Results consistently revealed that visuals, either still graphic or animated, in the on-line help instructions
enabled
the users to significantly perform more tasks in less time and with fewer errors than did users who did not
have visuals
accompanying the on-line help instructions. Although users receiving spoken instructions were faster and
more accurate for
the initial set of tasks than were users receiving written instructions, the majority of subjects preferred
written instructions
over spoken instructions. The results of this study suggest additional empirically-based guidelines to
designers for the
development of effective on-line help.
Keywords:
Graphical user interfaces, on-line help, visuals, user interface components.
Introduction
With the advancement of microcomputer technology and the use of graphical user interfaces, the number of
features
available within software applications has rapidly increased. Word 5.0a [22] for example, is a word
processing package for
the Macintosh that presents the user with a choice of 97 features in the opening screen of the text-based on-
line help system.
The Excel 4.0 spreadsheet package [23] lists 22 features in the first level of the text-based help directory,
and a second
listing appears containing 111 features for the spreadsheet's worksheet.
Unfortunately, advances in the design of on-line help systems have not matched recent advances in the
design of interface
technology. Researchers have observed that subjects in need of assistance often do not utilize the on-line
help available
[9,20,32]. Subjects cited "poor quality of the help information" as a common reason for avoiding on-line
help. Researchers
have found that the common textual form of on-line help tends to hinder rather than help the user in the
performance of
procedural tasks [9,10,15,32]. The visual and dynamic nature of modern graphical user interfaces is not
captured by the
traditional form of on-line help [33,37].
The on-line help of this study presented instructions for completing procedures involving the creation,
editing, and deletion
of five types of objects within a HyperCard environment. HyperCard is an integrated object-oriented
programming system
for the development of documents with linked segments of information of text, sound, and graphics [25].
The visual
illustrations presented within the on-line help of this study included still graphics and animated visuals in
the form of a
computer-generated demonstration using animation.
The relative effectiveness of each design of the on-line help within this study was evaluated by the speed
and accuracy in
which the user performed each procedure. As suggested in performance measurement guidelines by Bovair
and Kieras [8],
dependent measures of completion rates, elapsed time, and error rates were used to evaluate the
effectiveness of written or
spoken instructions with or without still graphic or animated visuals in enhancing the user's performance of
computer-based
procedures within a graphical user interface environment.
BACKGROUND
Designers of on-line help presenting procedural instructions "should concentrate on ensuring that the
procedure construction
goes smoothly" [8,p.227]. Procedure construction is part of the process of acquiring a procedural skill - a
process that
researchers recognize as an aggregate of three phases. In the first phase relevant facts for performing the
skill are encoded
and committed to memory in a declarative form [3]. The learner focuses on understanding the skill by often
verbally
rehearsing the information as the skill is performed. Strategies for executing the procedure are formulated,
tested, and
executed in a consciously controlled way [31] during this phase, resulting in slow and error-prone
performance [1]. During
the second phase, execution errors are gradually detected and eliminated as the learner practices the skill
[24]. The
execution of the procedure becomes automated and faster through extensive additional practice as the set of
individual steps
are collapsed into one large unit during the third phase [2].
The amount of practice the user invests in learning a procedural skill is the user's choice, not the
researcher's or the
software designer's. This study investigated the effectiveness of design elements in on-line help without the
variable of
practice. By excluding the use of practice, this study was able to focus on the design effects of on-line help
rather than on
practice effects.
PREVIOUS RESEARCH
Only a few studies reported in the literature employed the use of on-line help for computer-based
procedures and only three
of those studies [27,28,37] investigated the use of graphics and animation in on-line help. These studies
incorporated the
use of practice in addition to the on-line help as a means of aiding the user in acquiring the procedural
skills.
Shneiderman [33] contends that use of a graphical demonstration is the most direct way for novices to learn
the basic
functionality and steps necessary to perform procedures within the graphical user interface environment. A
few researchers
[26,28,37] have investigated the potential use of demonstrations within on-line help for presenting
procedural information.
Such demonstrations of real-time instantiations of computer-based procedures have been termed "animated
demonstrations"
[28]. Like still graphics, animated demonstrations immediately show what objects are used within the
procedures. This
approach was shown to be partially effective [27,28]. Segmented animated demonstrations, presenting one
step of a multi-
step task at a time, may aid the user in developing discrete steps of the procedure. Unless the animation is
segmented, the
user may not be able to identify and encode specific aspects of the procedure [8,26]. No studies were
found, however, that
investigated the use of segmented animated demonstrations.
Overall, the research with on-line help indicates that the current use of text-based on-line help is generally
ineffective
[10,11,12,15,18]. Researchers agree, however, that the effectiveness of on-line help may be improved if
on-line help was
properly developed using empirically based guidelines. The empirical evidence from research investigating
the use of
instructions, audio, visuals, and animation in the design of on-line help suggest that on-line help for
procedural tasks may be
more effective when:
- Explanations providing goal structure or organizational information are presented at the beginning of
the procedural task
[13];
- Procedural instructions are divided into small steps with each step labeled and presented in the order
they are to be
executed [8];
- Information within each step is explicitly stated, and the most important information is presented first
[14];
- Visuals accurately depict the procedural step since users often prefer to follow the visual examples
rather than the
instructions [19] and visuals tend to help eliminate orientation errors [36];
- Visuals are accompanied by some form of written or spoken instructions in order to cue the user to the
important aspects
of the visuals [35];
- Spoken instructions begin simultaneously with or slightly after the visuals are presented [6,21];
- Animation is segmented to focus the user's attention on specific parts of the animated displays [29,30];
- Instructions are segmented to reinforce the concept of chunk or steps for completing the procedure [8];
and
- The opportunity to perform the procedure is delayed until all steps of the procedure have been
illustrated [7].
By following these guidelines for the design of the on-line help, the user's mental representation of the
procedure should be
more accurately constructed, the referential links to the physical objects should be better connected, and the
individual
segmented steps of the procedure should be better organized within memory thus providing a proper
foundation for the
successful performance of the procedure.
EMPIRICALLY-BASED ON-LINE HELP DESIGN
The design of the elements in the on-line help adhered as close as possible to the empirically-based
guidelines listed. The
order and the language of the instructions, in both written and spoken forms, were based on the guidelines
presented. Still
graphic or animated visuals were added to the instructions. The spoken instructions accompanying the still
graphics or
animation were designed to begin immediately after the first visual appeared on the screen or as soon as the
animation
sequence started. The step-by-step procedural information was presented in a segmented form. In-depth
information
pertaining to the studies associated with the guidelines listed is found in [17].
The combinations of written or spoken instructions with or without still graphics or animated visuals in a
segmented form
made this study one of the first experiments of its kind to investigate the effectiveness of on-line help on the
performance of
computer-based procedures within a graphical user interface environment. This study was designed to
investigate the
relative effectiveness between:
- the use of on-line help with and without visual illustrations;
- the use of on-line help with still graphics and with animated visuals; and
- the use of on-line help with written instructions and with spoken instructions.
METHODOLOGY
The experimental context of this study was a HyperCard emulator with on-line help in which subjects
received instructions
for completing seven procedural tasks. This study used a 2 x 3 randomized factorial design with Verbal
Instruction and
Visual Illustration as the two treatment factors. The Verbal Instruction factor was the presentation of
instructions either in
spoken or written form. The Visual Illustration factor presented either no illustrations, still graphic
illustrations, or animated
graphics.
Participants
Participants for this study were drawn from an available pool of 215 undergraduate students enrolled in the
Science
Curriculum and Instruction course within the School of Education and in the Introduction to Computers and
Their Uses
course within the School of Arts and Science at a university of approximately 10,000 students. To be
eligible to participate
in the study, each student was required: (a) to demonstrate a basic ability in the use of a Macintosh
computer, keyboard,
mouse, and pull-down menus; (b) to have no previous experience in the authoring techniques of HyperCard
[5]; (c) to have
completed the HyperCard Tour 2.0v2 tutorial [4]; and (d) to have demonstrated an understanding of the
basic computer and
HyperCard concepts. These requirements reduced the potential sample size to 198. Equipment failures,
language barriers,
and lack of attendance during the experiment resulted in 176 participants being randomly assigned to one of
the six
treatment groups.
Materials
All materials used in this study, including the computer program, were designed and developed by the
experimenter in
consultation with experts in the areas of survey construction and simulation programming. A set of eight
HyperCard
authoring tasks were used in this study. An example of a task explanation, specific task description, and
task steps for one
of the tasks is shown in FIGURE 1.
FIGURE 1. Details of Change Button Icon Task.
HyperCard Emulator Design. A HyperCard emulator was developed in SuperCard [34] to
avoid potential
difficulties involved in programming and controlling menu responses and capturing timing information and
mouse locations
while using the actual HyperCard application. The HyperCard emulator included the screen images, menu
options, and
dialog boxes normally seen during the process of performing the HyperCard procedures used as tasks in the
study. The
basic layout of the screen shown in FIGURE 2 was divided into four functional areas [16].
FIGURE 2: Screen Design with Four Functional Areas.
The large main functional area, the work space, covered three-fourths of the upper portion of the screen.
The work space
was used to display the simulated HyperCard menu bar and all still or animated graphics. Subjects
performed all tasks in
this area using the simulated menu bar and related dialog boxes. Welcome Information and Task
Explanations covered this
area when such information was displayed.
Information and options relating to the current step of a task being explained or performed were placed in
the second
functional area positioned at the bottom of the screen. To the left of this area, the number of the current
step and the option
to repeat a step were located. The main portion of the informational area contained the current step
instruction. An
abbreviated version of the task requirement was shown below the current step instruction.
Supplemental information concerning the current phase of the study was displayed in the third area located
in the upper
right corner of the screen. When the Abort option appeared after four incorrect attempts were made to
complete a step, it
also appeared in this area. The fourth area located at the bottom right of the screen contained the current
navigational
option. Options, displayed one at a time, included advancing to the next step, beginning the task test,
continuing, beginning
Help, and exiting.
Color was added to the screen design to help divide the functional areas. The dark areas shown on the
screen were
displayed as a medium to light blue. The background of any informational field was a light yellow. The
colors for the
screen design were specifically selected as a result of the user evaluations. The final version of the
emulator, complete with
all sounds for the spoken instructions and converted to a stand-alone application consumed 1,287K on a
high-density
diskette.
Treatment Designs. All steps for a given task were presented consecutively in all treatments.
To maintain a
sense of realism, a simulated HyperCard menu bar was visible across the top of the screen for all treatment
groups. Subjects
using spoken forms of the Verbal Help or seeing animated forms of the Visual Help were allowed to replay
the current step
of the task. Immediately after the Help presentation for the task, subjects were required to complete the task
once without
additional access to help. The completion of the task was part of the Immediate Exercise set, and the
screens that were
accessed were the same for all treatment groups.
The Written, Animated Treatment, WA, animated the cursor movement required to perform the instruction.
This movement
provided subjects with a visual cue to the important point of the instruction. Options on pull-down menus
were highlighted
as the cursor moved down through the choices. FIGURE 3 demonstrates the way the cursor moved across
the screen and
activated a pull-down menu on the help screen for the WA treatment. Although four screens are used in
FIGURE 3, all
cursor movement in the treatment occurred on the same screen.
FIGURE 3: Four Screens Demonstrating an Example of Written Animated Help Screens.
A still graphic that corresponded only to the written step instruction was shown in the work area of the
screen for the
Written, Still Graphic Treatment, WS. The graphic illustrated the way the screen looked as that step of the
procedure was
completed, screen (d) of FIGURE 3.
Subjects assigned to the Written, Nonillustrated Treatment, WN, were provided written step instructions,
one at a time, in the
same location at the bottom of the screen as illustrated in screen (a) of FIGURE 3. As in other written
treatments, the task
requirement was displayed below the written instruction and the step indicator was located to the left of the
written
instruction. Graphic examples were not provided for the WN treatment.
Subjects in all spoken treatments were presented each step instruction through a set of headphones. Spoken
instructions
were identical to the written instructions presented to written treatment groups.
No written instructions appeared on the screen in the Spoken, Nonillustrated Treatment, SN, treatment. The
area normally
used for written instructions was covered with a medium blue color. The task requirement, however, was
shown at the
bottom of the screen Graphic examples were not provided for the SN treatment. The Repeat Step option,
located below the
step indicator, was available to allow the subjects to replay the instruction.
Both the Spoken, Still Graphic Treatment, SS, shown in FIGURE 4 and the Spoken, Animated Treatment,
SA, provided
visuals similar to those described for the WS and WA treatments, respectively. Like the SN treatment, step
instructions were
verbally presented through a set of headphones and no written instructions appeared on the screen.
FIGURE 4: Spoken, Still Graphic Treatment Screen.
Questionnaire. Seven sets of mirror images of statements assessing the subjects' reactions to
the help
instructions were created. For example, if the original statement said "I prefer to receive help instructions
in a written rather
than spoken form," the mirror image said "I prefer to receive help instructions in a spoken rather than
written form." The
questionnaire used a six-point scale ranging from Strongly Agree to Strongly Disagree.
Procedures
Eight groups of subjects, composed of 20 to 30 people, completed the study using Macintosh LC II
computers in the
university computer laboratory. The experimenter arranged the subjects in the room so that subjects
assigned nonillustrated
help could not see the screens displaying of illustrated help and subjects assigned still graphic help could
not see the screens
displaying animated help. Subjects were guided through the steps of completing one practice task using the
on-line help
instructions. After all subjects had tried the task and questions were answered by the experimenter, subjects
continued with
the seven tasks of the Immediate Exercise set.
The following steps were repeated for each task in the set.
- On the computer screen, the usefulness of the task was explained and the specific task requirement for
completing the task
was stated.
- A written or spoken help instruction for completing the first step of the task was presented by the
computer. At the
subject's request, each additional step was given, one at a time, until all steps of the task were presented.
- After the last step was presented and when the subject was ready to continue, a message reminding the
subject of the task
requirement appeared on the screen.
- Subjects attempted to complete the entire procedure at their own pace. After four incorrect attempts,
the option to abort
appeared until the subject chose to abort or until the correct action was selected.
After the series of tasks were completed, subjects were asked to complete the questionnaire about help
instructions.
Dependent Measures
Audit trails, commonly used in analyzing learner-computer interactions [38], were included in the
HyperCard Emulator to
record completion rate, elapsed time, and error rate. Values for each category were directly collected or
calculated by the
HyperCard Emulator for each step of a task and for each task of the set. Responses to the questionnaire
were entered into a
data file and verified by the experimenter.
RESULTS
Data were analyzed for 176 subjects randomly assigned to the six treatment groups as shown in FIGURE 5.
FIGURE 5: Two by Three Randomized Design of the Study.
A priori orthogonal contrasts, shown in FIGURE 6, were used for analyzing the data of the six treatment
cells. Contrast
Visual (1) compared treatment groups receiving nonillustrated and illustrated (still graphic and animated)
forms of help.
Visual (2) compared treatment groups receiving still graphic and animated forms of help.
FIGURE 6: A Priori Orthogonal Contrasts Used in Analysis.
Although multiple contrasts were used in the analysis, only the comparison of the nonillustrated and
illustrated groups
revealed any significant differences in the study.
Effectiveness was measured in terms of the dependent variables for completion rate and error rate.
Efficiency was measured
in terms of the dependent variable for the time the subject used in attempting to complete a task or set of
tasks.
Completion Rates for Exercise Sets
Completion rates were inspected in two ways: the percent of each treatment group completing all seven
procedural tasks of
the set and the average percent of tasks completed within each treatment group.
Percent of treatment group completing all tasks. A log-linear comparison of the cell
proportions across
treatment groups for subjects who completed all seven tasks of the Immediate Exercise set with subjects
who did not
complete all seven tasks revealed a significant difference for the nonillustrated versus illustrated contrast
(coef. = .19, 95%
CI = .08 to .30). An examination of cell frequencies showed that only 22 of the 58 subjects or 37.9% of the
nonillustrated
groups (WN, SN) completed the entire Immediate Exercise set. With the illustrated (WS, SS, WA, SA)
groups, 78 of the 118
subjects or 66.1% completed the entire Immediate Exercise set.
Average percent of tasks completed. The number of tasks completed out of the seven possible
tasks of the
exercise set was converted to a percentage for each subject. A significant difference for the nonillustrated
versus illustrated
contrast was detected (F(1, 170) = 12.99, p < .001). Cell means for the percent of tasks completed within
the Immediate
Exercise set by treatment group reveal that the nonillustrated groups (WN, SN) have a combined mean
completion rate of
83.5% (n = 58, SD = 18.69). The illustrated groups (WS, WA, SS, WA) have a combined mean
completion rate of 92.5% (n
= 118, SD = 13.82).
Error Rates for Exercise Sets
Since subjects were allowed to abort the performance of a procedural task on any step of the procedure,
calculations for
error rates were based on errors per step attempted. When error rates were evaluated, a significant
difference was found for
the nonillustrated versus illustrated contrast (F(1, 170) = 22.48, p < .001) when an a level of .01 was used.
An evaluation
of cell means shows the combined nonillustrated (WN, SN) conditions to have a higher error rate (M =
1.83, SD = 1.33, n =
58) than the error rate (M = .95, SD = 1.08, n = 118) of the combined illustrated conditions (WS, WA, SS,
WA).
Elapsed Time for Exercise Sets
Because subjects did not always complete all of the steps of a given procedural task within the exercise set,
elapsed times
were measured based on the number of seconds a subject spent in attempting to complete a step of a
procedure. To
calculate elapsed time per step, the total amount of time elapsed while the subject attempted to complete all
task steps was
divided by the total number of steps the subject attempted to complete. Results of the MANOVA test for
the illustration
contrast and an inspection of the cell means indicated that the significant differences occurred (F(1, 170) =
28.99, p < .001)
with nonillustrated groups having a mean elapsed time of 13.80 seconds and illustrated groups having a
mean of 9.58
seconds.
DISCUSSION
Nonillustrated versus Illustrated
The findings of this study add support to the notion that the presence of a visual, either a still graphic or an
animation, aids
the subjects in acquiring accurate representations of the steps involved in the procedure. At least two
possible explanations
exist for why subjects receiving illustrated on-line help exhibited faster and more accurate initial
performances of the
procedural tasks than did the other subjects.
First, the visuals contained pictures of the objects in the graphical user interface that were to be used in
performing the
procedure. Phrases such as Edit Menu, the OK button, and Field tool, scroll down arrow, and Shared Text
box, mentioned
in the instructions, could have been internally mapped to the actual object. Second, the visuals provided an
example of the
system response to the user input so the subject would not be surprised by the screen image displayed.
Subjects in the
illustrated groups were perhaps more prepared for the next step of the procedure since they would have
known what image
would appear as a result of their action. The presentation of the illustrations during the on-line help may
have allowed
subjects to visually rehearse and plan the actions necessary for the completion of the procedure. It is likely
that subjects in
the nonillustrated groups were unable to be prepared since they lacked the ability to mentally picture the
screen image
associated with the action's response.
Still Graphic versus Animated Illustrations
An unexpected result of this study was the lack of a significant difference between the subjects'
performance in the still-
graphic conditions and the animated conditions. However, the data suggested a slight advantage for
animation over still
graphics as expected. One possible explanation for the lack of difference is that the task performed in this
study did not
require the coding of motion information which would have been provided by the animation. In this study,
the main
difference between the information provided by the still graphic condition and the information provided by
the animated
condition was the movement of the cursor. It is likely that subjects inferred the movement of the cursor by
mentally noting
the change in cursor location from the graphic image illustrating the first step to the graphic image
illustrating the second
step. If the procedural tasks had involved the movement of an actual object such as the resizing of a button
or the changing
of a picture's location, it is possible that significant differences would have been found between the still-
graphic and
animated conditions.
A second possible explanation for the similar results between the still-graphics condition and the animated
condition relates
to the use of segmentation. Individual task steps were emphasized by each image presented in the still-
graphics condition.
If the animated presentation had been presented in a continuous form, like the animated demonstrations in
the Palmiter [26]
study, it may have been difficult for the subjects to identify the individual procedural steps. However, since
the animation
provided in this study was segmented to emphasize each step of the procedural task, subjects in both the
still-graphics and
animated conditions probably were equally able to identify each procedural step. The segmentation of the
animated
illustration may be one reason why the animated conditions consistently performed better than the other
conditions across
time and type of tasks. Although the better performance was nonsignificant when compared to the other
groups, results
showing that the animated groups tended to maintain their high performance across time are noteworthy
since those results
contradict the findings of the Palmiter [26] studies.
Implications of the Findings
Caution should be exercised in generalizing from this population to a larger population of college students
and from the
procedural tasks included in this study to other computer-based procedures. The robustness of the findings
of this study,
however, provides a strong basis for suggesting that instructions for computer-based procedures presented
via on-line help
should be accompanied by some form of visual illustration. Designers of current on-line help facilities in
the graphical user
interface environment should be encouraged to consider changing from the standard from of text-based on-
line help to a
form of on-line help that incorporates graphics.
Recommendations for Future Research
Based on the results of this study, the review of the literature, and the general observations made by the
experimenter during
this study, two areas of research for investigating the effectiveness of on-line help are recommended. Since
tasks in this
study did not involve the aspect of motion beyond the movement of the cursor, no significant advantages
were revealed for
the use of animation. Research needs to be conducted using a set of computer-based procedures involving
the movement of
different objects on the screen to help determine the effectiveness of animation in on-line help.
When the on-line help information in this study was presented, the work space on the subject's monitor was
completely
filled. Since the size of the typical computer monitor is rapidly increasing as technology advances, different
conclusions
about the effectiveness of on-line help may have been reached if the on-line help information remained
visible on the screen
as the subject completed the procedure. A study could be conducted to compare how the presence or
absence of the on-line
help during task performance affects the completion of a computer-based procedure.
The results from this study can be summarized in one statement: The use of visuals within on-line
help instructions for
computer-based procedures enabled adult subjects to perform more procedural steps in less time and with
fewer errors than
subjects who received no visuals within on-line help instructions. Insights gained from this
investigation should
contribute to the current knowledge base used by designers to guide their development of on-line help and
ultimately help
the end-user.
References
1. Ackerman, P. L. (1989). Individual differences and skill acquisition. In P. L. Ackerman, R. J.
Sternberg, & R. Glaser (Eds.), Learning and individual differences: Advances in theory and research (pp.
165-217).
New York: W. H. Freeman.
2. Anderson, J. R. (1982). Acquisition of cognitive skill. Psychological Review, 89(4), 369-406.
3. Anderson, J. R. (1990). Cognitive psychology and its implications (3rd ed.). New York: W. H.
Freeman.
4. Apple Computer (1990b). HyperCard Tour 2.0 [Macintosh program]. Cupertino, CA: Apple
Computer.
5. Apple Computer (1991). HyperCard 2.0 [Macintosh program]. Cupertino, CA: Apple Computer.
6. Baggett, P. (1984). Role of temporal overlap of visual and auditory material in forming dual
media associations.
Journal of Educational Psychology, 76(3), 408-417.
7. Baggett, P. (1988). The role of practice in videodisc-based procedural instructions. IEEE
Transactions on Systems,
Man, and Cybernetics, 18(4), 487-496.
8. Bovair, S., & Kieras, D. E. (1991). Toward a model of acquiring procedures for text. In R. Barr,
M. L. Kamil, P. B.
Mosenthal, & P. D. Pearson (Eds.), Handbook of reading research: Vol. II (pp. 206-229). White Plains,
NY: Longman.
9. Carroll, J. M., & Mazur, S. A. (1986). LisaLearning. IEEE Computer, (November), 35-49.
10. Cohill, A. M., & Williges, R. C. (1985). Retrieval of HELP information for novice users of
interactive computer
systems. Human Factors, 27(3), 335-343.
11. Czaja, S. J., Hammond, K., Blascovich, J. J., & Swede, H. (1986). Learning to use a word-
processing system as a
function of training strategy. Behaviour and Information Technology, 5(3), 203-216.
12. Czaja, S. J., Hammond, K., Blascovich, J. J., & Swede, H. (1989). Age related differences in
learning to use a text-
editing system. Behaviour and Information Technology, 8(4), 309-319.
13. Dixon, P. (1987a). The processing of organizational and component step information in written
directions. Journal of
Memory and Language, 26(1), 24-35.
14. Dixon, P., Faries, J., & Gabrys, G. (1988). The role of explicit action statements in understanding
and using written
directions. Journal of Memory and Language, 27(6), 649-667.
15. Dunsmore, H. E. (1980). Designing an interactive facility for non-programmers. In Proceedings
of the 1980 Annual
Conference of the Association for Computing Machinery, (pp. 475-483). Nashville, TN: Association for
Computing
Machinery.
16. Grabinger, R. S. (1993). Computer screen designs: Viewer judgments. Educational Technology
Research and
Development, 41(2), 35-73.
17. Harrison, S. (1994). Still, animated, or nonillustrated on-line help with written or spoken
instructions for performance
of computer-based procedures (Doctoral dissertation, University of Minnesota, 1994).
18. Hicks Jr., J. O., Hicks, S. A., & Sen, T. K. (1991). Learning spreadsheets: Human instruction vs.
computer-based
instruction. Behaviour and Information Technology, 10(6), 491-500.
19. LeFevre, J.-A., & Dixon, P. (1986). Do written instructions need examples? Cognition and
Instruction, 3(1), 1-30.
20. Mack, R. L., Lewis, C. H., & Carroll, J. M. (1983). Learning to use word processors: Problems
and prospects. ACM
Transactions on Office Information Systems, 1(3), 254-271.
21. Mayer, R. E., & Anderson, R. B. (1991). Animations need narrations: An experimental test of a
dual-coding
hypothesis. Journal of Educational Psychology, 83(4), 484-490.
22. Microsoft Corporation (1992a). Excel 4.0 [Macintosh program]. Redmond, WA: Microsoft.
23. Microsoft Corporation (1992b). Microsoft Word 5.0a [Macintosh program]. Redmond, WA:
Microsoft.
24. Neves, D. M., & Anderson, J. R. (1981). Knowledge compilation: Mechanisms for the
automatization of cognitive
skills. In J. R. Anderson (Ed.), Cognitive skills and their acquisition (pp. 57-84). Hillsdale, NJ: Lawrence
Erlbaum
Associates.
25. Nielsen, J., Frehr, I., & Nymand, H. O. (1991). The learnability of HyperCard as an object-
oriented programming
system. Behaviour and Information Technology, 10(2), 111-120.
26. Palmiter, S. (1991). Animated demonstrations for learning procedural, computer-based tasks
(Doctoral dissertation,
University of Michigan, 1991). Dissertation Abstracts International, 52, 03B.
27. Palmiter, S., & Elkerton, J. (1991). An evaluation of animated demonstrations for learning
computer-based tasks. In
S. P. Robertson, G. M. Olson, & J. S. Olson (Eds.), Proceedings of CHI ‘91 Human Factors in Computing
Systems, (pp.
257-263). New Orleans, LA: Association for Computing Machinery.
28. Palmiter, S., Elkerton, J., & Baggett, P. (1991). Animated demonstrations vs. written instructions
for learning
procedural tasks: A preliminary investigation. International Journal of Man-Machine Studies, 34, 687-701.
29. Rieber, L. P. (1989). The effects of computer animated elaboration strategies and practice on
factual and application
learning in an elementary science lesson. Journal of Educational Computing Research, 5(4), 431-444.
30. Rieber, L. P. (1991a). Animation, incidental learning, and continuing motivation. Journal of
Educational Psychology,
83(3), 318-328.
31. Schneider, W., & Shiffrin, R. M. (1977). Controlled and automatic human information
processing: I. Detection,
search, and attention. Psychological Review, 84(1), 1-66.
32. Sellen, A., & Nicol, A. (1990). Building user-centered on-line help. In B. Laurel (Ed.), The art of
human-computer
interface design (pp. 143-153). Reading, MA: Addison-Wesley.
33. Shneiderman, B. (1983). Direct manipulation: A step beyond programming languages. IEEE
Computer, 16(8), 57-69.
34. Silicon Beach Software (1991). Aldus Super Card 1.6 [Macintosh program]. San Diego: Aldus.
35. Spangenberg, R. W. (1973). The motion variable in procedural learning. Audio Visual
Communication Review, 21(4),
419-436.
36. Stone, D. E., & Glock, M. D. (1981). How do young adults read directions with and without
pictures? Journal of
Educational Psychology, 73(3), 419-426.
37. Sukaviriya, P., Isaacs, E., & Bharat, K. (1992). Multimedia help: A prototype and an experiment.
In P. Bauersfeld, J.
Bennett, & G. Lynch (Eds.), Proceedings of CHI'92 Conference on Human Factors in Computing Systems,
(pp. 433-434).
Monterey, CA: Association for Computing Machinery.
38. Williams, M. D., & Dodge, B. J. (1993). Tracking and analyzing learner-computer interaction. In
M. R. Simonson &
K. L. Abu-Omar (Eds.), Proceedings of the 1993 Annual Convention of the Association for Educational
Communications
and Technology, (pp. 1115-1129). New Orleans, LA: Iowa State University.
