Making paperless work.

Full text

Page 1

1
Permission to make digital or hard copies of all or part of
this work for personal or classroom use is granted without
fee provided that copies are not made or distributed for
profit or commercial advantage, and that copies bear
this notice and the full citation on the first page. To copy
otherwise, to republish, to post on servers or to redistribute
to lists, requires prior specific permission and/or a fee.
CHINZ 2007, July 2-4, 2007, Hamilton, New Zealand
© 2007 ACM 978-1-59593-836-7/07/0007...$5.00
Making Paperless Work
Beryl Plimmer
Department of Computer Science
University of Auckland
New Zealand
beryl@cs.auckland.ac.nz
Mark Apperley
Department of Computer Science
University of Waikato
New Zealand
m.apperley@cs.waikato.ac.nz
ABSTRACT
Despite well documented advantages, attempts to go truly
"paperless" seldom succeed. This is principally because
computer-based paperless systems typically do not support
all of the affordances of paper, nor the work process that
have evolved with paper-based systems. We suggest that
attention to users' work environments, activities and
practices are critical to the success of paperless systems.
This paper describes the development and effective
utilization of a software tool for the paperless marking of
student assignments which does not require users to
compromise on established best practice. It includes a
significant advance in the task management support.
Author Keywords
Paperless environments, affordances of paper, pen
computing, annotation.
ACM Classification Keywords
H5.2. User Interfaces, Input devices and strategies.
INTRODUCTION
There are many good reasons why we might aspire to
paperless environments, including reduction of the
environmental harm of paper consumption and the
economic cost of paper production, transfer and storage.
Digital environments free us from the location and physical
constraints of paper and provide better support for updating,
filing, and comprehensive searching of documents. Yet,
rather than decreasing, paper use is increasing in parallel
with the increased use of computer systems.
Sellen and Harper [17, p76] suggest that there are four
affordances of paper that are absent from most digital
systems: flexible navigation; cross-referencing of multiple-
documents; easy document annotation; and the modeless
interleaving of reading and writing. However, they propose
(p48) that for paperless environments to succeed, human
systems must adapt to match the computer system. They
evidence this with DanTech’s relatively successful
reduction in paper usage which they attribute to changes in
work practices. We agree with Sellen and Harper that
‘going paperless’, as the two organizations they studied
attempted, is doomed to failure in the short term. We
believe this is because there is insufficient understanding of
how paper supports various activities for that support to be
replicated in computer software: additionally ideal
computer hardware is not yet available. We argue that long
term success will depend on computer systems better
supporting human systems, and we demonstrate that this is
possible.
There are two research challenges which must be met for
such computer systems to emerge. First, we need to fully
understand the role of paper documents in particular work
practices [8]; this suggests an activity-centered model of
analysis and design [9]. Second, technology must be
developed (both hardware and software) to support
equivalent paperless environments.
Clearly both of these are complex issues. Here we present a
successful exemplar, Penmarked, a paperless environment
for managing, annotating and grading student assignments.
We suggest that our experiences will be of assistance to
others developing paperless environments.
BACKGROUND
The use of paper is deeply integrated within our work
practices; to convert tasks from paper-based to paperless we
must understand papers essential qualities and uses. Some
of paper’s affordances need to be preserved in paperless
equivalents, yet virtual paper can overcome the physical
limitations of paper.
The materiality of paper documents is useful in a number of
respects [8]. The physical dimensions and binding of a
document convey meaning; a hard-cover book is very
different in presentation from a hand-written note. The
spatial position of documents on a desk may be significant,
as may the order of the documents within a stack.

Page 2

2
Documents left on a work surface act as a reminder for an
unfinished task. Document handover may signify a social
process (eg the passing on of responsibility, completion, or
collaboration) [6, 17].
However paper’s predominant affordance for document
storage is as an ink repository [17]; it represents an object
from which we can read and upon which we can write,
often interleaving these two activities. People prefer reading
from paper to reading ‘on screen’. A widely held belief is
that this is because of the quality of screens and the fact that
the projected light they emit makes reading more difficult
than reading from paper. Yet, there are other characteristics
of paper that are important. It is easier to quickly get an
overview of the layout and content of a paper document by
flipping through the pages [8]. Active reading includes
activities such as scanning from the current point of focus
backwards and forwards in the current document and
between related documents. We often spread documents
around our work surface and use spatial and tactile clues as
markers to particular parts of the documents. For example
we can arrange documents in a mind-map configuration or
temporarily mark a passage by placing a pen on it.
Active reading includes annotating of the document or the
writing of summary notes in a new document [17]; the act
of annotation is a very powerful active reading and learning
mechanism [21], and the resultant annotated document is a
different artifact from the original document [18].
Although annotation techniques are both personal and
idiosyncratic, annotations can be broadly categorized as
emphases – achieved by highlighting, underlying, circling
or side bars; corrections – crossing-out, overwriting, and
inserting; or commenting – adding words or glyphs in the
margins to explain or summarize the document [7].
Subsequent readers of the annotated document (including
the annotator and the author) have a different reading
experience [20]. They concurrently see the original work
and the superimposed annotations. The annotations
embellish the original document guiding the reader to
points that the annotator considered significant.
There is a noteworthy difference between ink-over
annotations on paper and the types of reviewing supported
in word processors. With ink-over annotation one feels that
the author still ‘owns’ the work as the original document
remains fixed and the annotations are merely decorations.
In contrast word processor reviewing alters both the
appearance and layout of the original document so that
authorship is now shared. Ink also affords powerful
informal and iconic annotations that are difficult to replicate
with keyboard and mouse [16].
On the other hand, the physicality of paper imposes
limitations on paper documents that are not present in
digital environments. First, the physical paper object
requires physical creation, transportation, storage and
destruction; each of these processes is significantly more
expensive and time consuming than its digital equivalent
[17]. Copying paper documents is also a physical process,
while multiple copies of a digital document can be created,
emailed and stored in digital form at a fraction of the cost.
The fixed nature of the ink on paper documents makes
substantial revision impossible, whereas editing a digital
document is simple. Paper documents are, by their very
nature, linear, while the natural structure of some
documents we may wish to represent is non-linear (for
example object-oriented programs). A digital environment
better supports non-linear navigation, and the indexing and
searching of documents and document repositories.
However, electronic environments do impose different
constraints, such as the requirement for suitable hardware,
reliable power sources and network connectivity. Much of
this technology is currently available in the office, but is not
yet as mobile and ‘always on’ as paper. Marshall [8]
attributes the failure of e-books to the need for specialized
devices, and suggests that to be successful, this technology
must be supported by appropriate software on general
computers.
In fact, although we think of paper documents being a more
permanent record of a document than their digital
equivalent, a single copy of a paper document is more
expensive and far more fragile than the multiple copies that
usually exist of digital documents. Consider for example
the various versions and revisions of this paper; the cat
chewed hand-written notes left on the floor, but there are
probably 4-6 digital copies of it, on our hard drives, mail
servers and backups.
Assignment marking
Our goal with this research is to explore the requirements
for the complete and effective replacement of a paper-based
document system. We do this through the development of
an exemplar, a computer-based environment for marking
student assignments. Traditionally assignment marking is a
paper-based activity. The student completes his or her
assignment and submits the script to the teacher. The
marker (teacher or teaching assistant (TA)), reads,
annotates, ranks and grades the papers, records the grade in
a grade-book, and returns the script to the student. Thus,
paper is employed for a range of common reading, writing
and document management tasks.
Below we provide a detailed description of assignment
marking. We contend that an in-depth understanding of the
paper-based activity is an essential precursor to the design
of a paperless environment. Our approach is a combination
of autoethnography [1] and a review of earlier work.
Traditionally students handed their paper ‘scripts’ to the
teacher. In many cases this has already been replaced by
electronic submission [for example, 5, 13, 14]. Electronic
submission is easier for the student, and alleviates problems
of misplaced scripts and the time-stamping of submission.
Further, when the assignment is a computer program or
similar, a digital copy of the work is required for grading
purposes.

Page 3

3
The teacher may quickly scan the entire collection of scripts
before commencing grading. A survey of colleagues
suggested a number of preordering techniques they employ.
Sometimes they will select the scripts that they expect to be
best and/or worst and inspect, but not necessarily grade,
these first. On other occasions they may arrange the scripts
in a ranked order before assigning grades, or if the
assignment topics vary then the scripts may be grouped by
topic before grading commences. Both scanning and sorting
document collections are facilitated by paper scripts, but
can be difficult in standard digital environments.
Our observations of teaching assistants (TAs) suggest that
they take a more linear approach. In our departments the
teacher provides a detailed marking rubric for the TAs,
who, we have observed, start at the top of the stack and
mark each assignment in turn against that rubric. They
rarely compare scripts or undertake any scanning or pre-
ordering, overview activities in which they have little
interest.
The next and most important step in the marking process is
a careful examination of each student’s work. Traditionally
the marker annotated the work with a red pen. Such
annotations serve three purposes. First, they act as a
reminder to the marker of the critical points in the
assignment; it is common practice to use annotations as a
part of the grade decision process. Second, the ink records
the summative grade assigned to the work. Finally, and
most importantly, annotations provide feedback to the
student.
Individualized formative feedback on a student’s work is
acknowledged as a powerful learning intervention [14]. All
manner of electronic grading systems have been proposed
that can produce a fair summative grade [for example, 5].
The most sophisticated may generate specific feedback
depending on the errors detected; however, the
personalization is lost. The ability to easily annotate a script
is a capability that is needed in order to achieve a successful
paperless system in this application.
To move from one script to the next when working with
paper is simply a matter of lifting the next script off the
stack. Many paperless environments require significantly
more effort, including the closing and opening of folders,
applications and documents.
Once a grade has been determined for a script, the next step
in the marking process is the recording of the student’s
grade both on the script and in a grade-book. The modern
trend is to grade against a set of criteria and allocate a score
to each criterion. The grade is then derived by calculating a
weighted average of the criteria. A paper-based system
requires manual calculation of the grade and its double
entry on the script and in the grade-book. In contrast, with a
computer-based system, both calculating the grade and
producing multiple copies of it is trivial.
The final step is return of the script to the student. The
options for return of paper assignments are governed by the
physicality of paper; for example the assignments can be
distributed in class, held somewhere for collection or (snail)
mailed to the students. In contrast, electronic systems can
accomplish this automatically, either by email or placement
of the graded assignment into a digital repository.
Furthermore, with paper, there is only one copy of the
assignment; if it is lost, it is lost forever, or if the teacher is
required to retain a copy (for example, for moderation) then
it must be photocopied.
From this description of marking as an activity it can be
seen that the challenges in creating an electronic paperless
system for this application are support for work tasks such
as moving effortlessly between scripts, freehand annotation
of scripts, and simultaneous recording of scores. Our
objective is to provide both faculty and students with ‘the
best of both worlds’. This is to say, an electronic
environment that provides rapid, easy storage, transmission
of documents and recording of grades, and at the same time
an informal paper-like environment that affords quick
commenting and rich feedback.
DESIGN STRATEGIES
As computer scientists, most of the assignments our
students write are computer programs. For us, detailed and
in situ comments are as important in marking programs as
in the case of more traditional essay-style assignments.
Our initial focus was on supporting the marking of
programming assignments; this focus has both pros and
cons. The advantages are that our users (programming
teachers and TAs) are expert computer users accustomed to,
and unfazed by, new technology. Also as programming is
predominately a computer-based activity, the markers
expect to use a computer for at least part of the grading
process. The disadvantages are three-fold; program code is
often saved in multiple files (one per program class);
programs usually need to be compiled and executed as a
part of the marking process; and programs may have a
complex non-linear structure, in comparison with essays,
which are typically read in a linear fashion. However, we
also intend this prototype to be useful for grading more
traditional assignments such as essays.
Norman’s [9] activity-centered model of an activity, task,
action, hierarchy is used as a framework for this design
discussion. The activity that we are concerned with is the
marking of a set of student assignments, with special
consideration for marking programming assignments. The
set of tasks involved in this activity are collecting
assignments, pre-marking preparation, detailed examination
and annotation of each assignment, allocation of scores,
archiving of grades for the grade-book, and return of the
scripts to the students. There are tools and techniques
readily available for each of these steps – however many
markers rely on paper for part of the process because the
connections between the parts is not well supported or the
formal nature of the computer tools limits information
transfer. More detailed technical descriptions of Penmarked

Page 4

4
are available [10, 11], here we concentrate on the support
for paperless work.
Most institutions have student support systems that include
electronic drop-box facilities. We propose a flexible
assignment collection architecture that will support either a
web-based service or a customizable interface to an
electronic drop-box. Data such as student identifiers,
contact information and assignment file information, which
is required for marking, is customarily available from such
facilities. Essay assignments are usually submitted as a
single file, while computer programs are frequently
submitted as a zipped package of multiple files. In this
latter case, only selected files may be of interest to the
marker. A general marking system must deal with both
these situations, ie single and multiple file scripts.
Pre-marking activities include the preparation of a marking
rubric and pre-scanning and ordering of scripts. A wide
range of marking rubrics may be employed, from a simple
alphabetic grade (A+) to complex multi-criteria Likert
scales. These are easily accommodated in an electronic
system, although, as our intention is to support pen-based
interaction for annotation, recognition of handwritten scores
is desirable.
Providing an interface that makes it as easy to pre-scan
multiple scripts as it is to scan paper copies is a challenge
with current technology [8]. A large wall or tabletop display
could show multiple scripts on the screen in such a manner
that the marker could scan and sort them; interaction
techniques for flicking through digital documents have been
explored for use with digital books [16]. However, to
realize this design as a prototype, we restricted ourselves to
working with standard hardware such as a desk-top, lap-top
or Tablet PC. A viable software approach with such
environments would be to list all the scripts, and to allow
quick inspection and ordering by clicking and dragging
actions on this list.
The detailed examination of each script relates directly to
Sellen and Harper’s [17] notion of active reading. The
reader needs to be able both to read and annotate
simultaneously; written feedback is a vital ingredient of
assessment. Some have proposed off-line comments,
however these are clumsy and time-consuming to construct
because a point of reference must be established (e.g. on
page 3, paragraph 2...). Heinrich and Lawn [3] propose
typed annotations that lie over the document. While this
places the comments in situ, typed comments lack the
flexibility and expressiveness of ink. We suggest direct
inking on to the document on a Tablet PC monitor [16].
Grading computer programs calls for particular software
functionality [14], including provision for multiple file
scripts, and navigating between these, as well as provision
for program compilation and execution. The marker needs
access to the program code in both the marking software
(which supports annotation and the recording of scores) and
the programming environment (where they can execute the
program); our solution to this is the use of two monitors: a
tablet for annotating the script and a standard monitor to
display the programming environment (Figure 1).
Figure 1. Penmarked in use with dual monitor setup.
Rating the assignment against criteria is integrated with
active reading and feedback annotation. Scores must be
assigned to the work and recorded for use in a grade-book.
The ideal is for the marker to write scores once only.
Current recognition techniques could not reliably
disambiguate a score a written anywhere on the document
from other annotations. We considered three possibilities to
achieve score recognition. First, the marker changes ink
modes to write the score – this was discarded because
modal inking has been shown to be less than satisfactory for
users [12]. Second, a score zone (a square in one corner or
column down one side) is provided on the document and
any annotations in this zone are recognized as scores. Last,
a separate table is provided which contains a marking
schedule, into which markers can record scores with either
a stylus or keyboard. However inking accurately requires a
larger space than a type-written score takes, so to keep the
space for a score table small we suggest a separate inking
zone adjacent to the table. The marker writes a score in the
inking zone which the system recognizes and stores in the
table. Once recognized, scores can be totaled and archived.
The final marking tasks are the saving of the grades into a
grade-book and returning the scripts to the students. As
with collection of the scripts, most institutions have existing
student databases; saving grade-book information in
standard file format supports integration with other systems.
For reliability purposes the scripts need to be saved in a
fixed format that makes them difficult to change (this may
be contrary to the requirements of many other document
annotation activities). Two alternatives for return of work
are emailing and a reverse drop-box, depending on
standards and protocols for the particular institution.

Page 5

5
PENMARKED PROTOTYPE
Figure 2. Penmarked main interface in default layout.
Penmarked (Figure 2) has been developed for use on a
tablet PC running the Microsoft XP Tablet™ operating
system and the Visual Studio .Net Framework™. To
meet the design goals described above, the major focus
of the prototype development has been ensuring that all
stages of the assignment marking activity are
appropriately supported. The key components of the
system are: management of pre- and post-marking
activities for a class set of assignments; on screen
annotation of scripts and concurrent recording of scores;
and, for marking computer programs, easy transition
between Penmarked and a programming environment.
Pre-Marking Activities
Penmarked support starts with the setting up of an
assessment, with a wizard to step the user through this
process. First the user defines the marking schedule
(rubric); they may load an existing schedule and modify
it, or create an entirely new schedule. The schedule
specifies the assessment criteria and minimum and
maximum score for each criterion.
Next the user specifies the location of the student files.
The data that the system requires are a student identifier
and email address (for return of work), and the name(s)
and the location of submitted files. As the format of
electronic drop-boxes varies
implemented two plug-ins. One is tailored to a locally
developed drop-box and another based on XML files. In
either case, the specified directory structure is parsed
and information gathered to add each assignment to the
student list for display.
widely we have
The student list has become progressively more complex
with each prototype iteration. Initially it was a simple
list in alphabetic order. We then added check boxes so
the marker could check-off assignments as they are

Page 6

6
completed. Later we added a change of colour and font
for assignments that had been opened but not completed.
More recently teachers (rather than TAs) have used
Penmarked. In one case the students had a choice of
essay topic; the teacher would have liked to be able to
preorder the assignments by topic. We foresee no
difficulty implementing this functionality; we report it
here to emphasize the importance of replicating all of
the affordances of a stack of papers.
Marking an assignment
To mark an assignment the marker selects the student’s
list entry. Penmarked locates the student’s submitted
files; zip files are extracted and placed into a
subdirectory and the location of the student’s files is
associated with the folder icon on Penmarked’s toolbar.
A user defined filter is used to select appropriate files
(e.g. all those with a particular file extension except
xyz). We found that file type and name filtering is
particularly important for marking .Net programs as the
environment generates a large number of files in which
the marker has little interest, but which are required to
compile and execute the program. This is an example of
how closely the paperless environment must be tailored
to the task.
Selected text and rich-text files are displayed in the
annotation pane (Figure 3). If there is more than one file,
each is displayed in a separate tab. The marker can now
read, annotate and score the script against the marking
rubric. To navigate through the script the marker can
scroll within a window using the scroll bars, or jump
between windows using tabs. A find dialogue proved to
be useful for locating specific points in the programs.
Ink annotations can be placed anywhere on the
documents or in an auxiliary panel. The ink is held on a
transparent overlay that is superimposed over each
document. Positioning between the two layers is
synchronized so that they appear as one. An ink eraser is
also provided; users have commented on this as a real
benefit of a digital environment. The underlying script
cannot be changed by the marker.
Figure 3. Assignment annotation pane
The mark schedule (Figure 4) is created for an
assignment when it is first opened in Penmarked. The
marker navigates the marking rubric using the scroll bar
and can select a criterion to be scored. The score is
entered by either writing in the input box or via the
keyboard. Ink input is recognized using the operating
system recognizer. The input is validated against the
minimum and maximum, and if valid is stored in the
mark table. If the input is invalid the user is alerted to
this fact.
Figure 4. Marking schedule showing a score in the input
box
When marking programs the source files are accessed by
selecting the folder icon on the Penmarked toolbar. This
opens the Windows folder that contains the student’s
file. From here the marker can open the programming
environment and compile and execute the program. We
noticed that markers switch between the programming
environment and Penmarked frequently, as is typical for
a multi-document exercise [17]. We have implemented
multiple display support in Penmarked, with the tablet
screen used to display the marking software so that the
marker can annotate with a pen, while the second
monitor is used to display the folder of student work and
the programming environment (Figure 1).
We carefully designed Penmarked to fully support pen-
only interaction. However, when marking programs that
are not designed for the tablet, usability testing showed
that it was much quicker to interact with student’s (non
tablet) programs with a keyboard.
Post-marking tasks
When marking is finished the marker selects the
assignments for return, and chooses the method of
return. The system generates a PDF file for each student
that contains both the completed marking rubric and the
annotated assignment documents. The PDF file is
emailed to the student via either STMP or Microsoft
Outlook™. The method of return and mail client
information is specified in the Penmarked options.
Grades and identification data are exported in XML
format to provide maximum flexibility for interfacing
with other software.
Interaction design
Penmarked has been designed with pen-only interaction
in mind [4]. Access to the frequently used functions is
via the main interface toolbar. The interface is

Page 7

7
configurable with both the student list and marking
schedule held in floating, resizable and dockable
windows. These windows can be stacked on top of each
other; in this case tabs appear for the user to navigate
between the two. This provides maximum flexibility for
users to adjust the viewing spaces and adaptation for
right or left-handedness and screen orientation.
Stylus entry of scores requires clear writing for
successful recognition; this is easier to achieve when
writing with larger gestures. In situ writing in the
marking schedule was not used because of the screen
space that would be required. The XP on-screen
keyboard was discarded because it would reduce the
screen space available to Penmarked or require frequent
opening and closing. The panel we have implemented is
a balance between screen-space, accurate recognition
and unnecessary interruptions of the user. Potentially
users could allocate marks to the wrong category or not
notice when a mark was misrecognized. We analyzed
usability testing recordings of 12 assignments being
marked by 4 different markers, carefully observing this
task. There were no incidents of markers allocating a
score to the wrong criterion. There was one incident of a
marker not seeing that a score was misrecognized. The
system has been modified to check that each marking
schedule is completed before allowing the return of the
work to the student. This is still fallible; we are
exploring other options to fail-safe this feature without
interrupting the user with ‘ok cancel’ dialogues.
Penmarked supports a range of hardware configurations.
It can be used in tablet only mode, where all interaction
is via the stylus. This works extremely well when
marking essays and provides maximum portability. It
can also be used on a tablet with a keyboard attached, as
users have found a keyboard is useful for data entry
when marking programs. Finally, it can be operated with
a tablet, keyboard and second monitor, providing the
optimum configuration for marking programming
assignments, as Penmarked and the programming
environment can be seen and interacted with
simultaneously.
Penmarked has been extensively evaluated in use. Early
in the development we undertook numerous ad hoc
small evaluations with colleagues and TAs. We ran
three formal usability tests, a think-a-loud study and
focus group [self reference] when we had completed our
first prototype, and a lab observation using Morae™
[19] on our second prototype. Penmarked has been used
for marking about one thousand student scripts, over six
different assessments. The types of assignments include
Visual Studio programs, Java programs and essays.
Twelve different markers (both teachers and TAs) have
been involved in these trials. They have reported that
they do not print anything or feel any need to print the
documents.
DISCUSSION
The Penmarked prototype addresses many of the
affordances of paper that are not available in standard
systems. First, Penmarked makes it very easy to
navigate from script to script, and between script and
programming IDE. It is also easy to move between
multiple documents that make up a single assignment
script. The ‘find’ functionality implemented in
Penmarked is an example of search that is not available
on paper.
The marker switching between Penmarked and the
programming IDE is an example of multiple document
referencing [17]. The
provides a satisfactory solution to this problem. An
alternative possibility would be to integrate the
functionality of Penmarked into a programming IDE.
two-monitor environment
The interleaving of reading, writing and erasing are fully
supported. There are two distinct forms of writing,
annotation onto the script, and the entering of scores.
We have observed markers interleaving reading, both
forms of writing, and program execution without
difficultly.
Many potential users have asked whether there is a
clipboard for comments. This contrasts with the
observations of [7] that annotations are unique and
idiosyncratic. Also, a pasted comment would require
manual positioning and resizing [2]. This could take
longer than creating a new annotation, but it an area
worthy of investigation.
Marking student scripts involves not only reading,
commenting on and grading but also pre- and post-
marking tasks. Through sustained use of this system,
and discussions with colleagues, we have learnt much
about the types of tasks that people undertake both
before and after marking a set of scripts. The student list
has undergone a number of reviews during our
prototype development and we can see potential for
improved functionality, further enhancing the paperless
system by some of the affordances offered by
computerization; for example, the ability to reorder the
list as we commented above; adding comments related
to a student; or showing the total grade once the
assignment has been marked.
We noted that there are differences between how TAs
and teachers mark assignments. Similar differences are
likely to exist with other activities depending on the
user’s role and individual preferences.
One task associated with marking computer programs is
not well supported in Penmarked, compiling and
running the program. As a foray into this we
implemented ink annotation into the Visual Studio IDE
[15] and while this shows promise as an idea, the
environment proved to be very difficult to program. We
are currently exploring other environments.

Page 8

8
CONCLUSIONS
Penmarked supports educational best practice by
providing in situ ink annotation of a student’s work and
a marking rubric against which the marker can score the
work. Through formal evaluation studies and on-going
use of Penmarked, our understanding of the affordances
of paper that are important to this activity have been
honed.
Many attempts to move to paperless environments have
been unsuccessful. Often this is because the replacement
computer systems do not support the appropriate paper
affordances required for the activity. With this research
we began by examining the differences between paper-
based and standard computer-based document activities
and then focused specifically on assignment marking as
an activity. We have described this in detail, as we assert
that it is only with a comprehensive understanding of the
tasks involved in a paper-based activity that a paperless
system can be designed to support it. It is necessary to
identify the paper affordances that are critical to the task
to those that can be better replaced with digital
equivalents.
Assignment marking is just one example of a
traditionally paper-based activity that can now become
paperless. The technology is available; by unraveling
the entanglement of paper in our work practices we can
understand which affordances of paper to retain and
which to replace. Penmarked, like all systems, is
constrained to technology of today. As larger stylus
sensitive surfaces become available the desk-top
metaphor may become a reality. We can imagine a
digital desktop where we can indeed stack documents
and spread them around, annotate and edit them, in a
totally paperless, yet natural work environment.
REFERENCES
1.
Duncan, M., Autoethnography: Critical appreciation
of an emerging art, International Journal of
Qualitative Methods, 3, 4, Article 3, (2004),
2.
Golovchinsky, G., Denoue, L., Moving markup:
repositioning freeform annotations, in proc UIST
'02, ACM, (2002), 21-30
3.
Heinrich, E., Lawn, A., Onscreen marking support
for formative assessment, in proc Ed-Media, (2004),
1985-1992
4.
Jarrett, R., Su, P., Building Tablet PC applications,
Microsoft Press, (2003),
5.
Joy, M., Luck, M., Effective Electronic Marking for
On-Line Assessment, in proc ITiCSE, ACM,
(1998), 134-138
6.
Liu, Z., Stork, D. G., Is paperless really more?,
Communications of the ACM, 43, 11, (2000), 94-97
Marshall, C., Annotation: from paper books to the
digital library, in proc DL, ACM, (1997), 131-140
Marshall, C., Reading and Interactivity in the
Digital Library: Creating an experience that
transcends paper, in proc CLIR/Kanazawa Institute
of Technology Roundtable, (2003), 5.4.1-20
Norman, D., Human-Centered design considered
harmful, Interactions, (2005), 14-19
10. Plimmer, B., Mason, P., Designing an Environment
for Annotating and Grading Student Assignments,
in proc OZCHI, (2004), 45-53
11. Plimmer, B., Mason, P., A Pen-based Paperless
Environment for Annotating and Marking Student
Assignments, in proc AUIC, CRPIT, (2006), 27-34
12. Plimmer, B. E., Apperley, M., Software for
Students to Sketch Interface Designs, in proc
Interact, (2003), 73-80
13. Preston, J. A., Shackelford, R., Improving on-line
assessment: an investigation of existing marking
methodologies, in proc ITiCSE, ACM, (1999), 29-
32
14. Price, B., Petre, M., Teaching programming through
paperless assignments: an empirical evaluation of
instructor feedback, in proc ITiCSE, ACM, (1997),
94-99
15. Priest, R., Plimmer, B., RCA: Experiences with an
IDE Annotation Tool, in proc CHINZ, ACM,
(2006), 53-61
16. Schilit, B. N., Golovchinsky, G., Price, M. N.,
Beyond Paper: Supporting active reading with free
form digital ink annotations, in proc CHI 98, ACM,
(1998), 249-256
17. Sellen, A. J., Harper, R. H. R., The myth of the
paperless office, MIT, (2002),
18. Shipman, F., Price, M., Marshall, C., Golovchinsky,
G., Identifying useful passages in documents based
on annotation patterns, in proc ECDL, (2003), 101-
112
19. TechSmith, Morae, 2004, (2005),
20. Wolfe, J. L., Effects of annotations on student
readers and writers, in proc Digital Libraries, ACM,
(2000), 19-26
21. Yang, S. J. H., Chen, I. Y. L., Shao, N. W. Y.,
Ontology enabled annotation and knowledge
management for collaborative learning in a virtual
learning community, Educational Technology &
Society, 7, 4, (2004), 71-81
7.
8.
9.