I chose recursive feedback using Google Drive as my capstone project because 1) the technology-driven process seemed more efficient than an old-fashioned paper swap, and 2) peer feedback provides formative assessment data that benefit both teachers and learners.

Some background information: I teach third graders, some of whom are struggling readers and writers. I have only just begun to explore Google Apps, but I thought it would be a valuable tool to enhance the classroom environment, where collaboration, discussion and reflection are valued and expected. My students are always eager to interact and help one another, and they enjoy using technology.

Recursive feedback is an "affordance" from Dr. Bill Cope and Mary Kalantzis’s work “Towards a New Learning: The 'Scholar" Social Knowledge Workspace, in Theory and Practice.” This affordance is a quality in instruction that is enhanced by technology.

Recursive feedback is a type for formative assessment, allowing students to continually learn and show growth. Summative assessment, on the other hand, is usually an end-result of learning. With summative assessments, learning ends at a final exam or project. But with formative assessment, learners receive feedback on how to further their learning. Recursive feedback allows this, but includes all sources in the classroom, not just the teacher's review (see the diagram below).

“Teachers and learners are all assessing learning, and every one of their perspectives is of some value. In fact, as their perspectives vary, the feedback may be more extensive, more thought-provoking, more rapidly provided and thus more valuable than the most assiduous of lone teacher-markers.” (Cope and Kalantzis, 2013)

Feedback is a vital part of the learning environment. By implementing peer review, feedback becomes more timely, less intimidating, and more engaging. Students know their peers will see and comment on their work and that they are expected to improve. Students giving feedback are learning as well. They see others’ work and hone their own skills as they support fellow students.

According to Cope and Kalantzis, “A culture of mutually supportive constructive feedback not only models the ideals of a knowledge economy where teamwork and networked collaborations are more valuable than ever. Assistance helps the stronger as well as the weaker. It sets community standards, where the weaker see models in the works they review that are stronger.” (Cope and Kalantzis, 2013)

My objectives were to 1) enable students to provide meaningful feedback in a timely manner, and 2) encourage students to use feedback to improve their work.

In this intervention, students were expected to provide three types of feedback to student writing, particularly from writer's workshop pieces. The feedback included:

• compliments
• questions
• suggestions

The goal was to ensure student feedback be relevant and useful. Feedback must be detailed enough so the author knows exactly what the work’s strengths are and what needs to be changed or added.

Feedback could fall into one or more of the following categories:

• grammar, spelling and conventions
• content
• word choice
• other (visuals, layout, and so on)

The second goal was for students to analyze the feedback given, determine if it is useful, and make revisions to their work based on feedback. As time progresses, student feedback should become more detailed, precise, and useful.

“Imagine a learning environment in which every task and every piece of student work is undertaken for; and anticipating feedback from, multiple sources: teacher, peers, parents, experts, critical friends and self (thinking aloud in moments of self-reflection and self-assessment).” (Cope and Kalantzis, 2008)

These objectives fall in line with those of formative assessment. The National Council of Teachers of Mathematics define formative assessment as a way to help "students develop a clear understanding of learning targets and receive feedback that helps them to improve." (July 2013)

Asking strategic questions, providing students with immediate feedback, and engaging students in self-reflection are other aspects of formative assessment that are reflected in recursive feedback and this project.

Furthermore, the NCTM cited research that supports formative assessment in mathematics classrooms directly correlates with improvement in student achievement. I believe formative assessment is helpful in all classrooms whether the subject be writing, social studies, or math.

Peer revision has never been a perfect process in my instruction, especially with writer’s workshop. Peer review in a traditional paper trade can be time-consuming and difficult to manage. Lost papers, messy notes, scheduling, and noise level can become issues.

Google Drive solved many of the issues of traditional paper-traded peer review. Students either typed or uploaded their writing into Google Drive, then shared it with me and at least two other classmates.

Students participated in this review process twice. In the first session, I randomly assigned peer reviewers. In the second session, students were matched by ability (high performing students were matched with average students; average students were matched with low performing students). It was difficult to decipher if the assignments made any difference - mostly because I changed too many variables the second review session. It then became too difficult to compare some specific needs and successes. This will be explained later in the paper.

However, in both situations, the sharing process was fast and efficient. There was no need for a copy machine, nor worries that the student work would get lost.

Once work was shared, students could easily insert comments into their classmate’s work.

Here is an example (comments are on the right):

Google Drive allows users to share work and insert comments

Thanks to technology, no papers or comments were lost during both review sessions. Students could refer to comments as many times as they wished. I could easily access student papers to comment on student comments. The dialogue between the writer, reviewer, and me was open, organized, and accessible.

In the first review session, students drafted and typed their narratives straight into Google using notes from a field trip.

During the second review session, I scanned handwritten student papers as .PDF files on to our network. The class uploaded the work into their Google accounts. From there, students could still share and comment on each other’s writing.

I was very excited about the ability to scan and upload handwritten papers. Third graders at my school are not proficient typers. They are much faster at handwriting their thoughts than typing them. Conventions and capitalization are also better on paper than on the computer, since students are still learning to use the “shift” functions.

However, in hindsight, revising handwritten papers is much more complicated than revising work on a computer. With Google, students can easily add and delete without the need of an eraser or extra paper. Google’s history function also keeps a good record of revisions.

Here’s a sample of revised work from our first review session:

Google's revision history allows students to edit their own work

Yet, when student uploaded the handwritten .PDF files, they could not go back and revise the .PDF. The .PDF was like an image file.

Instead student had to literally cut and paste their handwritten work to show changes. Students used erasers, scissors, glue, and pencils as they sat in front of a desktop computer and revised.

Here’s a sample of handwritten revision:

Handwritten revision with scissors and glue

Yes, students could have skipped lines on their notebook paper. They could have used Post Its or drawn arrows to insert new details. But in the past, both of these processes have been messy and inefficient. Plus, the students enjoy performing “surgery” on their papers. However, revising this way is slow - too slow.

It would have been worth our time to type the narratives into Google before comments and revision occurred. Our Frankenstein cut-and-paste papers don’t make a clean final copy like Google does.

So, my haste to upload work did not pan out for us in the long run.

In both review sessions, students were confused about how to write effective comments, even after repeated modeling. To begin, students shared their work with the class on the Elmo and I verbally gave three comments (a compliment, question, suggestion). After a few papers, I let students verbally give comments. We discussed which comments were effective and which were not so helpful.

Only 57% of initial comments were specific enough to be helpful. Other comments were either too vague to be effective or blank.

See the data below:

I judged effectiveness by whether the comment was specific to the paper.

Here is an example of effective comments:

Rylee complimented the writer's use of specific details such as pumpkin seeds; and she asked a specific question about the bakery food (she highlighted the area of the text that could be elaborated). Finally, Rylee asked for more information about the bakery - a specific detail from our field trip that day.

However, some students left vague comments. These comments did not help the writer craft new details.

The assignment was to write about our field trip. When a reviewer asked "tell me more about the field trip", I chimed in by asking "which part of the field trip do you want the writer to discuss?" If a reviewer asked for more details, I asked which details do you want to know more about? Some students heeded my advice; while others didn't.

Here is a sample of vague comments:

The review process took us two days to complete because I stopped the class intermittently to share student comments - highlighting specific vs. vague comments to the whole class.

The work in this first review session was very short, about a paragraph long. But the work in the second session was considerable longer. One student even wrote four pages of narrative. The comments this time were less effective than the first round - only 43% effective as opposed to 57% in the first round.

It took some trial and error before I realized that in the beginning, I’d only modeled effective and ineffective comments. I did not model how to derive comments. I realized this was a reading comprehension issue.

Most children read the work once, and then struggled with a compliment, question, and suggestion. I modeled how to read a paper completely, then how reread the text, stopping and thinking during the re-read.

This simple explicit instruction helped, but I also did something new this time around. I went back and left comments on student comments. (I am Kristyn (Giles) Shelly, the instructor.)

Commenting on student comments took considerable time, but it was worth it. Effective comments rose from 43% to 71% after students revised their own feedback. I know this because I went back and rescored all the comments.

I also provided students with more structured responses - a suggestion from a Scholar reviewer. Together, the students and I wrote up sample comments that they could copy and complete on their own. These comments were derived from the minilessons already taught in class about writing strategies. Here are samples of student prompts:

• You really painted a picture when you wrote about....
  
• I liked when you used the spicy word...
  
• Your writing made me feel... becuase...
  
• Could you please tell me more about ... (which part)...?
  
• Try using new words to start your sentences like first, then, next, or last.
  

Once comments were made, student writers met face-to-face with their reviewers. This seemed to make a considerable impact on the writing, as I noticed students working collaboratively to enhance the work. It seemed like the writing belonged to both the writer and the reviewers, as each had an invested interest in making the original piece stronger.

In the first session, 79% of the students significantly enhanced their work after peer review. By enhanced work, I mean students added more details, specifically addressed the questions peers asked them, or corrected specific mistakes pointed out by peers (spelling, capitalization, etc).

A student survey also showed 72% of children found positive benefits in peer review. Each child completed this Google Form survey. Here are the results (scroll right):

This success was unexpected, and I contribute it partially to the face-to-face interactions. Initially I’d added the face-to-face component because a Scholar reviewer suggested it. I decided to try it since my students are not proficient typers, readers, or spellers. I knew some confusion could exist without literacy proficiency. The co-ownership was an unexpected benefit.

These points became crucial in the second peer review session, when we ran out of time and I took the face-to-face aspect out of our lesson. Students struggled with the comments left by students. I found myself personally assisting 7 of 18 students (almost 40% of the class) by interpreting the comments left by peers. In the first session, I did not help any students with their revisions.

Survey results showed students did not find feedback as helpful as they did in the first round. Only 33% of students said feedback helped them make their changes; while 47% said the feedback was not helpful and they made their revisions on their own (again scroll right).

Note: I did change the survey variables from the first questionnaire. I knew every child but one made 2-3 revisions. I knew this because I personally helped 7 children with revisions. Had I not intervened, these revisions would not have happened. So I need to add something about “independence” to this second survey. Otherwise, it would appear as if almost 100 percent of the class made significant revisions - when in actuality, only 55% of the class made independent revisions (as opposed to 71% in the first round - the round that included face-to-face interactions between writers and reviewers). Also, not all children took this second survey. Five students were pulled from class for other programs.

Here’s the data breakdown: The confusion with revisions in this 2nd session lay both on the reviewer, who may have lacked the spelling or grammar skills to write comprehensible comments, and the writer, who may have lacked the reading skills to understand what the reviewer meant.

The face-to-face contact helped eliminate confusion, and left the third graders more independent about their comments and revision. Plus, revising work with a partner seemed more engaging than revising alone.

Despite the technological success of Google Drive, the students are still too inexperienced with written communication to leave out face-to-face conversations. This could change as the year progresses, as many of my transitional readers and writers become more proficient.

In either case, Google was a great starting point for face-to-face discussions. Both reviewers and writers could take their own personal time to process information without the pressure of someone waiting. And again, there is a written record of the discussion - a valuable aspect for my own evaluation purposes.

Implementing recursive feedback has transformed my instruction. But I didn’t see the greatest benefits until recently, due in large part to computer technologies.

In the past, I’ve used peer review with rubrics, Post Its, and writer’s workshop. But the process was always a difficult organizational struggle.

Then, after my first New Learning course last spring, I started experimenting with recursive feedback using an online platform called Kidblog. Kidblog allowed my students to answer short essay questions online related to science or social studies. We participated in the same review process described here in this paper - only we did not extend our writing to longer personal narratives.

Google seemed like a better option for using recursive feedback for student writing. We could write and review a lot more content on Google than we could on Kidblog. Plus, the revision history provided a nice paper trail on how much students changed their original work.

Despite the online platform, the process of recursive feedback created a shift in power from teacher-directed to student-centered learning. Students began to look to each other for help and support. This transformation was empowering, allowing the students to direct feedback and outcomes.

Cope writes about this shift from a teacher’s authoritarian stance to an environment where “learners become teachers of each other.” (2008)

“Traditional teaching discouraged lateral communication between students. The New Learning thrives on enormous amounts of lateral communication between learners: face-to-face talk; Internet and mobile telephony; online chat; instant messaging; email and messaging around text, image, sound and video in content creation and sharing environments.” (Cope, 2008)

New technologies have helped enable this kind of lateral communication in my classroom. It empowers students by allowing them to collaborate and learn from each other.

Finally, I never commented directly on student work - only on student comments. I did not want students to copy my compliments, questions, or suggestions. I also don't want my voice to come across as more important than the kids' voices. This helped shift dependence from me to the students.

However, before printing final papers, I did sit down with each student and make some last revisions - mostly conventional ones like adding quotation marks (those have not been taught yet) and fixing spelling errors that neither the writer or reviewers could correct. This is third grade, and there are many opportunities to teach.

So, I still played the role of direct teacher but the peer review made my role a whole lot easier than without it.

The process of peer feedback, even with Google Drive, required a dedicated investment in teaching students about formative assessment, the writing process, and reading/writing strategies. It was not an easy process, but it proved to be a successful one for both me (the instructor) and the students.

I’ve definitely learned some lessons. First, revising work on a computer is easier than doing so on paper. Second, the face-to-face contact was an important follow-up. Third, not only can students provide effective feedback on their peers’ work, but they can also provide feedback on comments.

I played a rather authoritative role deciding which comments were effective and which were vague. As students become better at providing feedback, they can also provide feedback on feedback. Furthermore, they can let their reviewers know when comments helped or did not help them and why.

Finally, I need to spend more time on writing strategies. Students are still unclear about what makes good writing. They need more explicit exposure to quality literature and more instruction on the traits of writing. This takes time, usually half a school year.

In the past, I usually spent the first semester modeling feedback through small-group writing conferences. I used checklists and rubrics that constantly reflect our newest writing strategies.

It isn’t until the second semester that I release feedback to the students, letting them take on my role as reviewer.

For the purpose of this class, I nudged students into a review role before most were ready. But overall, the children enjoyed the task and loved sharing their work. The process was a good step toward where we want to go this year in writing.

* Parts of this paper were written collaboratively with Scholar classmate, Kimberly Martin, for Work 1.

Cope, Bill and Kalantzis, Mary. 2013. Towards a New Learning: The ‘Scholar’ Social Knowledge Workspace, in Theory and Practice.

Cope, Bill and Kalantzis, Mary. 2013. [Image of recursive feedback]. Towards a New Learning: The ‘Scholar’ Social Knowledge Workspace, in Theory and Practice.

Cope, Bill and Kalantzis, Mary. 2008. New Learning Elements of Science of Education. Cambridge University Press.

National Council of Teachers of Mathematics. 2013. Formative Assessment: A position of the National Council of Teachers of Mathematics. Retrieved from http://www.nctm.org/uploadedFiles/About_NCTM/Position_Statements/Formative%20Assessment1.pdf#search=%22formative%20assessment%22.