. https://doi.org/10.1007/978-94-007-
1048-1_10
Bouwer, R., Van Steendam, E., and Lesterhuis, M. (2024). Guidelines for the Validation of
Writing Assessment in Intervention Studies. In F. De Smedt, R. Bouwer, T. Limpo and S.
Graham (Eds), Conceptualizing, Designing, Implementing, and Evaluating Writing
Interventions (p. 199-223). Brill.
Bouwer, R., Koster, M., & van den Bergh, H. (2023) Benchmark rating procedure, best of both
worlds? Comparing procedures to rate text quality in a reliable and valid manner,
Assessment in Education: Principles, Policy & Practice, 30(3-4), 302-319,
https://doi.org/10.1080/0969594X.2023.2241656
Bruning, R., Dempsey, M., Kauffman, D. F., McKim, C., & Zumbrunn, S. (2013). Examining
dimensions of self-efficacy for writing. Journal of Educational Psychology, 105(1), 25–38.
https://doi:10.1037/a0029692
Bruning, R. H., & Kauffman, D. F. (2016). Self-efficacy beliefs and motivation in writing
development. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing
research (2nd ed., pp. 160–173). Guilford Publications.
Calkins, L. & Boland Hohne, K. (2019). Écrire son opinion pour changer le monde [Changing
the world: Persuasive speeches, petitions, and editorials] (A.-M. Kallemeyn, Trans.).
Chenelière Éducation. (Original work published 2013)
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-
multimethod matrix. Psychological Bulletin, 56(2), 81–105. https://doi.org/10.1037/h0046016
Cartier, S. & Mottier-Lopez, L. (2017). Moyens et dispositifs de l’apprentissage autoréglé et de
la régulation des apprentissages [Means and devices for self-regulated learning and
learning regulation]. In S. Cartier, & L. Mottier-Lopez (Eds.), Soutien à l’apprentissage
autorégulé en contexte scolaire. Perspectives francophones (pp. 1–26). Presses de
l’Université du Québec. https://doi.org/10.2307/j.ctt1vw0rmf.7
Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159.
https://doi.org/10.1037/0033-2909.112.1.155.
29 | JOURNAL OF WRITING RESEARCH

Collins, A.A., Ciullo, S., Graham, S., Sigafoos, L.L., Guerra, S., David, M., & Judd, L. (2021).
Writing expository essays from social studies texts: a self-regulated strategy development
study. Reading & Writing, 34, 1623–1651. https://doi.org/10.1007/s11145-021-10157-2
Crinon, J. (2012). The dynamics of writing and peer review at primary school. Journal of Writing
Research, 4(2), 121–154. https://doi.org/10.17239/jowr-2012.04.02.2
Crinon, J., & Marin, B. (2010). The role of peer feedback in learning to write explanatory texts:
why the tutors learn the most, Language Awareness, 19(2), 111–128.
https://doi.org/10.1080/09658411003746604
De Smedt, F., & Van Keer, H. (2018a). Fostering writing in upper primary grades: a study into
the distinct and combined impact of explicit instruction and peer assistance. Reading and
Writing, 31(2), 325–354. https://doi.org/10.1007/s11145-017-9787-4
De Smedt, F., & Van Keer, H. (2018b). An analytic description of an instructional writing
program combining explicit writing instruction and peer-assisted writing. Journal of
Writing Research, 10(2), 225–277. https://doi.org/10.17239/jowr-2018.10.02.04
De Smedt, F., Graham, S., & Van Keer, H. (2020). “It takes two”: The added value of structured
peer-assisted writing in explicit writing instruction. Contemporary Educational
Psychology, 60, 101835. https://doi.org/10.1016/j.cedpsych.2019.101835
Falardeau, É. (2021). Une difficile mise en œuvre d’un modèle interactionniste de
l’enseignement explicite des stratégies d’écriture. In S. Bissonnette, É. Falardeau, & M.
Richard (Eds.), L’enseignement explicite dans la francophonie. Fondements théoriques,
recherches actuelles et données probantes (p. 157–180). PUQ.
Fayol, M. (2016). From language to text. The development and learning of translation. In C. A.
MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (2nd ed., pp.
130–143). Guilford.
Festas, I., Oliveira, A. L., Rebelo, J. A., Damiao, M. H., Harris, K., & Graham, S. (2015). Professional
development in self-regulated strategy development: Effects on the writing performance
of eighth grade Portuguese students. Contemporary Educational Psychology, 40, 17–27.
https://doi.org/10.1016/j.cedpsych.2014.05.004
Finlayson, K., & McCrudden, M. T. (2020). Teacher-implemented writing instruction for
elementary students: A literature review. Reading & Writing Quarterly, 26(1), 1–18.
https://doi.org/10.1080/10573569.2019.1604278
Gillespie, A., & Graham, S. (2014). A meta-analysis of writing interventions for students with
learning disabilities. Exceptional Children, 80(4), 454–473.
https://doi.org/10.1177/0014402914527238
Gillespie Rouse, A. & Kiuhara, S. A. (2017). SRSD in writing and professional development for
teachers: Practice and promise for elementary and middle school students with learning
disabilities. Learning Disabilities Research & Practice, 32(3), 180–188.
https://doi.org/10.1111/ldrp.12140
Graham, S., & Harris, K. (2005). Writing better. Effective strategies for teaching students with
learning difficulties. P. H. Brookes Publishing Co.
Graham, S., & Harris, K. (2017). Evidence-based writing practices: A meta-analysis of existing
meta-analysis. In R. Fidalgo, K. Harris, & M. Braaksma (Eds.), Design principles for teaching
effective writing: Theoretical and empirical grounded principles (pp. 13–37). Brill Editions.
Graham, S., & Harris, K. (2018). An examination of the design principles underlying a self-
regulated strategy development study. Journal of Writing Research, 10(2), 139–187.
https://doi.org/10.17239/jowr-2018.10.02.02
Graham, S., Harris, K. R., Kiuhara, S. A., & Fishman, E. J. (2017). The relationship among strategic
writing behavior, writing motivation, and writing performance with young, developing
writers. The Elementary School Journal, 118(1), 1–23. https://doi.org/10.1086/693009
Graham, S., Harris, K. R., & Mason, L. (2005). Improving the writing performance, knowledge,
and self-efficacy of struggling young writers: The effects of self-regulated strategy
development. Contemporary Educational Psychology, 30(2), 207–241.
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 30

https://doi.org/10.1016/j.cedpsych.2004.08.001
Graham, S., Harris, K., & McKeown, D. (2013). The writing of students with learning disabilities,
meta-analysis of self-regulated strategy development writing intervention studies, and
future directions. In H. L. Swanson, K. Harris, & S. Graham (Eds.), Handbook of learning
disabilities (pp. 405–438). Guilford Press.
Graham, S., Hebert, M., & Harris, K. R. (2015). Formative assessment and writing. A meta-
analysis. The Elementary School Journal, 115(4), 523–547. https://doi.org/10.1086/681947
Guay, F., Gilbert, W., Falardeau, É., Bradet, R., & Boulet, J. (2020). Fostering the use of
pedagogical practices among teachers to support elementary students’ motivation to
write, Contemporary Educational Psychology, 63 101922,
https://doi.org/10.1016/j.cedpsych.2020.101922
Harris, K., Lane, K. L., Graham, S., Driscoll, S. A., Sandmel, K., Brindle, M. (2012). Practice-based
professional development for self-regulated strategies development in writing: A
randomized controlled study. Journal of Teacher Education, 63(2), 103–119.
https://doi.org/10.1177/0022487111429005
Hattie, J. & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1),
81–112. https://doi.org/10.3102/003465430298487
Koster, M., Tribushinina, E., de Jong, P. F., & van den Bergh, H. (2015). Teaching children to
write: A meta-analysis of writing intervention research. Journal of Writing Research, 7(2),
249–274. https://doi.org/10.17239/jowr-2015.07.02.2
Lundstrom, K., & Baker, W. (2009). To give is better than to receive: The benefits of peer review
to the reviewer’s own writing. Journal of Second Language Writing, 18(1), 30–43.
https://doi.org/10.1016/j.jslw.2008.06.002
MacArthur, C. A. (2016). Instruction in evaluation and revision. In C. A. MacArthur, S. Graham,
& J. Fitzgerald (Eds.), Handbook of writing research (2nd ed., pp. 272–287). Guilford
Publications.
MacArthur, C. A., & Graham, S. (2016). Writing research from a cognitive perspective. In C. A.
MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (2nd ed., pp.
24–40). Guilford.
Martinelli, M., & Mraz, K. (2017). Les tableaux d’ancrage. Des référentiels dynamiques pour
maximiser l’apprentissage [Smarter charts – K-2] (J. L’Italien, Trans.). Chenelière. (Original
work published 2015)
McKeown, D., Brindle, M., Harris, K. R., Graham, S., Collins, A. A., & Brown, M. (2016).
Illuminating growth and struggles using mixed methods: Practice-based professional
development and coaching for differentiating SRSD instruction in writing. Reading and
Writing, 29(6), 1105–1140. https://doi.org/10.1007/s11145-016-9627-y
McKeown, D., FitzPatrick, E., Brown, M., Brindle, M., Owens, J., & Hendrick, R. (2019). Urban
teachers’ implementation of SRSD for persuasive writing following practice-based
professional development: positive effects mediated by compromised fidelity. Reading
and Writing, 32(6), 1483–1506. https://doi.org/10.1007/s11145-018-9864-3
McKeown, D., FitzPatrick, E., & Sandmel, K. (2014). SRSD in practice: Creating a professional
development experience for teachers to meet the writing needs of students with EBD.
Behavioral Disorders, 40(1), 15–25. https://doi.org/10.17988/0198-7429-40.1.15
Ministère de l’Éducation, du Loisir et du Sport (2013). Faits saillants. Épreuve obligatoire
d’écriture de juin 2013. 2e année du secondaire. [Highlights. June 2013 mandatory writing
test. Secondary 2.] Gouvernement du Québec.
Mottier-Lopez, L. (2012). La régulation des apprentissages en classe [Regulation of learning in
the classroom]. De Boeck.
National Center for Education Statistics. (2012). The Nation’s report card: Writing 2011 (NCES
2012-470). Institute of Education Sciences, U.S. Department of Education.
Pajares, F. (2003). Self-efficacy beliefs, motivation and achievement in writing: A review of the
literature. Reading & Writing Quarterly, 19(2), 139–158.
31 | JOURNAL OF WRITING RESEARCH

https://doi.org/10.1080/10573560308222
Palermo, C. & Thomson, M. M. (2018). Teacher implementation of Self-Regulated Strategy
Development with an automated writing evaluation system: Effects on the argumentative
writing performance of middle school students. Contemporary Educational Psychology,
54, 255–270. https://doi.org/10.1016/j.cedpsych.2018.07.002
Panadero, E., Jonsson, A., & Alquassab, M. (2018). Providing formative peer feedback. In A. A.
Lipnevich, & J. K. Smith (Eds.), The Cambridge handbook of instructional feedback (pp.
409–431). Cambridge University Press. https://doi.org/10.1017/9781316832134.020
Prata, M. J., de Sousa, B., Festas, I., & Oliveira, A. L. (2019). Cooperative methods and self-
regulated strategies development for argumentative writing. The Journal of Educational
Research, 112(1), 12–27. https://doi.org/10.1080/00220671.2018.1427037
Rogers, L. A., & Graham, S. (2008). A meta-analysis of single subject design writing intervention
research. Journal of Educational Psychology, 100(4), 879–906.
https://psycnet.apa.org/fulltext/2008-16034-011.html
Rosário, P., Högemann, J., Núñez, J. C., Vallejo, G., Cunha, J., Rodríguez, C., & Fuentes, S. (2019).
The impact of three types of writing intervention on students’ writing quality. PLoS ONE,
14(7), e0218099. https://doi.org/10.1371/journal.pone.0218099
Van Gasse, R., Lesterhuis, M., Verhavert, S., Bouwer, R., Vanhoof, J., Van Petegem, P., & De
Maeyer, S. (2019). Encouraging professional learning communities to increase the shared
consensus in writing assessments: The added value of comparative judgement. Journal of
Professional Capital and Community, 4(4), 269–285. https://doi.org/10.1108/JPCC-08-2018-
0021
Wigglesworth, G., & Storch, N. (2012). What role for collaboration in writing and writing
feedback. Journal of Second Language Writing, 21(4), 364–374.
https://doi.org/10.1016/j.jslw.2012.09.005
Zimmerman, B. J., & Risemberg, R. (1997). Becoming a self-regulated writer: A social cognitive
perspective. Contemporary Educational Psychology, 22(1), 73–101.
https://doi.org/10.1006/ceps.1997.0919
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 32

Appendix A

Student Characteristics

Experimental Experimental Control group

group – peer group –
feedback individual
Variables n % n % n %
Gender
Girl 80 49.4 80 51.0 76 46.3
Boy 82 50.6 77 49.0 88 53.7
Total 162 100 157 100 164 100

Age
9 – 10 years 108 66.7 132 84.1 145 88.4
11-13 years 54 33.3 25 15.9 19 11.6
Total 162 100 157 100 164 100

Home language
French 143 88.3 130 88.8 140 85.4
English 4 2.5 1 .6 2 1.2
Other 15 9.3 26 16.6 22 13.4
Total 162 100 157 100 164 100
33 | JOURNAL OF WRITING RESEARCH

Appendix B

Self-Efficacy scale adapted from Bruning et al., 2013.

We used a Likert scale ranging from 0 (Not confident at all) to 100 (Absolutely
confident).

IDEATION
1. I can think of many ideas for my writing.
2. I can put my ideas into writing.
3. I can think of many words to describe my ideas.
4. I can think of a lot of original ideas to write about.
REGULATION
5. I can concentrate for at least one hour when I write.
6. I don’t get distracted while I write.
7. I can control myself when I write, even when it’s difficult.
8. I can keep writing even when it’s difficult.
VARIED TYPES OF SENTENCES
9. I can use interrogative sentences (questions) in my text to make my reader
think.
10. I can use imperative sentences (orders, advice) to encourage my reader to
act.
11. I can use exclamatory sentences (e.g.: What a great idea!) to encourage my
reader to act.
READER’S AWARENESS
12. I can think about my reader, his/her ideas, his/her age, to convince him/her
of my opinion.
13. I can speak directly to my reader to better convince him/her.
CONSISTENCY OF THE ARGUMENTATION
14. I can find convincing reasons (arguments) to support my opinion.
15. I can develop the reasons (arguments) that support my opinion, giving
justifications.
STRUCTURE
16. I can write an introduction announcing my opinion and my reasons
(arguments).
17. At the beginning of each paragraph, I can write a reason (argument) that
supports my opinion and justify it correctly.
18. I can write a conclusion that restates my opinion and my reasons
(arguments) in words different from those I’ve already used.
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 34

PEER FEEDBACK
19. I can offer another student clear comments and examples to help him/her
improve his/her text.
20. I can suggest solutions to help another student improve his/her text.
21. I can ask students who comment on my text for clarification when I don’t
understand what they’re saying.
22. I can pay attention to advice given to me by other students about my text.
35 | JOURNAL OF WRITING RESEARCH

Appendix C
Example of Detailed Indicators for the Criterion “Reader’s awareness” According to
the Ratings
A-B-C-D-E

A (10/10):
The student takes the recipient into account throughout the text: they address the
recipient several times at appropriate moments; they greet the recipient at the
beginning and end of their text according to the formulas of the letter; we feel that
they have thought about the characteristics of their recipient in order to write a text
that is interesting for this person in particular. This is reflected, among other things,
in the chosen arguments: the student has chosen relevant and convincing
arguments, considering that their addressee is the school principal. As a rule, they
are polite and respectful. If one puts oneself in the shoes of a school principal, one
is fully satisfied with the proposed letter (in terms of taking the recipient into
account).

B (8/10):
The student takes the recipient into account: they greet the recipient at the beginning
and end of the text according to the formulas of the letter. They address the recipient
at least once in the text in an appropriate manner. In general, we sense that the
student has thought about the characteristics of the recipient in order to make the
reading of the letter interesting. An argument may be less related to the concerns
of a principal and more related to those of a 5th grader. Keep in mind that at this age,
students are still very egocentric and have a hard time putting themselves in the
shoes and head of someone else, especially someone older than they are with
duties quite different from what they know as children. If we put ourselves in the
shoes of a school principal, we are satisfied with the proposed letter (as far as taking
the recipient into account).

C (6/10): The student takes minimal account of the recipient: they greet him at the
beginning and/or end of the text, but may do so awkwardly (in some cases, he may
omit one of the greetings). They may or may not address the recipient in the body of
the text: in the majority of cases, we feel that it is “tacked on” and that the student
has taken little account of the characteristics of the recipient to make the reading of
their letter interesting. If we put ourselves in the shoes of a school principal, we are
more or less satisfied with the proposed letter, something seems to be missing for us
to be convinced (as to the consideration of the recipient).

D (3/10):
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 36

The student takes very little account of the recipient: they may greet the recipient at
the beginning and end of their letter, but in an awkward manner. They never address
the addressee in the body of the text in a judicious manner. One senses that they have
given little thought to the characteristics of the recipient to make their text interesting.
If we put ourselves in the shoes of a school principal, we are not satisfied with the
proposed letter, it lacks several elements and interpellations for us to be convinced (as
to the consideration of the recipient). We can also find clumsiness such as lack of
respect which shows that the student did not take into account the status of their
addressee.

E (0/10):
The student ignores or misidentifies the recipient (e.g., addresses parents, friends or
teacher instead of principal); fails to greet the recipient at the beginning or end of
the letter. Never addresses the recipient in the body of the letter. They do not take
into account the characteristics of the recipient to find relevant and convincing
arguments. If we put ourselves in the shoes of a school principal, we are not at all
satisfied with the proposed letter: we feel that the student wrote the letter without
taking into account that they had an addressee, namely the school principal.
37 | JOURNAL OF WRITING RESEARCH

Appendix D
Correlations Between the Correction Criteria
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
1. Reader’s
.32/.34/.29 .35/.53/.44 .45/.26/.36 .34/.27/.45 .28/.39/.26 .22/.24/.22 .25/.13/.10 .29/.23/.19 .03/.01/.10 .26/.32/.19 .23/.26/.26
Awareness
2. Progression of
- .49/.40/.43 .30/.40/.44 .47/.30/.40 .10/.20/.21 .21/.12/.15 .15/.20/.23 .29/.21/.28 .08/.21/.32 .34/.23/.38 .11/.14/.34
ideas
3. Development of
- .38/.34/.46 .44/.38/.45 .29/.45/.27 .24/.29/.22 .22/.28/.08 .30/.29/.34 .07/.18/.24 .32/.34/.31 .26/.16/.29
ideas
4. Structure and
- .22/.22/.46 .15/.29/.17 .29/.26/.17 .34/.25/.09 .28/.26/.31 .30/.22/.10 .20/.34/.34 .20/.21/.36
elements
5. Vocabulary - .42/.43/.48 .19/.31/.22 .13/.20/.24 .31/.17/.32 -.01/ .26/.22 .42/.44/.36 .33/.18/.29
6. Types of
- .01/.15/.02 .09/.10/ -.05 .20/.14/.19 -.05/.21/ -.05 .21/.14/.17 .15/.10/.24
T1

sentences

7. Reader’s
- .36/.37/.14 .45/.61/.31 .32/.46/.38 .52/.67/.29 .52/.59/.20
Awareness
8. Progression of
- .38/.47/.31 .37/.38/.38 .45/.46/.33 .22/.19/.14
ideas
9. Development of
- .25/.38/.35 .56/.57/.31 .48/.42/.30
ideas
10. Structure and
- .20/.46/.27 .15/.31/.00
elements
11. Vocabulary - .46/.62/.39
12. Types of
-
sentences
T2

Note. Correlations before the first slash are for the EWI+PF group, between the two slashes for the EWI group and after the second slash for the control group.

Specifically, Campbell and Fiske (1959) proposed the following guidelines to assess the construct validity of a multidimensional instrument: 1) convergent validity occurs when the correlation between scores of two
different methods assessing the same construct is high and significant (see correlations in bold in Appendix D); 2) divergent validity occurs when correlations between scores of two constructs assessed by a different
method are lower than convergent correlations (convergent correlations are in italic in Appendix D; 3) a method effect exists when correlations between constructs’ scores of the same method are higher than
convergent correlations (same method correlations are underlined in Appendix D); 4) the pattern of correlations among different constructs’ scores should be similar for different methods. In sum, construct validity
is supported when a) convergent correlations are high, b) divergent correlations are lower than convergent correlations, c) when the method effect is low, and d) when the pattern of correlations among different
constructs is similar for different methods. Although it is impossible to present results in detail here regarding these four guidelines, most of them are respected except that there is a method effect because correlations
among scores of the seven constructs of the same method (see underscored correlations in Appendix D) are sometimes higher than convergent correlations (see correlations in bold in Appendix D).
 FALARDEAU  TEACHER-IMPLEMENTED EXPLICIT WRITING INSTRUCTION | 38

Appendix E
Effects of Covariates on Self-Efficacy, Assessment Criteria and Global Scores
Parameter p
Self-efficacy
Age -5.82 <.001
Gender -4.56 <.001
Reader’s awareness
Age -.30 .003
Gender -.64 <.001
Consistency: Progression of ideas
Age -.13 .12
Gender -.24 .001
Consistency: Development of ideas
Age -.07 .62
Gender -.59 <.001
Structure and elements
Age -.37 .01
Gender -.45 <.001
Vocabulary
Age -.17 .21
Gender -.52 <.001
Use of various types of sentences
Age -.48 <.001
Gender -.43 <.001
Global score
Age -2.09 .02
Gender -3.51 <.001

i
POW = Pick an idea or opinion; Organize and generate notes and ideas; Write and say
more.
TREE = Topic sentence; reason; Explanation; Ending.
All French language materials available to teachers and their students are freely available
ii

online: https://www.strategieslectureecriture.com/groupe-exp%C3%A9rimental-collaboratif