Post on 29-May-2019
Improving cohesion in student writing: A professional experiment to improve
teacher knowledge
Peter Farrell and Glenda Telford
Zeerust Primary School
Zeerust Victoria Australia 3634
Introduction
Under the auspices of the Australian National Curriculum, literacy includes speaking
& listening, reading, literature, language conventions, and writing. Writing is further
divided into five domains, these being: (1) Purpose, Audience and Structure of Texts,
(2) Text Cohesion, (3) Concept of Print and Screen, (4) Creating Text, and (5) Editing
Text (ACARA, 2012). All government schools in Victoria, Australia, where the
authors teach, invest considerable time in teaching literacy. There is an expectation
by the education department that the first two hours of each day will given over to
teaching and learning literacy. The significance of literacy is further emphasised by
an annual test of numeracy and literacy undertaken at grades 3, 5, 7 and 9 across
the nation. As might be expected, the literacy part of the test includes a writing
component. For our small rural school, improving writing is both important and
challenging.
By the age of five a child will be able to orally make and comprehend grammatically
correct sentences, and are likely to have mastered several thousands words
However, the performance of writing is more culturally based and it can take up to
three decades, and tens of thousands of hours of practice, for a person to be
considered an ‘expert’. Kellogg (2008, p.2) likens this to the effort needed to become
a grand chess master or professional musician and he suggests that it is only those
individuals for whom writing is a significant part of their day-to-day work that
achieve such a level of competency. Of course, primary aged students are not
expected to reach this standard but it is our job as teachers to ensure they are
placed upon a trajectory towards it.
As self-study teacher-researchers our work is highly contextual, and any theory we
might generate is only applicable to our own situation (Farrell, in press). Our theory
of practice is underpinned by the idea of the professional experiment, first
described by Donald Schön (1987, p.70) in his book, “Educating the Reflective
Practitioner”. There are three types of experiment. The first, an exploratory
experiment is where the professional implements a change with no expectation for
the outcome. The second, is called a move testing experiment, and is similar to an
exploratory experiment except that the professional does have a desired outcome in
mind. The third kind of experiment called hypothesis testing but unlike an academic
experiment, where variables are controlled and the test is expected run its course, a
significant feature of this method is that the professional will manipulate the
variables and interfere with the experiment in order to force a desired outcome.
2
Schön (1987) is perhaps better known for his endorsement of reflection as the
means by which the professional person improves their professional knowledge and
practice. He suggests two iterations of reflection to occur after an intervention. The
first reflection, happens quite soon after the intervention, and is a relatively simple
event looking at the process used and the outcome achieved. The second reflection
occurs much later, and is a much more thoughtful and comprehensive process.
At our school there are six steps in our professional experiment and we make use of
the reflective process as part of our approach. These are: (1) Identifying the
problem of practice, (2) Knowing-in-action, (3) Make a plan, (4) Implement, observe
and/or manipulate the plan, (5) Reflection-on-reflection-in-action (the first
iteration), and (6) Reflection-on-reflection-on-action (the second iteration).
Step 1: Identifying the problem of practice
For us, it begins with an intuition, a feeling that something isn’t quite right, and, on
further exploration, this feeling is backed up by the evidence. For a couple of years
now we have felt that writing at our school has been stymied by the reluctance of
our students to write complex text. Last year we tackled the issue through an
experiment about teaching and learning grammar (Farrell 2015). Our measure for
that experiment was the language conventions component of the annual NAPLAN
(national assessment program for literacy and numeracy). All the senior students in
the school undertake the test even though only grade three and five students are
required to do so. The 2015 test provided data for an effect size of 1.16 over the
3
previous 12 months that compared very favourably with the 0.6 and 0.62 results of
previous years. Hattie (2009, p.17) who writes about effect sizes explains that the
hinge point is 0.4, which he says, is the effect a competent teacher would normally
realise. So, on the face of it, our students had responded well to our grammar
experiment, but we did not feel we had finished the job.
Improving our students understanding and application of language conventions is a
technical approach. It is more about addressing the secretarial components of a
writing task. We wondered how we might improve the authorial aspect of our
students’ writing. Authors control what they write, they know where they are going
and they are aware of their audience and try to meet the needs of that audience.
How could we move our focus from the technical (secretarial) to the creative
(authorial) aspects of writing without losing that we had gained in the previous
year.
Step 2: Knowing-in-action
Schön (1987, p.13) says professionals think on their feet, calling on their intuition,
experience, knowledge, and personal theory of practice to solve issues in their
professional lives, Schön (1987, p.22) calls this practice knowing-in-action. One of
the unanticipated outcomes of our earlier experiment was the manufacture of a
spread-sheet for the collection of data about our students’ understanding and
application of grammar (Farrell 2015, p.7). Across the top of the spread-sheet were
recorded all our students names and what grade they were working towards. Down
4
the sides it showed a break down of the learning area into various domains and the
expectation of each grade level. These were sourced from the Australian National
Curriculum and they worked so well for grammar we applied the same model to
ascertaining where our students were performing in writing.
In 2015 our target for a foundation level subject was 100%. That is we expected all
of our students to perform at this level. However our target for grade six learning
was just 14.3% - this is because we only had enrolled two grade 6 students (14.3%
of the total enrolment). Students achieving a particular expectation would be scored
1, for those students not performing to the level expected, their cell in the spread-
sheet would be left blank. Individual student scores could be calculated against the
entire requirement of the learning area. For example, the foundation student was
expected to score 14% to be on track while a grade 3 student needed to score 57%
of the total curriculum to be adjudged as up to standard. Naturally, grade 6 students
were required to perform all of the expectations (100%). With this instrument we
could ascertain what areas needed to be taught and how individual students were
performing. The spread-sheet for grammar proved so useful in our small school
setting that others were created for other parts of the literacy program including
writing.
Zeerust Primary School has just two teaching staff including the principal but
despite our small size we do hold curriculum committee meetings on a weekly basis
and once each fortnight we meet as a professional learning team (PLT). The
5
curriculum committee pays attention to what is happening with individual students,
aspects of our teaching practice, and to our interventions. The PLT is about
collecting data and responding to that, it is also about improving our teaching
practice and it was this ‘group’ that undertook a review of the students’
performances against the writing expectations of the national curriculum. The
domains of general concern to us were Creating Text, Text Cohesion and Editing
text. Notwithstanding the fact that these analyses were conducted early in Term 2
and we would anticipate that any student in a particular grade might be expected to
demonstrate an improvement in performance by later in the year, some data
indicated that this was by no means a certainty.
Figure 1: Showing school performance against one of five domains of the national curriculum for writing, this being Create Text. The left hand column shows grades and the two right hand columns show our target based upon student enrolment in different grades, and what was being achieved at the beginning of our experiment. The large column shows the criteria some of which paraphrase those in the national curriculum. Teacher judgement indicates an overall weakness in the creating text domain throughout the school.
Figure 1 relates to the domain of Creating Text in the writing domain of the national
curriculum. At grade 1 the expectation is that all students in grade 1 and above,
which is 93% of our 2015 enrolment, will be able to write short imaginative and
information texts. Teacher assessments, as at early Term 2, are that just half of our
students can do this. The grade 2 expectation is the same as for grade 1 but with
6
evidence of structure, grammar and meeting the needs of an audience. 85.7% of our
students should be doing this but just a third are. The situation is more concerning
at grade 4, 5 and 6 level (it should be noted we have no grade 5 students enrolled at
our school in 2015 when this intervention took place).
Figure 2: Showing school performance against one of five domains of the national curriculum for writing, this being Text Cohesion. Teacher judgement indicates an overall weakness in the text cohesion domain throughout the school.
Turning now to the domain of Text Cohesion in the national curriculum for writing
teacher judgements indicate that apart from a high level understanding around
sentence structure and changing the emphasis, text cohesion is not understood by
the students and neither, as it turns out, is it specifically taught apart from some
extensive work on paragraphing in the senior class (grade 3-6) early in the school
year.
Figure 3: Showing school performance against one of five domains of the national curriculum for writing, this being Edit Text. Teacher judgement indicates an overall weakness in the text cohesion domain throughout the school.
7
Editing Text is a domain in the national curriculum for writing and a quick study of
these data might indicate that the school has failed to teach this domain at all. Small
school data sets are notorious for their fluctuations and, had we been able to include
the data for the three students who graduated at the end of last year, then these data
would have looked much better. What is apparent, for those students who remain, is
that Editing Text, will be a serious intervention.
At our school the teaching of writing is broken down into three basic forms,
narrative, exposition, and persuasive, and particular genre are taught where
appropriate. On reflection it would fair to say that our students are not really
challenged with their writing unless they already generating written product that
competently done, but where the student has not pushed any boundaries. It was this
response to simple writing tasks from a number of our senior students that led to
the previous experiment about grammar (Farrell 2015). What those students, in
that experiment, had in common, were that they were already quite creative with
their writing and they needed to master more complex secretarial writing to do with
spelling, punctuation and grammar. However, those students have since graduated
and of those who remain, their current writing needs relate to mastering more
complex authorial matters; these being matters of organisation, coherence and
detail.
We believe that a professional experiment in a school should achieve two outcomes:
(1) Address a problem of practice, and (2) Improve the professional knowledge of
8
the teacher. Looking at our practice it would be fair to say that our teaching
approach began with teaching a structure, next we would look at coherence and
how the piece makes sense, or not. The details were given fairly short shrift by
comparison. At our school writing exposition was given more weight than writing
narrative. We needed to improve our professional knowledge of authorial skills
around the narrative form.
Step 3: Make a plan
Our working hypothesis was simple. We would focus on narrative writing and we
would reverse the order of significance in teaching our students about writing
complex text. We would start with the details, and work hard to ensure that these
were coherent. Only then would we have the students make a choice about which
structure would be used to tell our narrative. We believed we would get more
content (creating text) and that the stories would be more coherent and that the
editing process would be more effective.
We also determined to become authors of fiction ourselves and write alongside our
students (Kelly 2011) and we were to apply this simple model (Black 2015):
Character + Setting + Event = Plot. The teaching goal here was to encourage the
children to think about the details first and how certain characters are likely to
respond in an acceptable way in a particular setting or, to put it another way, that
the links in a story can be highlighted ahead of time; this was about encouraging text
cohesion in a piece of writing. Paralleling the work in the classroom our PLT
9
(professional learning team) read and discussed a paper by Kellogg (2008) that
examines the uptake of writing skill from a cognitive perspective.
Step 4: Implement, observe and/or manipulate the plan
Senior Class (grade 3-6)
Written in first person by the first author
In this class the students were instructed and encouraged to develop a character,
setting and event based upon three dot points for each. I would model the writing of
each of these in turn. My example character was 50 years old, happy, and a teacher.
My setting was a large dilapidated shed, in an isolated area where the wind blew
most of the time. The event I chose was concerned with the arrival of a stranger, a
car that wouldn’t start, and no mobile phone service. I would write at the same time
as the children with my work being displayed on a large screen. I would share my
authorial thinking before, during and after a writing session. My final act of planning
was to demonstrate how to write a plot based upon my character, setting and event.
This is my character:
Character
50 year old
happy
teacher
10
The children only watched the start of this process and then, using word-
processors; they worked on their own character. I brought them back near the end
of my session to share what I had done.
Wendy was a teacher in a rural secondary school where she taught home economics.
She had always loved cooking and especially eating, and she was just a little bit dumpy
in shape. OK. Wendy was fat!
Wendy had recently turned 50 and her friends had sprung a surprise party on her. It
had been a wonderful event and she had never been happier. Many of her old friends
from around the country had been able to come and it had been great to catch up with
them. And the food was heavenly.
Her husband, William, was always telling her off about eating too much. “Your heart,
Wen, you have to look after your heart,” he would say. It was the only time Wendy felt
unhappy. William remembered his own heart attack all too well and everyday he took
a cocktail of drugs to stay healthy. Wendy knew she ate too much and did get
breathless easily, but she did so enjoy a little cake, or two. It was easier when William
went off to his shed.
One of her children, Michael, lived nearby, by himself. He said he was too busy for a
girlfriend. Wendy suspected he did not like anyone but himself. Tom lived in another
state, happily married with a beautiful child that Wendy would like to see more often.
11
I needed Wendy to struggle with physical activity, and being an overweight cook
was acceptable enough. I also needed an excuse for someone to be at a remote shed,
and I needed a convincing cause for Wendy to go there. A poor phone service and a
husband who had forgotten his heart medication was a very good motive in my
honest opinion. I also needed to give a heads-up that a fit young son existed and
lived nearby.
One of the difficulties encountered by some students during the planning phase is
that they simply wanted to get on with writing the story and so, in order to value
this, I instructed these children not to delete their work, but park it somewhere in
another file, as a text fragment. The argument was they might refer to it later and it
was a simple task for a word processor.
Peer-assistance is reported as having a 0.75 effect size on improving people’s
writing (Kellogg 2008, p.20) and so the senior class were instructed to review one
other person’s writing and write post-it notes to the writer on how that description
could be improved. The idea here was that this action would put editing and
audience awareness in more pre-eminent positions than was usually the case.
We began with reviewing each other’s character descriptions and some comments
were useful: (1) I would like you to explain how she started ballet, (2) Why does
James hate people? (3) Please provide more of an explanation about why he likes
12
being tall, (4) This is a very big paragraph, (5) Tell me about his emotions and, (6)
Why did Jean hit her boss? Some comments were less useful: (1) More about
character, (2) Spelling and, (3) How old he is [sic].
Peer commentary around the setting included: (1) ‘Good Job!’ which was
encouraging but not particularly informative, (2) Punctuation, was less encouraging
and not all that specific, whereas, (3) Consider reading this out loud provides sound
advice to the writer, (4) ‘On all sides of the court there was Lakers painted in yellow’
… that sentence could shorten up a bit Its very very long [sic], (5) The sign saying
Jackson Town has a red paint timber roof what shines on the pet store. This is a bit
like your story. The peer was changed around for each reviewing session.
When auditing the event descriptions, the more useful comments were: (1) You got
into your story, (2) You don’t have to but maybe you could write a bit more about
how it started, (3) Too big, it makes no sense, things are out of order, and (4)
Describe the start more. Less useful comments were: (1) Good work, (2) Really,
really good job. Although, these latter comments may be suggesting that at least two
of the children are writing better.
What I was beginning to see were signs of quite specific advice about what was
wrong and what might be done about it. It was also fairly obvious by this time that
some students were doing almost as much writing as reviewing and were able to
shift their attention between these tasks at need.
13
I was particularly taken by the idea that writing imposes a huge cognitive load on
the author (Kellogg 2008). I shared this information with the class and insisted they
work silently and not-be distracted by others. They would call this being in the
‘zone’. My observations suggested that by this time, we had been writing for six days
(three x 80 minute sessions per week), the stories had taken on a life of their own.
The students were producing up to 150-250 words per 80-minute session. Text
production was definitely up.
Writing the event was more difficult for everyone including myself. I ended up
writing individual pieces about my stranger, the poor phone service, and had to
describe two cars that were difficult to start. I noticed that what a student had
planned as an event at the beginning of the experiment did not quite fit in with the
character or the setting they had now. Could they change their event? Yes, I replied,
or alternatively, they could alter their setting or character to make their texts ‘hang-
together’ (be more coherent). I also suggested that, like me, they might have to
separate their event into the three dot points.
At this point I had written three descriptions as part of my planning but had not yet
written my story. I encouraged the children to hold off from doing this for as long as
possible. At the end of the second week I printed out what the students had
produced at that point. What was obvious to me was while nearly all the students
struggled with my direction not to write their story yet, it was the boys who
14
invariably started writing the story. The girls, for the most part, had kept each
component separate.
And so now we come to the plot. Our working hypothesis is that, character + setting
+ event = plot (Black 2015). The plot is the storyline or scenario for a story. It should
be outlined with an end in mind. I made the point to my class that the plot was like
telling someone about your story rather than telling the story itself. Like explaining
a movie you had seen to someone who had not seen it. I found writing the plot very
easy, it practically wrote itself. However, this is where my students began to blur the
line between planning and production.
As happens in professional experiments, things go wrong, and two of my students
had managed to delete a considerable amount of their planning. Fortunately I had
been keeping hard copies of their work. However, It meant that that continued use
of a word processor was not feasible. The curriculum committee discussed the issue
and determined that the senior students would hand-write their real story using the
same time (40 minutes) and space limitations (four pages) afforded by the NAPLAN
writing test for comparative purposes. They would have access to the latest hard
copy version of their plans.
Personally, I found the limitations (40-minutes and limited amount of paper)
restrictive and I had to end my story somewhat more quickly than I planned.
However, I was able to retain and integrate the more significant parts of my plan
15
and the task was easy. I wrote 1780 words during planning and production. The
completed handwritten story comprised of 505 words – about a third of all the
words I had written. This was my first draft:
The party had been wonderful. Wendy and William had enjoyed it so much, after all it’s
not every day that you have a 50th birthday. All her friends from school had been there
and yes she had eaten far too much food. Wendy planned to spend the morning
pottering around and tidying up but William had already headed off to his shed.
William stopped his ute outside his rented shed. That track is getting worse he thought
to himself. A bit like this shed really. The shed was a dilapidated looking thing that
stood in lonely isolation next to hill in a paddock. It used to hold farm machinery but
now held William’s caravan, a few shelves filled with stuff, a workbench and his pride
and joy, the little red sports car he was restoring. It was this car that brought to this
shed in this windy place. William stopped by the generator to check the fuel level. Low!
He would have to head off the farmer’s house and scrounge some more later.
Wendy had just put her feet up and was enjoying a nice cup of tea and one or three
cakes when William’s medical kit caught her eye. Oh no! He’s left his heart medicine
behind. Wendy’s own heart seemed to beat a little faster. William was planning to stay
in the caravan overnight. He would never notice his pills were missing until the
morning, and that was far too late. What to do? The phone signal was awful out there.
16
“Michael, it’s Mum. Dad’s forgotten his medicine. Oh, I hate leaving messages,” Wendy’s
son, Michael wasn’t answering his phone. Wendy would have to take care of this
herself.
Some time later Wendy stopped her car at the last farm gate. She would have to walk
from here. The day was hot and windy and so she took some water and wore a large
hat and set off for William’s shed. A hot and bothered Wendy arrived at the shed about
twenty minutes later only to find William was gone. He’s probably visiting the farmer
she thought. Wendy decided to wait.
A short time later Wendy stepped out of the caravan with a steaming cup of tea and a
biscuit where she met a most unexpected sight. The stranger was dressed in a business
suit. It was rumpled and dusty and it had a tear in the left leg. The man’s shirt was
unbuttoned and his bright red tie was askew; he looked done in.
“My car ran out of fuel,” he croaked. And then he collapsed.
Wendy was in a pickle. No phone service, no car, except William’s ridiculous little red
car, and no help until William came back. Wendy started first-aid, it was apparent that
the stranger had a serious case of heat-stroke and an ambulance was needed.
17
Just then, her son Michael arrived. He had been concerned by his Mother’s call and
together they were able to save the stranger. William arrived soon after much to the
relief of Wendy.
The following day the students looked at someone else’s work and once again
provided post-it notes with suggestions to make the story better. Useful comments
of an authorial nature included: (1) Do not make it happen so farst [sic]; and (2)
Maybe a better thread, where did the man come from with the girl? Why is he
robbing still?; and (3) Write how much times his team has won a game. Good job.
Comments relating to secretarial matters were, (1) Better grammar, better spelling,
What does “don’t get your hair in a mair” mean? Could you add quotation marks
(“”)? It is a saying; (2) Talk 4 clothes. So are all dogs boys? Proper erasing. Spelling.
One student just commented, “good job”.
The students then edited their own work and wrote a new draft. The students spent
some 20-25 minutes going over their own story and then 40-60 minutes re-writing
it and, with one exception the final drafts were much improved. I was really pleased.
Junior Class (grade F-2)
Written in first person by the second author
In the junior class I modelled the writing process to be followed using supersized
post-it notes. I decided my characters were to be Granddad, Grandma and a 10-year-
18
old girl. My event was to be a possum (a small marsupial) in the roof of the house
that was to be my setting. The house was set just out of town. My students and I
wrote by hand and it was immediately obvious that the pace and volume of work
was to be much slower than in the senior class. I demonstrated to the children how I
would describe this house in writing. I was asked, “Is it a real house?” I said it was a
house I had seen before and so it was real.
There is an old house next to a long driveway. The driveway was straight and in some
places there are potholes and there was brown gravel with some bigger rocks amongst
it. The little house is made of weather boards that were painted darkish grey a few
years ago. The paint was bubbling off the west-facing wall. The house was starting to
show its age with boards near roof rotting and gaps appearing. The cement sheets
under the eave has been broken with some triangular holes where bits had fallen
down. Over near the driveway are two pink lady apple trees growing on the lawn.
The students were given a sheet of lined paper to write their own descriptions of the
house to be used in their story. I collected their work at the end of the session: G1,
B1 and G2 each wrote a list while G3 wrote sentences. One of my students, B2, a
foundation boy, did not have the ability to write his story, and so I acted as his
scribe. He wanted to write about his old house rather than make one up. I said that
was OK.
19
B2’s description was vivid and detailed. It was clear to me that he had a clear mental
picture of the house in his head and he was able to communicate this to me in a very
effective manner. He described the house in sentences that were simple, compound
and complex.
I also teach an
intellectually disabled
student G4, and she asked
if she could describe the
house where she lived now
rather than making one up;
I agreed. G4 wrote a single
sentence and had to be
encouraged to write more,
which she did: “and the
wood is painted blue.” G4’s
example shows a
compound sentence with
two ideas linked together.
This contrasted with my other students who had decided to write about an
imaginary house. What was interesting about all of the writing produced by these
students was that they were lists, a mismatched arrangement of simple sentences
describing a complete mess of a house.
Figure 4: B2's detailed description of a house he knows uses simple, compound and complex sentences
20
I later discussed what my students had done with the first author, and we realised
that here was a way to determine if a student did have a clear mental representation
of what they were describing. Simply by looking at the mix of sentence structures
they used. The more advanced the sentence structures the more capable the
student.
A related issue is when ‘orphans’ appear in the student’s writing. Orphan is the term
applied to those little bits of information that appear unrelated to the text around
them. I told my students that if they are writing about the driveway then don’t add
other bits of information about the driveway somewhere else.
In the subsequent sessions I observed that some students seem to go off on tangents
adding more and more details to their list of what the house in their story was
about. Some were beginning to write their story and I cautioned them against this,
saying: “You may be developing ideas as to what your finished story will be about.
We are not writing the actual story but it is important only to write about the items
that are going to be significant to your end story.”
The next phase was to write descriptions of the characters in our stories. Once again
I modelled the process. Similar problems were encountered here as were thrown up
before. Some children wrote lists, some children wanted to start writing their story,
21
some students did not seem to ‘own’ their own character, except where that
character seemed remarkable similar to themselves (B1).
It was during this time I finished reading the paper by Professor Kellogg and I began
to understand just how ambitious I was for the students in my class. Writing is
difficult for teenage authors!
When we met as a curriculum committee to discuss the intervention. The end of
term was fast approaching and report-writing time loomed. The senior class were
ready to write their actual story now but my class was a long way off, even after
seven sessions of writing. We were yet to describe our event let alone work out the
plot. In the interests of expediency I decided the children had to attempt a finished
story. I gave the children 50 minutes, about the same time as the national writing
test, to hand-write their stories.
At last the day came where the students were to write their story. I began the
session by modelling the process they should follow. First, I re-read all my writing
continually referencing how much information I had. Next, I demonstrated how I
would start my story:
Emma had arrived at Nan and Pop’s house for the Christmas holidays. Her mum and
dad couldn’t look after her during the week because they had to go to work. Emma
22
didn’t mind going to Nan and Pop’s as they always did different things and Nan liked
to cook.
The next day Emma got her pink bike from the back veranda and went riding doen the
long straight driveway. She liked to drive down the potholes, especially if it had just
rained and they had water in them.
I left off writing at this time and explained how I had used some of my planning in
my real story. At last the children were then sent off to begin their stories for real
and I returned to my own writing:
Emma’s long brown hair that was sticking out from under her helmet, was bouncing
up and down as she rode in and out of the potholes it was lots of fun.
Pop was outside mowing the lawn around the house. They had a large area of lawn all
round the house. It took Pop ages as he pushed the mower up and down in rows. Pop
got very hot so Nan would bring him a drink of water. I think this is why Pop is so very
tanned because he spends so much time in the sun.
Nan came back outside with a straw broom. She started to sweep the grass off the
paths. Emma went and got another broom to help Nan.
23
It was getting late in the afternoon but the yard was looking neat and tidy. It had a
smell of freshly mown grass too.
During this time some students were reporting to me that they had finished the task,
but a quick study of their work indicated problems with relevance to planning. I sent
them back to try again.
I also had the task of supporting my non-writer, B2, with getting the text down. We
used the notes app in a standard iPad but we had mixed success. It turned out my
dictation to the machine was more successful than his and this took me away from
my own story. However, his ideas did link together in a coherent way, but I am
uncertain whether this was due to my questions or his planning.
All too soon the session had finished and the results were interesting to say the
least. Only one student made any reference to their character and the setting for
their story. For the rest of the students, their first draft beared little if any
relationship to their planning. However, G2, had started her story with, ‘Once upon
a time,’ and her story had a natural place where the setting might have been
described but was not.
I wondered to myself about what this all meant and came up with the following
possibilities: (1) We did not spend enough planning time on preparing for the event
in their story, (2) I should have completed the modelling of the writing of my whole
24
story to the class highlighting the links as I went, and (3) Even though lots of time
was spent planning, the students needed lots of time to produce a first draft.
I need to make a comment now about the editing process post first draft. I made
copies of each student’s writing and took the students through some things to think
about when editing someone else’s work like full stops, capital letters, and better
expression. I stressed how important it was to be able to explain your editing
purpose. The students were paired off with a blue pencil and went off together to
edit each other’s stories. This turned out to be a very cooperative and worthwhile
exercise. Before they started on their second draft I finished my story as they
watched:
Nan and Emma went inside to have a cold drink while Pop the mower away. He soon
joined them and then headed for the shower. Emma switched on the TV and Nan
started to get dinner ready.
After dinner Pop and Emma sat and watched TV while Nana cleared away the dishes.
It was not long before Pop announced he was tired and heading for bed. Emma had
had a busy day and it certainly past her bed time now. Nana switched off the TV.
Emma cleaned her teeth and got ready for bed. Nan tucked her in and said,
“Goodnight.” Nan turned off the light and went to bed herself.
25
Emma was woken in the night. There was a scratching thumping noise outside her
window. She got out of bed and ran into Nan and Pop’s bedroom. They both woke up
immediately and asked Emma what was the matter. She told them about the noise that
had woke her. Nan and Pop and Emma went into the other bedroom to listen to the
noise.
Pop then got his torch and went outside to investigate what was making the noise. Nan
and Emma went outside too. Up under the eve of the house were two beady eyes
peering out from a triangular hole where the piece of cement sheeting was missing.
Pop said, “There’s the problem. It is a possum that must live up in the roof of the house,
and it was scrambling back to get back to where it was living.” You may not know that
possums are nocturnal and so go out during the night time.
Now knowing what was making the noise Nan, Pop and Emma went back to bed.
Tomorrow, at breakfast, they would think about how to solve this problem.
As I wrote I referenced my planning, and made the links clear to my students. On
completion of my writing I passed out fresh sheets of lined paper to the students.
They were allowed to refer to draft one of their stories. Again they were allowed 50
minutes to write and the results were interesting:
(1) B1 wrote a new story that linked his character with activities he has personally
experienced. He did mention a setting but did not describe it. He actually wrote less
26
than in his first draft but it wasn’t a list this time. He used capital letters, full stops
and conjunctions.
(2) G2 was absent.
(3) G3 wrote a new story where the characters in her plan heard a sound in the
garage at night. Her sentences showed considerably more complexity with up to
three ideas linked together. She wrote two pages and full stops were used
consistently but not capital letters.
(4) G4 basically wrote the same story but edited out some details.
(5) G1 wrote a new story where the technical aspects of the writing were well
handled but the authorial aspects, which included the same characters as her plan,
were under done. She wrote less words in her second draft.
(6) G5 wrote a completely new story (she needed to) based on two characters and a
mouse in a bedroom. There were three pages of text each with good punctuation
and frequent use of compound sentences. The story was linked to her planning.
My observations are as follows: (1) that the second draft, after peer editing, and
seeing a completed project, was much improved from a technical standpoint; (2)
that the planning process led to much more writing; (3) many children wrote stories
27
with a similar plot to my own and this might have happened because they never
planned their own event or plot.
Professional learning
Written in third person
The article we read as part our professional learning as teachers was by Professor
Ronald. T. Kellogg, a cognitive psychologist specialising in writing. Kellogg (2008,
p.3) proposed three states in a writer’s development: (1) writing to tell what one
knows, (2) transforming what one knows for the author’s benefit, and (3) crafting
what one knows for the reader’s benefit. Each stage, he says, takes approximately 10
years of dedicated effort to achieve. This indicated to us that our work is about
encouraging our students to better interact their planning, production and
reviewing process. We determined that knowledge crafting was probably beyond
the scope of the students in our school but might be within the reach of the two
teachers, particularly the first author who has been writing exposition at a high level
for many years.
Kellogg (2008, p. 5) discusses in detail the impact of different kinds of memories on
the cognitive processes of a writer. Where processes and concepts are embedded in
and retrievable from long-term memory, working memory is freed up to account for
reader requirements. Where these processes and concepts are not embedded then
the student writer finds it an impossible task to meet the reader’s needs. At a lesser
level the student writer may even struggle to make the text sensible to his or herself.
28
Kellogg (2008) suggests that awareness of one’s audience, even if that audience is
the writer, is a significant marker in determining how coherent a text is. Associated
with this increasing awareness of audience is the demonstration of mastery around
details, which become richer as the child develops as a writer. This is something to
be considered by us as we move forward with our writing experiment. Details are a
precursor to audience awareness. Have we, fortuitously, intuited an approach to
writing which matches the cognitive steps in learning writing?
Step 5: Reflection-on-action
Not withstanding our intention not to engage with the literature in terms of
understanding our experiment. Professional learning did take place in parallel with
our experiment and this had outcomes for our practice. Learning that a person may
need thirty years of sustained practice where a significant component of that
person’s work is concerned with writing is, to say the least, a bit of a revelation. Let
us assume most people start to learn to write at the age of five, in 10 years, at the
age of 15, say grade nine, all going well the individual may just be capable of
producing written text, covering a range of genre and audiences, that makes sense
to them.
This iteration of reflection is about just about outcomes. It is not intended to engage
with the literature at this point just to share our headline discoveries, which were:
(1) Top-down planning increases word count, (2) Some students struggle with not
writing the story, (3) What to leave out, and what to keep in? (4) Students with a
29
clearly defined mental picture use a mix of simple, compound and complex
sentences, (5) An improved second draft indicates? And, (6) Top-down writing suits
better writers while bottom-up suits less able authors.
Our hypothesis was that if the students had enough time to plan their writing we
would see improvements in text production, text coherence and better results from
editing. We did.
(1) Top-down planning increases word count
By adopting a 40-minute hand-written format we were able to make some
comparisons with the NAPLAN writing all the senior students had engaged with
earlier in the year. That task was a persuasive text and they had five minutes of
planning. Our task was a narrative text and the children had had some 480 minutes
of preparation time. Comparing the word counts after the first draft (only one draft
of writing is assessed in NAPLAN) it was noteworthy that the senior girls had
written more words this time than last time. The best increase was 50%, while the
least was 25%, and two students increased their word count by 35%. The boys on
the other hand wrote about the same number of words, or substantially less in spite
of the significantly increased planning time. However, the very fact that all students
are involved in a planning process that requires significant word production during
the planning process for character, setting and events, means they are writing more
words than required for a straightforward draft and redraft approach (a bottom-up
approach).
30
(2) Some students struggle with not writing the story
It was noteworthy that some students struggled with not writing the story – this
was especially true for the junior students and the boys in the senior class. We
wondered whether a large text fragment was an artefact of wanting to get the words
down on paper as per the knowledge-transfer approach exhibited by novice writers.
(3) What to leave out and what to keep in?
After some 480 minutes of planning most senior students had a lot of material to
choose from. However, they had just 40 minutes and limited space in which to write
their narrative. For a lot of students, decisions had to be made, and this presented a
challenge that the better writers were able to resolve under the conditions set by
our writing task. The better writers were able to select those actions and
motivations that advanced their first draft and enhanced its coherence. Poorer
writers did not make good decisions and tended to keep extraneous detail and lose
track of where they were going.
(4) Students with a clearly defined mental picture use a mix of simple, compound
and complex sentences
This observation appeared to be true for each of the planning, production and
revision phases. The phenomenon was first noticed by the second author, who
observed that when her young writers described a real place their sentences were
longer and more detailed, contrasting with those students who were describing an
31
imaginary place. These students tended to make lists using short sentences. We also
noted that where a complete piece of writing was available as a model for the
students then many students would create work that followed a similar plot.
(5) An improved second draft indicates?
The editing of the first draft with a peer and subsequent re-draft provided for a
range of responses from fixing spelling mistakes and grammatical errors, through to
adding in more detail or removing extraneous information. The technical aspects of
writing received a lot of attention and the authorial concerns were addressed too.
Some students carried out substantial revisions for example, one student, a grade 4
boy, re-wrote his entire story going from a third-person to a first-person
perspective. The result was a much more coherent narrative and a better
observation of language conventions. Another boy substantially changed the end of
his story after being told by a peer not to “make his story happen so farst [sic]”. His
re-draft was much improved by not writing his main character out of the action. One
of our oldest writers realised that she had missed subtle opportunities for irony and
with just a few small changes her story became much less obvious and more
interesting for the reader as a result. In nearly every instance, except one, the
second draft was a major improvement on the first. Opportunities to redraft one’s
work after a rigorous editing process conducted with a peer are a must in the
writing classroom.
(6) Top-down suits better writers while bottom-up suits less able authors
32
What we noticed with our students in our classrooms was that top-down planning
seemed to bring about better writing from our more able writers while the bottom-
up approach of draft and re-draft seemed more suited to less able writers who only
wanted to get the text written. The junior class students for the most part
completely ignored their earlier planning and wrote stories unrelated to that work.
It should be noted however that an event and plot had not been modelled to the
juniors at all. When they had seen the complete story modelled then they were able
to write a story with a very similar plot to the one modelled.
Results
We created a second version of our writing spread-sheet and re-entered data and
then calculated effect sizes. We realised a 0.90 effect size for Create Text and the
children did especially well with grade 1, 2, 3, and 4 requirements (see Figure 1) but
in some ways this this was to be a self-fulfilling prophecy given the plan we had
implemented and subsequently manipulated to achieve the result we wanted. The
top-down approach does lend itself to lots of writing especially in the planning
phase.
Text Cohesion realised an effect size of 1.22 especially in the skills required at grade
1 and 2. The junior class teacher had pushed for the writing of sensible sentences
where there was an apparent storyline rather than a mere list. In the senior class the
call was for the story to hang together, and for clearly discernable threads to run
between the character, setting, event and plot, and thus into the story itself. We
33
think that the frequent use of a peer for the editing process provided a real and
present audience.
Edit Text saw an effect size of 1.21. Once again it was a grade 1 requirement
‘rereading to implement changes to improve meaning, spelling and punctuation’
that produced one of the biggest changes. The grade 2 criteria of ‘edit for structure,
sentence composition, spelling and punctuation’ also saw much improvement.
Overall, looking at what our students needed to do to meet their obligations for the
writing curriculum we saw an effect size of 0.83 for this intervention. The hinge
point is 0.4 (Hattie 2009, p.17)and we doubled this. Our experiment was a
worthwhile intervention but will its effect last?
Step 6: Reflection-on-reflection-on-action
With the passage of time we come to step 6 of our professional experiment. This is
where we re-examine the process and the outcomes in light of the literature read at
the beginning, during and since we commenced out work. As professionals we do
not simply confine ourselves to peer-reviewed journals but seek out blogs, articles,
websites, and forums. We are looking for relevant materials to legitimize, augment
and challenge our professional knowledge and thus refine our own theory of
practice (Farrell 2015 in press).
34
A research report (Deane, Odendahl, Quinlan, Fowles, Welsh and Bivens-Tatum
2008) into cognitive models of writing was a larger more broad-ranging document
than Kellogg’s (2008) and they suggest five explanations for why a student is unable
to make the leap from the knowledge telling approach of a novice who is concerned
with getting the words written, to the knowledge transformation of the skilled
writer whose concerns are knowledge development and knowledge expression in a
recursive way (p. 35-38). These are: (1) Interference effects, (2) Lack of strategic
writing skills, (3) Insufficient topic specific knowledge, (4) Weak content-reasoning
and research skills, and (5) Unformed or rudimentary rhetorical goals.
(1) Interference effects are about how writing processes compete for working
memory and knowledge transforming is particularly demanding for the novice
writer, especially where that writer has not yet achieved automaticity in text
production processes, and/or is unable demonstrate higher decoding and
comprehension skills when reflecting on their own writing (p. 36).
(2) The lack of strategic writing skills is about the writer’s ability to appropriately
select a particular writing task (planning, production, and evaluation) at a particular
time and switch between these at need, or to devote extended time to a particular
task where it is effective to do so. Novice writers do not display this control (p. 36).
35
(3) ‘Topic knowledge is a major predictor of writing quality (p.37)’ irrespective of
whether the approach to text production follows a bottom-up or top-down
approach.
(4) Weak content-reasoning and research skills relates to the act of writing as being
about solving problems using critical thinking. These skills are likely to be missing
in the novice writer’s repertoire. Problems to be solved by the essayist may be
concerned with audience, purpose, ideas, or a lack of information while the
narrative story-teller must be concerned with action and motive (p. 37).
(5) Unformed or rudimentary rhetorical goals are about the extent to which the
writer simply puts words down on a page to share his/her own knowledge, or
writes in a way to transform that writing into something that improves his/her own
understanding, or ultimately the understanding of others – knowledge crafting as
Kellogg calls it. The writer’s awareness of, and obligation to, the intended audience,
is mediated by his/her understanding of the structure provided by particular genre
(p.38).
The point made here is that all writing instruction needs to overcome these barriers.
It would appear, on the surface at least, that a bottom-up approach by-passes many
of the barriers described here. Free-writing is a ‘text on paper’ approach that
facilitates word production, pre-selects the writing task (planning), allows for
differences in topic knowledge, leaves off problem solving until later drafts, and
36
finally, ignores any rhetorical goals in the first instance. For readers unaware of the
free-writing process a short explanation of how we do it at our school might be
helpful (Farrell, 2014 p.41). The students are provided with a topic, and required to
generate at least three ideas about it along with supporting details. There is a five-
minute time limit and the students are required to write as fast as possible.
Punctuation, spelling, full sentences are not important, orphans are OK too. Just
write as fast as possible. A further five minutes is allowed for finishing off ideas and
adding other details. Then coloured highlighters are used to identify each idea and
their associated details. The second draft happens at a much more normal pace. It is
particularly useful for the five-paragraph essay form. Personal editing and multiple
revisions plays a major part in our process although we observe that our more
advanced writers can, and do, make effective authorial decisions as they produce
the first draft while the less able are left with a less then helpful first draft because
there is so much still to do in order to make it an effective piece of writing.
Modelling, planning and revision are proven to be highly effective ways for better
writers to further develop their writing but the processes for doing so need to be
taught, as do the genre structures that mediate the writing (Hoogeveen 2012, p.
129-130). We also feel this approach ameliorated, for the less able writers, many of
the issues described by Deane et al. (2008, p. 35-38). We required each author to
demonstrate the process of planning, production and reviewing. Kellogg (2008,
p.17) discusses the difference between learning by doing and learning by
37
observation and suggests that writing training adopts a blended approach where
both methods are present in good measure.
The point to be made here is that learning by observing can be combined
effectively with learning by doing at a different step of the training regimen.
The writing intervention data from Zimmerman & Kitsantas (2002) show
that observational learning from a model can produce large effect sizes.
However, when feedback is provided then the effect of practice can combine
with the effect of the model to yield truly impressive gains in writing skill.
(Kellogg 2008, p. 20).
In the present experiment we have broken down the process of narrative writing
into steps, where by continuously revising the details of character, setting (juniors
and seniors), event and plot (seniors only), we hope our students are committing
this ‘topic knowledge’ to their long-term memory where they may be accessed as
required. Thus planning is compartmentalised. The production process begins from
a position of deep topic knowledge and reviewing becomes an integrated feedback
process where our students are sense-making for themselves every decision they
make about their character, their setting and the actions that will or might occur.
However, not all students have made the progress we might have hoped for but
given the demanding nature of writing it might take some time and practice for
them to do so.
38
What was apparent in the junior room was that many students took little or no
account of what they had written during the planning phase. This was especially
true of their setting, which disappeared from the initial draft of their story. Some
students were able to make effective revisions of their second draft, but once again
it was the better writers who responded best to any constructive criticism. A way
forward might be to not rely on completely fictional characters, places, events or
ploys during the planning, production and revision phases.
We came to understand that the greatest impact on a piece of writing was from the
provision of a real and present audience for our student writers – in the senior class
we used post-it notes and each person’s work was looked at by at least three other
people over the process of planning and production, and there were times when the
senior student’s work was looked at by the whole class. For the junior class it was
the teacher who tended to review student work, except between the first and
second draft when pairs of students were hand-picked. It was observed that that
peer-edit process was particularly effective for improving both the technical and
authorial aspects of writing. By virtue of continuous modelling both teachers wrote
for an audience.
The knowledge-telling practice of the novice writer of simply putting the words on
paper with little thought to their meaning for a reader is a telling detail for the
working teacher. One senior student was very guilty of this, with both drafts of her
story being little more than streams of consciousness. Unsurprisingly this student
39
struggled with language conventions too. The words-on-paper mind-set resulted in
some students producing large text fragments during the planning phase, indicating
that their control of the writing process wasn’t all it might have been.
Writing with the end in mind has been part of the practice of the senior class for
some years even to the point where students write down their intended last
sentence. Our focus was very much on sense-making and perhaps we did not give
this strategy quite the attention we should have in our experiment. It may have
helped some students stay focussed upon their rhetorical goals. The more successful
narratives did end with a more interesting last sentence, “Like they say, practice
makes perfect” and “James is now a very loveable person but every once in a while
he will rob a store” or, “Bzz, bzz! Oh gosh!” Looking at the students’ plans and first
and second drafts would indicate these final lines appeared earlier rather than later.
Deane et al. (2008, p. 35-38) provides five explanations for a student not achieving
the knowledge-development practice of the better writer and these do present a
challenge for us as teachers. For convenience we repeat them here: (1) Interference
effects, (2) Lack of strategic writing skills, (3) Insufficient topic specific knowledge,
(4) Weak content-reasoning and research skills, and (5) Unformed or rudimentary
rhetorical goals. This professional experiment did address the needs of students
suffering the interference effects of a lack of atomicity around text production
(providing an iPad to B2). It did attempt to address issues of control around
strategic writing skills. It did attempt to build strong topic knowledge and content
40
reasoning. It did show up weak rhetorical reasoning in the early drafts of some
stories.
Reflective practice, as described by Donald Schön (1987) expects professional
people to make decisions, to make judgement calls about how certain matters will
be dealt with. We decided upon a top-down approach to writing and, our
hypothesis-testing experiment, with its many flaws, did bring about an
improvement in the writing quality of our students when viewed through the lens of
the Australian National Curriculum. This will, of course, be in part due to the very
fact that we brought our collective attention to improving writing at our school.
Whether our intervention will have a more lasting effect will be linked to the impact
the experiment has had on our own professional learning and our theory of action.
Schön’s (1987) reflection-on-reflection-on-action step, ensures that our experiment
goes beyond mere problem solving and becomes something more meaningful. A
process that legitimises, augments and challenges our professional knowledge,
values and attitudes and leads us to further refine our own theory of action.
Conclusions
Writing involves advanced levels of thinking and makes great demands on working
and long-term memory and so the belief, “What you can say, you can write” does not
ring true. There are many cognitive challenges and to become an effective writer
involves decades of practice of planning, producing and revising writing – all the
time maintaining a focus on ideas, audience, and meaning. Include spelling,
41
grammar, punctuation, and handwriting or typing, and one soon appreciates what a
task becoming a writer is.
Teaching writing to young children as we do is arguably one of the more difficult
tasks we undertake as educators. Children need to see writing being done, they need
to write for a real and present audience, and they have to learn to plan, produce and
revise their own writing. They need to shown how texts hold together and make
sense to the writer and the reader. They need time to do all of this this well.
References
ACARA (2012), The Australian Curriculum, English Scope and Sequence: Foundation
to Year 6. http://www.australiancurriculum.edu.au/australian%20curriculum.pdf?
Type=0&s=E&e=ScopeAndSequence
Black, Jason (2015). Plot to punctuation: How settings make or break your
characters. http://www.plottopunctuation.com/blog/show/90 on Saturday, 16 May
2015
Deane, Paul., Odendahl, Nora., Quinlan, Thomas., Fowles, Mary., Welsh, Cyndi., and
Bivens-Tatum, Jennifer. (2008). Cognitive models of writing: Writing proficiency as
a complex integrated skill. Educational Testing Service. Princtown, NJ.
http://www.ets.org/Media/Research/pdf/RR-08-55.pdf
42
Farrell, Peter (in press). Life outside academia: On becoming an expert teacher-
researcher. International Journal for Transformative Research
Farrell, Peter (2015). Teacher-research and the art of the professional experiment:
Reflective practice in the practical-knowledge tradition. International Journal for
Language Studies 9(1): 1-22. https://www.academia.edu/9629761/Teacher-
research_and_the_art_of_the_professional_experiment_Reflective_practice_in_the_pr
actical-knowledge_tradition
Farrell, Peter (2014). The six Rs: Pedagogical advice for the very small school.
https://www.academia.edu/7918295/The_six_Rs_Pedagogical_advice_for_the_very_
small_school
Hattie, John. (2009). Visible learning: A synthesis of over 800 meta-analyses relating
to achievement. Routledge: London
Hoogeveen, M. (2012). Writing with peer response using genre knowledge: A
classroom intervention study. PhD Thesis, University of Twente, Enshede, The
Netherlands. http://doc.utwente.nl/84101/1/thesis_M_Hoogeveen.pdf
Kellogg, Ronald. T. (2008). Training writing skills: A cognitive development
approach. Journal of Writing Research 1(1): 1-26.
http://www.jowr.org/articles/vol1_1/JoWR_2008_vol1_nr1_Kellogg.pdf
43
Kelly, James. B. (2011). When the Instructor Writes alongside the Student.
Compendium2 4.1: 43-49. https://ojs.library.dal.ca/C2/article/view/3838
Schön, Donald. O (1987). Educating the Reflective Practitioner. Jossey-Bass higher
education series: San Francisco, USA.
Peter Farrell
Dr. Peter Farrell, EdD is the teaching-principal at Zeerust Primary School in rural
Victoria, Australia. He has worked at this very small school for 13 years. Peter was
awarded his doctorate in 2009 by La Trobe University. Peter’s current research
interests are professional experimentation and reflective practice.
Glenda Telford
Glenda Telford has been a classroom teacher at Zeerust Primary School for more
than 25 years where she specialises in teaching in the early years of primary school.
Glenda’s professional interests include teaching and learning literacy and reflective
practice.
44