Brandeis University



IIM 98b: Capstone Project

“Learning Science and Skill Acquisition”

Summary Report

Nathan Hakimi

May 13, 2011

Under the Supervision of Prof. Rick Alterman

I. Introduction

This report summarizes the outcomes of my research experience, independent study, and capstone project on the topic “Learning Science and Skill Acquisition”, under the supervision of Professor Rick Alterman.

As a capstone project, this experience concludes my undergraduate work in the field of Cognitive Science. This major was created under the Independent Interdisciplinary Major program, with a curriculum designed by myself and my advisors, with the stated purpose of exploring consciousness, intelligence, and computation, including coursework from Neuroscience, Philosophy, Computer Science, and Psychology.

Several components of that curriculum informed the work I conducted this semester: perception, motor articulation, attention, and memory in Behavioral and Cognitive Neuroscience; agents, systems, and environments in Informatics 2D: Reasoning and Agents; scaffolding and learner support in Adaptive Learning Environments; cognitive models in Advanced Topics in Cognition. Most directly, this work was predicated upon subject matter from a course taken first semester senior year, with Prof. Alterman, called Computational Cognitive Science.

In that class, we learned about the processes that underlie intelligence in terms of activity, context, and interaction – a distributed view of cognition. Among the learning goals for the course were to gain understanding of tools, artifacts, language, embodiment, culture, scripts, and environments as they relate to the nature of human cognition and activity.

From this perspective, the use of tools such as computer applications is a worthy topic of research. This semester, I participated in a research project on skill acquisition for the use of computer applications. The focus was on developing a training method the supports the users skill acquisition vis-a-vis the use of an application. This a problem of cognitive engineering that requires insight and expertise with regard to human cognition.

A critical issue in approaches to training is the Active User Paradox, which relates to users’ tendency to sacrifice long-term efficiency for short-term goals (see Section III). In class, Prof. Alterman presented a program he and his PhD supervisee Brian Krisler had developed called HotKeyCoach (HKS), which attempts to resolve the active user paradox by inserting learning events and training “in the flow” of activity [1].

One published study had already been completed on HKC [1]. There it was established that HKC successfully helps users develop efficient motor routines for software operation, by introducing learning events to reinforce keyboard shortcuts.

However, HotKeyCoach also has a second major functionality, which assists users in developing conceptual understanding of software features, and in learning visual-motor routines for locating relevant items in the software’s menu structure. It does this by introducing “cribsheets” explaining the major concepts involved in an application and highlighting the location of relevant menu items. My role this semester was to guide development and testing of this second functionality of the training tool HotKeyCoach.

My involvement was threefold: I scripted the trainer to support users in a specific Mac OS X application, I participated in the design of a pilot study to test the trainer, and I collected data for the pilot study.

Throughout the semester, Professor Alterman, his PhD student Brian, and myself met weekly. During those meetings we not only managed the software implementation, design, and data-collection, but also engaged in discussion of practical and theoretical issues. Many ideas related to skill acquisition, embodiment, concept learning, the learning sciences, and so on. In this way, this experience constituted both a hands-on research experience, and an academically and intellectually relevant independent study.

Ultimately, we successfully developed a version of HotKeyCoach which trains users in Preview, the Mac OS X image and pdf-reader application, designed an experiment to test how well it trains users’ conceptual and embodied understanding of the software, and collected data which yielded promising results and will serve as the foundation for later development and testing.

II. Background

The theoretical underpinnings for this work derive from the fields of Learning Science, Cognitive Science, Software Design, and Human-Computer Interaction. It concerns the way computer users learn, understand and use software. Literature in these fields is extensive, encompassing many facets of the seemingly simple interaction which occurs between an agent, an artifact, and an activity. While the entirety of that background work cannot be adequately summarized in this report, a brief overview will here be presented.

Mastering a novel domain or skill requires building of schemata [10]. In the case of computer skills, learning is not only intellectual but enacted physically, through an artifact (i.e. mouse and keyboard). Patterns of visual-motor operation must be learned, in order to carry out higher-level actions which satisfy the goals of the highest-level activity [5]. This process is embodied skill acquisition, and it occurs in stages, beginning with formulation of explicit mental production rules and subsequently leading to internalization of those rules at an implicit and autonomous level [8, 3].

The nature of the educational interaction poses its own set of demands and understanding. Building of schemata must proceed in an orderly and structured way – scaffolding. Schemata are compositional, and constituent parts must be learned individually before the complex whole can be understood as its own discrete entity [9, 12]. Additionally, throughout this process, it is necessary to take into account the amount of “cognitive load” required of a learner at any given time. During training, cognitive load can be intrinsic to the activity being learnt, germane insofar as it goes toward schema construction, or extraneous in that it has only to do with the exercise of the training tool itself. It is desirable to maximize the former two and minimize the latter [11].

Finally, during learning, and during any kind of activity, it is critical that a person maintain focus, alertness, and as complete as possible attention. Ideally, a person in the midst of an activity achieves a state of “flow”, which is a type of heightened awareness in which one experiences one’s consciousness merging with the activity being performed [7]. Characteristics of flow include: challenge, immediate feedback, full control, high skill level, and intrinsic motivation, among others.

All these theories come to bear on the development of products and tools which are effective and usable; it is not only the research program of the theoretician, but the livelihood of many a software developer to design hardware and software tools that are user-friendly in all these respects. When it comes to scenarios involving real-life users and applications, the fields of academic sciences and applied software development are intricately intertwined.

III. Active User Paradox and HotKeyCoach

The active user paradox [6] has two phenomena at its heart. The first is the production bias: users are biased against having to invest time and effort into seeking new solutions to solve a problem, even when they are aware that a more efficient method exists. The second is the assimilation bias: when confronted with a new situation, users are biased toward applying existing knowledge and methods which have been previously successful [1, 4].

Additionally, it has been found that when confronted with a new task, users tend to attempt to discover new solutions on their own, rather than consult assistance [6].

Existing training methods are ineffective because they do not adequately deal with the active user paradox. Consulting a written manual violates the production bias; it requires too much additional effort in the short-term to go through the steps necessary to learn a new solution this way than to use a familiar solution, even if it will be less efficient in the long-run. Furthermore, the effort exerted would be disruptive to the flow of activity [1,2].

HotKeyCoach is a training tool that was developed in 2008 by Alterman and Krisler in an attempt to circumvent these vexing problems by providing on-line training and assistance to users of computer software, non-disruptively providing assistance “in-the-flow” of the user’s activity.

The way HotKeyCoach assists users is two-pronged. One is to scaffold mastery of an application’s underlying conceptual structure, and the other is to train optimally efficient visual-motor operation routines for interacting with the application. As explained above, there are essentially two distinct aspects of the trainer which accomplish these two ends respectively.

The first round of data collection was primarily focused on training at the operational level [1]. Users were trained to transition from menu selection to using hot keys (hence the name of the trainer), thereby reducing the “articulatory distance” involved in software use. HotKeyCoach works automatically with any application on Mac OS X in this regard, introducing “coaching events” when users select a menu item, displaying the keyboard shortcut for that action if one is available (Fig. 1). The trainer allows the user to either dismiss the coach or to practice the shortcut. It carries out the intended action upon input of the proper keystrokes, or triggers an alert if incorrect input is made. It can be “ignored” and thus never appear again for that specific menu item, or else persist until the user stops using the menu item completely. Thus is allows the user to practice until the shortcut is mastered or training is declined.

[pic]

Figure 1: Learning evens in original HotKeyCoach

In the meanwhile, additional functionality was being developed, which would train users’ understanding of an application conceptually, using a Concept Bar and Cribsheets [2]. For this feature of the trainer, when the user needs help finding an action in the software, she uses a shortcut (ctrl-tab) to summon a bar displaying the major concepts of the application (Fig. 2). This was modeled after the built-in “application switcher” feature of Mac OS X which uses the shortcut command-tab. Repeatedly pressing tab while holding down the control key cycles through the items in the bar, and releasing the shortcut keys displays the cribsheet for the selected concept.

[pic]

Figure 2: General Formula for Concept Bar

Each cribsheet contains information regarding the features and actions associated with the given concept. The cribsheets have three major components: First, it explains the concept, and breaks it down into sub-concepts if necessary. Second, it shows relevant menu items in a facsimile of the application’s menu structure, with only pertinent items displayed. Third, it provides a key to the symbols used in keyboard shortcuts to facilitate their use (Fig. 3). These are displayed in a consistent layout, varying only on the location of the symbol key which is placed as close as possible to the menu items.

[pic]

Figure 3: General layout for concept cribsheets

Theoretically speaking, whereas the first aspect of HKC was concerned with embodied skill acquisition, this aspect involved schema-building, scaffolding, and germane cognitive load. In terms of activity, the keyboard-shortcut learning events occurred in the Post-Action-Selection phase of activity, and this component introduced itself at the Pre-Action-Selection phase. Users previously encountered the trainer only once they had already selected a menu action; this training interface could be summoned at any time by the user when they need help to find the appropriate action in the first place.

HotKeyCoach was developed with a number of key principles in mind. These principles took theoretical issues and findings into account and translated them into actionable constraints for the HKC trainer design. In the first round there were four such principles, and by the second there were six. Paraphrased, they are as follows [1]:

• Introduce small learning events into the activity stream:

Rather than require users to read manuals or undergo formal training, HotKeyCoach enables the user to train themselves as needed, little by little, over the course of their existing interaction with the software.

• The principle of the heap:

Information is conveyed gradually, so that it does not at any point overwhelm the user, but builds up cumulatively. Software features are broken down into major concepts, and further into sub-concepts, so that “scaffolding” may occur from relatively little or no knowledge, up to the point of mastery.

• Training as focus-shifts:

In keeping with the first principle, training does not occupy its own space in terms of attention, but rather blends smoothly with the task. Importantly, the training does not at any point pose a major interruptions or distraction from the task at hand, comprising a “focus-shift” rather than a focus breakdown. The user’s flow is never disrupted.

• The user as locus of control:

At all times, the user retains control over the flow of both the task and the training. The trainer never takes over.

• Control costs at the level of visual motor action:

The trainer should help the user achieve optimal efficiency at the level of execution of intended goals. HotKeyCoach’s cribsheet layout supports learning of keyboard shortcuts.

• Use a consistent training style across applications:

HotKeyCoach provides the structure for a single unified method training style that can be implemented on any piece of software. This lends consistency to the training experience which in turn makes training more effective.

Altogether, HotKeyCoach was already well-established with theoretical underpinnings, prior literature, and one successful study already completed. It remained only to implement the concept-training functionality on a particular piece of software and collect data testing its use.

IV. HotKeyCoach Implementation: First Attempt

Though it did not ultimately go through as planned, the original plan for this study was for me to implement HotKeyCoach on Final Cut Pro, the digital film editing software used by many professionals and taught at Brandeis in the Film Studies department (Fig. 4). I have great interest in multimedia content and digital creativity tools, and completed a Film major in addition to my study in Cognitive Science. Hence it seemed that this would be an interesting way to find an intersection between fields.

[pic]

Figure 4: The Final Cut Pro Interface

In my experience with Final Cut Pro, I witnessed and experienced firsthand what I went on to learn with technical and theoretical precision in the Computational Cognitive Science course: semantic and articulatory distance; activity, action, and operation levels; focus and flow; cognitive skill acquisition; and finally, the active user paradox. Initially I learned it as a student in the Motion Picture Editing class. Since then I have continued to use it as an editor for several extracurricular film projects. I have also been a TA for other editing classes were students first learned to use the technology and tutored many novices as an employee of the Brandeis Digital Media Lab, Through these experiences I had become not only familiar with the intricacies of the software conceptually, but acutely aware of the difficulties and challenges the interface poses to people learning to use it.

The question becomes whether it is possible for a program to automate what an expert human tutor can provide intuitively.

The idea that a piece of software could provide online training to users of this highly complex and interesting piece of software was intriguing. Having learned about the theoretical background related to these issues, and about the HotKeyCoach trainer that Prof. Alterman and Brian Krisler had developed, the idea for an independent study germinated. It would allow me to both develop deeper understanding of the experimental and theoretical work and possibly to facilitate real-life skill development in another field of personal interest. The independent study was thus formulated with the goal of implementing HotKeyCoach on Final Cut Pro, and providing it to students in the Motion Picture Editing class for data collection.

At this stage of the research the software had only been tested for its potential to insert learning events at the Post Action-Selection phase of user activity. The next stage was to test HotKeyCoach’s functionality in providing support at the Pre-Action phase, by helping the user locate items within an application’s menu structure and scaffolding a grasp of software features conceptually. Thus it was necessary to install both the menu structure of Final Cut and to begin developing a breakdown of the software’s major components in order to create cribsheets explaining it conceptually.

So I began by inputting information about Final Cut Pro’s menu structure into an XML definitions file for HotKeyCoach. It was simple in premise, but nonetheless proved challenging because Final Cut has an extremely extensive menu structure. This was completed in the first two weeks of the study.

However, problems persisted. Due to the unusual complexity of the software, it proved extraordinarily difficult to break down the software successfully into manageable concepts and sub-concepts. What was truly fatal in the end was the fact that Final Cut Pro is fundamentally unusual in terms of its interface design. At the operation level, actions are often performed not by using menu items, but by manipulating elements such as clips of video directly with the mouse. HotKeyCoach is geared only to explain concepts in terms of the location of relevant tools within the menu structure, and ultimately seeks to reduce the articulatory gap by training the use of keyboard shortcuts. Final Cut Pro’s design proved itself too dissimilar from the conventional software interface to be effectively manageable by the training tool in its current design.

Though the original plan did not come to fruition, the process of the first attempt at implementation was a valuable learning experience. Having to think deeply about the conceptual and theoretical issues related to a piece of software which I care about very much and had already become familiar with allowed me to better understand those issues and the way training tools can improve user experience and productivity. To this day it puzzles me whether it would be ultimately possible to implement a trainer like HotKeyCoach on a formidably complex application such as Final Cut. Knowing what I know now, having successfully implemented on a simpler piece of software, I think it would be possible to re-approach the problem and gain more traction.

V. HotKeyCoach Implementation: Second Attempt

As the attempt to build cribsheets for FinalCutPro petered out, I began to think of scenarios in which HotKeyCoach might prove more easily testable, usable, and helpful. It occurred to me, when I was annotating PDFs for class reading, that the built-in Mac PDF and image-read application Preview might prove extremely appropriate.

Professor Alterman agreed that it was necessary to scale back and target an application of greater simplicity, but encouraged me to brainstorm a whole list of applications that could also be appropriate. Apple Mail, iCal, iTunes, and Safari were chosen as possibilities. In order to determine which of these would be most conducive to training with HotKeyCoach, Prof. Alterman had me draw up for each of these applications a Conceptual Model, and a Task Activity. The Conceptual Model described major concepts involved in the use of these applications, and the Task Activity consisted of a series of tasks which constitute exhaustive use of the software’s major features (Fig. 5).

Available Functions/Features:

-New Event, New To-Do, New Calendar, New Calendar Group...

-Tab Between Events, View Event Details (command-i), Tab Event Details, Edit Event Details

-Setup Alarm, Setup "Repeat""

-Navigate Time: Change Between Month, Week, and Day View (command-1, 2, 3), Cycle Between Currently Viewed Week/Day/Month (command-left,right), Go To Today (command-T), Go To Specific Date (shift-command-T)

-Export/Print

Conceptual Model:

-Events and Event Details, To-Dos and To-Do Details

-Time Frames/Cycles

-Navigation among time-frames, events, todos

-Calendar as a whole to be shared, exported

Task:

1) given a physical calendar with several events over the course of a week or two, both individual and "repeat", in 2 or 3 different "calendars" (e.g. work, school, home)...

2) input events into calendar with listing of Name, Location, Time, and Calendar

3) cycle between events and change their details

4) setup alarms for reminders of specific events

5) navigate timeframes to look through events across varying timespans

6) export and print a calendar, share with other viewers, integrate with Mail

Figure 5: Example Conceptual Model and Task Activity for iCal

Having developed these for all the candidate applications, they were presented for discussion the following week. Preview was decided upon in the end; it had the amount of balance of simplicity vs. complexity in its design and ease vs. challenge in its usability. Moreover, it was relevant and practically, for in an academic environment, the annotation tools which Preview provides are invaluable to myself and to other students, yet seem under-utilized.

The next step was to develop a more polished and formalized set of task procedures for which could be used to test people in an experiment on the efficacy of HotKeyCoach training in user activity and learning. Once this was done, it was possible to develop cribsheets explaining major concepts with visual guidance showing relevant menu items and shortcuts for each concept. This, in turn, required that we finalize a list of concepts.

In the initial specifications for Preview, two major concepts had been specified. These were 1) whole documents including both PDFs and image files, and 2) the selectable text within a document itself. Each of these concepts had tools associated therewith – documents are rotated, text is highlighted – and menu items that could be called up to do so.

Once Preview had been chosen, this was developed further and the picture became more complex. Preview allows the user to “navigate” the document in multiple ways, provides three major “tools” for interfacing with the document, allows different options for “annotating” not only text itself but on top of and around text, and so on. The concepts ended up being: Search, Annotate, View, Document, Tools, and Navigate. These were implemented in a Concept Bar as per the HotKeyCoach design (Fig. 6). In addition, cribsheets were developed for each of the concepts (Fig. 7).

[pic]

Figure 6: Concept Bar as implemented in Preview

[pic]

Figure 7: Cribsheet for the “Navigation” concept in Preview

When everything was ready – the concepts, the cribsheets, the task activity, surveys for before and after the task – it was time to design an experiment to test the success of the trainer in concept-building and menu-locating.

VI. Experimental Design

We investigated the effectiveness of HotKeyCoach as a tool for aiding activity and for learning application concepts. In particular, we hypothesized that users who completed a structure task using Preview in Mac OS X with the guidance of HotKeyCoach would more efficiently complete that task and would demonstrate better knowledge of the features and concepts in Preview immediately subsequent to the task as well as two weeks later.

To test this idea, we provided participants with a PDF document of a scientific journal article, and another PDF listing task instructions. Half the participants were required to complete these instructions in whatever way possible, with any help necessary from Preview’s buit-in help system or online documentation, but without the assistance of the HotKeyCoach training tool. The other half were taught how to use HotKeyCoach by a short instructional video and used the training tool in the course of their task completion.

This design allowed us to examine whether there is a linkage between use of a training tool in-the-flow of activity, and successful use and understanding of a computer application.

Participants

We recruited 20 participants from within the Brandeis community using the Brandeis classifieds website and personal communication. The ages of participants ranged from 18-40.

Data Collection

Pre-, post-, and two-weeks-delayed survey data was collected using a web-based service for adminstering surveys ().

HotKeyCoach (HKC) also is designed to collect data regarding user behavior and transfer this data to a database for analysis. For each interface interaction, we collected these items:

• Timestamp of the interaction

• Method of interaction (mouse or keyboard)

• Name of the interaction (Copy, Paste, etc.)

• Name of the application (Word, Safari, etc.)

Materials

All instructions and materials were included on an Experiment Webpage to which all participants were directed: iceland.cs.brandeis.edu/hotkeycoach/experiment/experiment.html

The following elements were included there and accessed by each participant:

● HotKeyCoach software

● PDF file “Mastery.pdf”

● Task Instructions (see Appendix 1)

● HKC Tutorial Video

● Pre-Experiment Survey: hosted by LimeSurvey (see attached appendix)

-Items G0-G4: Demographic data

-Items CS1-CS8: Computer skills

-Items D1-D2: Documentation Usage

-Items P1-P8: Preview Experience and Concepts

● Post-Experiment Survey hosted by LimeSurvey (see attached appendix)

-Items P1-P8: Preview Concepts

-Items T1-T20: HotKeyCoach Feedback

● Two-Weeks-Delayed Survey: hosted by LimeSurvey (not complete yet)

Items P1-P8: Preview Concepts

Procedure

Participants were directed to the experiment webpage which contained a complete set of instructions, and thereby enabled them to complete the task in its entirety independently. They first read and informed consent form. Then, they downloaded and installed HotKeyCoach and located their unique randomly generated and anonymous Participant Identifier ID.

On the basis of the first digit of their Identifier ID, participants were directed to a page which gave them instructions on completion of the experiment.

Group 0: Control Group

Were provided with the PDF and Task Instructions, and completed the task without the aid of HotKeyCoach

Group 1: Experimental Group

Were shown an instructional on the use of HotKeyCoach, and then downloaded the PDF and Task Instructions and completed the task with the aid of HotKeyCoach.

VII. Where we stand now

The pilot study is mostly complete. Below is a review of where we stand. At this point has been know analysis of the data.

Before continuing with results and data, it is relevant with respect to the independent-study and capstone-experience aspect of this summary to comment on the experience of running this experiment and collecting data.

The creation and implementation of a complete web-based experiment was a complicated process requiring problem-solving and close collaboration. First, task instructions were drawn up and tested on Prof. Alterman himself and tested on a single participant during our meeting. In this way we were able to proofread the surveys

VII. Data

Pretest

Pretest; Posttest, Survey; here is a link to a screen cast;To be completed! Show some data. Show the survey.

VIII. Results

To be completed! No analysis yet.

VII. Reflection

Psychology and Technology

For a long time I have been interested in the relationship between Psychology and Technology. The tools we use must synchronize harmoniously with the structure of our mental and manual abilities, obviously, or else we would not be able to use our tools effectively. Developing a trainer requires insight into cognition. This is true both theoretically and practically. Theoretically, work like the study I did in my senior year has direct bearing on theoretical topics like skill acquisition, embodiment, distributed cognition versus individualistic pschology, external/internal representation…. As a practical matter to develop a training tool like hot key coach, one must work as a cognitive engineer. Through application of theories and empirical results from cognitive science, I gained new insight and understanding of the field.

AlsoMoreover, the insights we gain in Cognitive Science as to the brain’s computational mechanisms help us to develop more sophisticated Artificial Intelligence tools, and vice-versa. My favorite example of this is the connection between the Google search algorithm “PageRank” and human long-term memory retrieval (Griffiths and Tenenbaum).

Situated Activity

Approaching the brain as a computational system was fascinating, but things took an even more compelling turn when I learned that not only the brain, but the entire environment in which it lives and interacts, can be analyzed as a cognitive system.

Mind, Body, and skill acqusition

reference papers for 118

Methods: Statistics and Experimental Design

The experimental evaluation of the trainer required the use of methods that I learned in the course I took in psychology on statistics and experimental design. … reference textbook

Overall his study has entailed connections with myriad components from the entirety of the Cognitive Science major. Statistics and Research Methods were employed in the design and execution of a scientific experiment. Experience with recruitment and testing of research subjects, as conducted in an assistant research position at the Brandeis Vision Neurophysiology Lab under supervision of Prof. Bob Sekuler, was invaluable here.

Gee whiz design experiment, developing the material, working subjects, sure turned out to be a lot more work than I thought. A lot of things went wrong, there was scrambling. I sure learned a lot from doing this … now when I read an experimental paper Ihave completely differend understanding.....

Methods: Cognitive Engineering

As a practical matter to develop a training tool like hot key coach, one must work as a cognitive engineer. Through application of theories and empirical results from cognitive science, I gained new insight and understanding of the field. For example, subjects will only use the trainer if it is less work to achieve their immediate goals than other options. As a cognitive engineer, we applied the principles of control costs at the level of visual motor routines and focus shifts not interruptions. Both of the principles required that we reason cognitive scientists about things like visual motor routines and attention, topics which I had learned previously about in name of course. references

Elegant Designs

In the case of Human-Computer Interaction and software design, what we find is that the most intuitive and “elegant” inventions are those which most harmoniously complement the mind’s natural tendencies. Hence the insights Cognitive Science makes with regard to brain function are very important to technology development. In return, advances in technology make it more possible for research in Cognitive and Brain Science to proceed, yielding data-analysis packages, testing environments, and other useful innovations.

Overall Value

Altogether, this capstone experience served as a fitting conclusion to the Cognitive Science major both academically and experientially. It was at once intellectually challenging and “hands-on”, and manifested tangibly the interactions I was interested in between psychology and technology. Though it did not proceed as initially planned, the experience of beginning with one idea, being confronted with and attempting to deal with serious difficulties, and subsequently having to revise the original course of action, was obviously very valuable. From exploring theoretical underpinnings, to participating in experimental design, to collecting data, to reflecting and synthesizing the entirety of the experience in a process of meta-cognition, the entirety of the experience constitutes a rigorous and thorough culmination of undergraduate work and study.

In the course of my studies in Cognitive Science I have learned about the mind and brain from multiple perspectives. Primary among these are Neuroscience, Psychology, and Computer Science. The brain is an organ with emergent complex properties and can be analyzed from each of these levels: biology, computation, and behavior. At the primary level are neurons and cytoarchitectural Brodmann areas, at the next level are networks and structures, and finally there are recognizable patterns of demonstrable behavior. Vision scientist David Marr depicted three “levels of explanation”: computation, algorithm, and implementation.

IX. References

1) Alterman, R. and Krisler, B., Training towards mastery: Overcoming the active user paradox. Proceedings of the 5th Nordic conference on Human-computer interaction, 2008: p. 239-248.

2) Alterman, R. and Krisler, B., Concepts and cribsheets: A concept-based training architecture. Brandeis University Volen Center for Complex Systems, 2011.

3) Anderson, J.R., Acquisition of cognitive skill. Psychological Review, 89:4, 1982, p. 369-406.

4) Anderson, J.R., Rules of the Mind. Lawrence Erlbaum Associates, 1993, p. 320.

5) Bannon, L. J., & Bødker, S., Beyond the interface: Encountering artifacts in use. In Carroll, J. M. (Ed.), Designing interaction: Psychology at the human-computer interface, 1991, p. 227-253.

6) Carroll, J.M. and M.B. Rosson, Paradox of the active user. Interfacing thought: Cognitive aspects of human- computer interaction, 1987: p. 80-111.

7) Csikszentmihalyi, M., Flow: the Psychology of Optimal Experience: Steps toward Enhancing the Quality of Life, New York: Harper and Row, 1990.

8) Fitts, P.M., Perceptual-motor skill learning. In Melton, A.W. (Ed.), Handbook of learning and cognitive processes (Vol. 1), 1975.

9) Pea, R., The social and technological dimensions of scaffolding and related theoretical concepts for learning, education, and human activity. The journal of the learning science, 13:3, 2004, p. 423-451.

10) Rumelhart, J., Schemata: the buildings blocks of cognition. In Spiro, R.J. et al. (Eds.), Theoretical issues in reading comprehension, 1980.

11) van Merrienboer and Sweller, “Cognitive Load Theory and Complex Learning: Recent Developments and Future Directions.” Educational Psychology Review, 17:2, June 2005

12) Vygotsky, L., Interaction between learning and development. In Mind and Society, Cambridge, MA: Harvard University Press, 1978.

X. Appendices

Appendix 1: Task Instructions

Part 1

1. Open the document named Mastery.pdf

2. Rotate the document until it is correctly oriented.

3.Save the document with a new name, Mastery-Upright.pdf.

4.Display the Thumbnails Sidebar and use it to navigate to the third

page of the document.

5.Without using the mouse, move forward and backward a page.

6.Return to page 1.

7.Page through the document without using the mouse and find the

page with the section titled: Active User Paradox.

8. Assign the Bookmark ʻActive User Paradoxʼ to this page.

9. Page through the document without using the mouse and find the

page with the section titled Experimental Design.

10. Assign the Bookmark ʻExperimental Designʼ to this page.

11. Navigate from the page with the Experimental Design section to

the page with the Active User Paradox section, using your bookmarks.

Part 2

1. In the section Abstract, highlight the sentence which begins with the phrase, “In this study”.

2.Use the search term “Active User Paradox” [including quotes], to find occurrences of that phrase in the document, and underline each occurrence.

3.Find the section Circumvention of the Assimilation Bias and underline the first sentence of the paragraph that explains Table 1.

4.In the section Minimal Task Interruption, the difference between breakdowns and focus shifts is defined. Add a Note to the margin of the document, near the definition of the term, that says, “definition of focus shift”.

5.Hide your Note.

6.Show your Note.

7.Print the document (do not actually print, hit cancel on the print

page).

Part 3

1. Find Table 1.

2. Draw an annotation oval around the highest number in the Learned

column.

3. Draw an annotation arrow pointing to lowest Usage percent.

4. Navigate to page 7.

5. Using the mouse, select the text Assimilation Bias in the section

heading.

6. Find the definition of the term selected in the previous step.

7. Using annotations, highlight the definition.

8. Email the document to nhakimi@brandeis.edu

9. Complete the end of experiment survey.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download