Content uploaded by Julián Hernández-Serrano
Author content
All content in this area was uploaded by Julián Hernández-Serrano on Apr 18, 2023
Content may be subject to copyright.
Journal of Computer Assisted Learning (2003) 19, 103-114
2003 Blackwell Publishing Ltd 103
The effects of case libraries on problem
solving
J.
Hernandez-Serrano & D.H. Jonassen
University of Puerto Rico and University of Missouri
Abstract The purpose of the study was to investigate the effects of
providing access to a case library of related stories while undergraduates
solved ill-structured problems. While solving complex food product
development problems, the experimental group accessed experts’ stories
of similar, previously solved problems; the comparable group accessed
fact sheets (expository representation of stories’ content); and the control
group accessed text selected at random from a textbook dealing with
issues unrelated to the stories. On multiple-choice questions assessing
processes related to problem solving (prediction, inferences, explanations,
etc.), experimental students out-performed the comparable and control
groups. Performance on short-answer questions also assessing problem-
related skills was not significantly different, in part because of test fatigue.
Analysis of interviews identified a number of factors that students used in
deciding how to apply their study strategies, including causal factors,
grounding phenomenon, grounding in context, and outcomes.
Keywords: Case-based reasoning; Case libraries; Modelling; Problem
solving; Post-secondary
Introduction
This study examined how stories, in the form of a case library, affected
undergraduate novices’ abilities to solve complex and ill-structured problems. It is
thought that in order to develop professionals equipped to deal with the complexity
of workplace situations (i.e. to deal with ill-structured problems) they should be
supported by providing stories generated at the workplace itself. Why? Practitioners
must face a workplace filled with ‘complexity, uncertainty, instability, uniqueness,
and value conflict’ (Schön., 1993, p. 18). The workplace is perpetually overflowing
with ‘situations of practice [that] are inherently unstable’ (p. 15) and ‘are
characterised by unique events’ (p. 16). In professional practice, situations abound
with indeterminacy, value conflict, and conflicting views. In these ill-structured
contexts, the successful practitioner must be able to ‘choose among multiple
approaches to practice or devise his own way of combining them’ (p. 17).
Unfortunately, novices in schools are only allowed to work on problems that are
decontextualised and well-structured, while problems in everyday and professional
contexts are complex ill-structured (Lave, 1988; Jonassen, 1997, 2000). Unlike well-
Accepted 31 August 2002
Correspondence: David H. Jonassen, University of Missouri, School of Information Science and
Learning Technologies, 221C Townsend Hall, Columbia, Missouri 65211 Email: jonassen@missouri.edu
104 J. Hernandez-Serrano & D.H. Jonassen
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
structured problems encountered in formal education, ill-structured problems do not
have single solutions; they are open-ended and composed of many subproblems; they
frequently have many possible solutions paths; and they possess no clear beginning
or end (Jonassen, 1997; Sinnott, 1989; Kolodner et al. 1996). Therefore, when
learning business management skills, Schön (1993) recommends conducting a
‘carefully guided analysis of innumerable cases drawn from real-world contexts in
order to help students develop the generic problem-solving skills essential to
effective management’ (p. 30).
Why are stories so potentially effective? Bruner, (1986), Randall (1999) and
Polkinghorne (1988) argue that since the sharing of stories is so important, there
must be some kind of ‘narrative intelligence’ that allows humans to formulate or
follow a story. Some psychologists argue that acquiring an experience indirectly, by
listening to and understanding a story of what other people went through while
performing a task, is tantamount to experiencing the phenomenon oneself (Ferguson
et al., 1991). Therefore, presenting the experiences of skilled problem solvers in any
field to novice problem solvers in that field can provide learners with important
scripts about how to solve problems.
The theoretical rationale for the use of stories in learning is provided by case-
base reasoning (CBR). CBR assumes that cases in the form of stories are useful for
supporting problem solving by focusing the novice’s attention on what is important,
making available ideas on how to move forward, and giving grounds for preassessing
the consequences of their decisions or actions (Kolodner, 1997). The process of
understanding and solving new problems in terms of previous experiences has three
parts: recalling old experiences, interpreting the new situation in terms of the old
experience based on the lessons that were learned from the old experience, or
adapting the old solution to meet the needs of the new situation (Kolodner, 1997).
Given the lack of previous experiences among novices, substitute experiences
available through a case library are expected to augment learners’ repertoires of
experiences by connecting to the experiences of others (experts), forewarning us of
potential problems, realising what to avoid, and foreseeing the consequences of
decisions or actions. Reasoning from stories or cases supports ‘inferences necessary
for addressing the kinds of ill-defined or complex problems that arise in the
workplace, at school, and at home’ (Kolodner, 1997; p. 58).
Shinn et al. (in press) have shown that the skills required to solve well-structured
problem are different than those required to solve ill-structured problems.
Researchers and designers believe that instructional materials supporting ill-
structured problem-solving skills should incorporate ‘cases that represent probable
real-world problems in the domain, that is, that are authentic’ (Jonassen, 1997). Case
libraries made available to students while they are learning can scaffold memory by
providing representations of experiences that learners have not had. Collins (1991)
states that by having an expert reflect on how he solves problems, learners can
benefit greatly by ‘seeing’ this otherwise covert problem-solving process at work.
Case libraries can also be used to represent complexity in learning environments by
providing multiple perspectives, themes, or interpretations to the problems or issues
being examined by the learners. Instruction often filters out the complexity that
exists in most applied knowledge domains, causing learners to develop shallow
understanding of domain knowledge.
Most of the published literature on CBR and case libraries describes how they are
Case libraries 105
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
implemented in learning systems, such as intelligent tutors (Riesbeck & Schank,
1991) and goal-based scenarios (Schank et al., 1993). A paucity of empirical
research on case libraries has shown that the use of stories in problem-solving
education increases problem-solving skills, helps address misconceptions, and
contributes to the changing of attitudes (Brown, 1992; Kearney & De Young, 1995).
However, these studies were primarily concerned with either well-structured forms
of problem solving (e.g. physics problems) or with attitudinal change (e.g. whether
or not to carpool as a way to help the environment). Most research related to cases or
stories (Kass et al., 1993; Edelson, 1993; Schank et al., 1993) has been primarily
concerned with interface issues or designing complex learning environments, not
effectively isolating or assessing the potential effect that stories may have on
learning. Although CBR is a coherent and cogent form of knowledge representation,
its efficacy in supporting ill-structured problems has not been validated empirically.
Therefore, the purpose of this study was to examine the effects that (case libraries
of) stories have on novice learners’ problem-solving skills while solving ill-
structured problems. The research questions included:
• Does a case library have any effect on the problem-solving skills of
undergraduates (junior/seniors) when working on ill-structured food product
development problems?
• How do novice learners make meaning of experts’ stories while problem solving?
Method
Participants
Forty-nine juniors and seniors were recruited from two sections of a food product
development course regularly offered at a mid-sized, Catholic university located in
the north-eastern part of the U.S. These undergraduates were preparing to be hired
by food companies in sales, marketing, research, food product development, etc., so
the experiences simulated in this study were authentic. These novice learners had
previous experience with the use of cases and the case method and some experience
in providing solutions (orally, in writing, individually, and in groups) to ill-defined
food product development problems. All students volunteered to participate in the
study in exchange for extra course credit.
Materials
The goal of the instruction in this study was to solve a complex food product
development problem. The problem was presented in a 15-page narrative with 11
pages of exhibits about how the Nestlé Refrigerated Food Company introduced two
refrigerated pasta products into the U.S. market (see Fig. 1 for a screen shot of the
task environment). The narrative provided background information about the
company, the product development process as implemented by the company, the
state of the refrigerated food business, the Contadina brand and related branding
issues, market potential, positioning, manufacturing and distribution, sales, product
launch, competition, and product extensions. Three subject matter experts deter-
mined that the section ‘Market Potential’ exemplified the objectives of this study
because it laid out the largest number of ill-structured problems and issues in the
whole case. The learning issues contained in this section of the case constituted the
learning objectives for the course assignment: evaluate (scan) data on existing retail
products; determine potential market size for a product; select and use fore-casting
106 J. Hernandez-Serrano & D.H. Jonassen
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
methods, determine consumer characteristics and collect and analyse demographic
data; assess consumer perception on various product positioning; determine first year
trial volume and repeat purchase volume; make and justify product launch
recommendations and positioning recommendations; and plan market activities.
Three levels of instructional support
were provided to students as they worked
through the problem case individually.
While solving the food product
development problems, the control group
received as instructional support text
selected at random from a textbook
dealing with issues unrelated to the stories.
This treatment functioned as the control
treatment (no instructional support),
however, the possibility exists that the
irrelevance of the information may have
been misleading. The experimental group,
on the other hand, had access to a case
library of 24 stories (see Fig. 2) told by expert food product developers (all currently
or formerly employed by four different major U.S. food companies) about how they
solve similar problems. Stories were collected by presenting the practitioner with the
‘Market Potential’ section of the Contadina case one paragraph at a time.
Based on each learning issue, each practitioner was asked if the issues being
presented reminded them of a professional experience (see Jonassen & Hernandez-
Fig. 1.
Partial screen of task environment
Understanding Market Strength, Market Share and Market Size
Coca-Cola’s Share of the Stomach
In 1980, less than 15% of Coca-Cola’s sales came from outside the U.S. Having
been a dominant player in the U.S. for so long, the company was having a hard
time finding opportunities for growth despite its enormous market strength.
Coca-Cola’s CEO, Roberto Crispulo Goizueta, challenged its employees to make
the company a dominant player all over the world. Without changing the
product, he drove Coca-Cola’s marketing machine to push the product nationally
and internationally by simply modifying Coca-Cola’s definition of market
strength, which up to that moment was defined as the company’s share of the
cola market.
His strategy was very simple. He called it “the share of the stomach.” That is, if
any person around the world drank liquids 20-25 times a day, Coca-Cola’s target
was that one of those drinks be a Coke. Mr. Goizueta’s insight was that Coca-
Cola was in the business of selling liquids and it had to compete with coffee,
water, juice or any other liquid consumed by an average person around the
world. He changed the thought process of the entire organization from the
perception that the company was a dominant player in the cola market to a
company that was measuring its potential market size by focusing on Coke’s
share of the stomach. This prompted Coca-Cola executives to venture into the
business of juice, water and other liquids in order to increase Coca-Cola’s share
of the stomach.
In 1982, 80% of Coca-Cola’s sales came from the U.S. and 20% from overseas.
By 1997, Coca-Cola’s sales were 20% from the U.S.—while still holding the
dominant market position—and 80% from overseas. This had been one of the
most successful growth strategies in recent times. Coca-Cola’s success made
many executives in other business rethink their definitions for market strength,
market share and market size.
Fig. 2. Sample story
Case libraries 107
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
Serrano (2002) for a description of this process). Twenty-four of the stories were
transcribed and edited into story form and assembled into a case library. Students
were not required to access these stories, however, they were told that these stories
may contain advise that could help them to solve the problem Finally, the
comparable group received fact sheets (expository descriptions of the issues
contained in the stories, with episodic information deleted (see Fig. 3). In this study,
an abstract exposition was contrasted with concrete narratives (experiential cases).
What has not been examined in the use of case libraries is whether it is the narrative
presentation, the variety of cases, or the extra context that potentially make case
libraries effective?
Instruments
The pre-test and post-test examinations consisted of 11 multiple choice and 7 short-
answer test questions. Both tests consisted of case scenarios requiring interpretation
and application of problem solving processes, not recall of any information
presented in the stories. Figure 4 illustrates one of the multiple-choice questions.
The multiple-choice questions assessed the following skills:
Reminding (identify example/outcome): item 1
Identify/explain failure: items 2, 6
Identify/recognise problem: items 3, 5, 10
Select/recognise solution, adaptation: item 4
Identify/predict outcome: item 7
Understanding Market Strength, Market Share and Market Size
Some Facts
Sometimes companies have a difficult time focusing on growth strategies simply because
they do not really understand the nature of their markets. Typically companies define their
market strength based on the share of the market they currently hold. But sometimes this
definition can be restrictive and may prevent marketing personnel from considering great
growth opportunities.
It is critical for companies to find innovative ways to define market strength, market
share and market size. When companies develop a more appropriate definition of what
their market is, they can transform themselves simply by acquiring a new perspective on
the demand they are trying to satisfy. This in turn can prompt marketing personnel to
venture into related business areas that otherwise may go unnoticed.
Fig. 3. Sample Fact Sheet
5.*
A major U.S. food company that markets a successful line of salty snack products created a
series of snack products to tap the aging baby-boomer population who are starting to
experience higher cholesterol levels. Medical research demonstrates clearly that higher levels
of cholesterol can contribute to an individual’s incidence of heart disease and stroke. As
everyone else, this company regards this population segment as the best hope to fuel growth
for the next decade. The products are one of a kind using a new cholesterol-lowering
ingredient, Cholestra, produced by a major pharmaceutical firm. The decision was made to
sell the new snack as a line extension of one of its traditional and highly successful products.
Against all expectations, the product is not selling. The company is not able to figure out the
problem. After spending millions of dollars in R&D, the company watches in horror as its
products are being returned even before the shelf life cycle ends.
What do you think is the problem the company is facing?
a. They do not possess the know-how to market this unique snack product to baby-boomers.
b. There is probably no market for a snack health product in the U.S. or anywhere else.
c. They are not marketing the product to retailers appropriately.
d. They perceive its market strictly in terms of snack food consumption.
e. They are operating in a market that is over saturated with snack products.
* The ‘preferred’ answer is <d.>
Fig. 4. Multiple-choice test item as used in pre-test and post-test
108 J. Hernandez-Serrano & D.H. Jonassen
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
Identify/explain success: item 8
Identify/judge alternate strategies/actions: item 9
Identify information needed (on obstacle): item 11
Short answer questions (see Fig. 5) also used scenarios but asked open-ended
questions that required learners to:
Strategize (describe, state, judge): item 12
Evaluate the quality of the solution(s): item 13
Recognise problem(s): items 13, 16, 17
Predict outcome (through analogy): item 14
Evaluate (criticise) the quality of the solution(s): item 15
State/explain obstacle(s): item 16
State/justify solution(s), adapt (Kolodner, 1997): item 17
Reminding (produce example): item 18
The student responses to the short-answer tests were evaluated by three faculty
judges using a rubric. The judges were chosen for their academic and industry
expertise in food systems, for having experience in teaching a food product
development course similar to the one that was the object of this study, and for their
commitment to this task. The scoring rubric was developed in concert with the
faculty judges and an instructional design expert.
Following the post-test, two learners randomly selected from each of the three
groups (control, comparable and experimental) were asked to participate in an in-
depth interview. The contents of the interview were audio-taped and transcribed.
Procedure
All novice learners started working on the Contadina case. The whole case was read,
briefly analysed and discussed during class except the section ‘Market Potential.’
Prior to starting the study, all learners were trained on how to use the environment.
All participants completed the pre-test to assess their pre-treatment problem-
solving skills. The pre-test consisted of two parts: multiple-choice test and short-
answer questions.
Each participant in the experimental group was given a unique access code that
gave him or her access to the case library and the task environment for the ‘Market
Potential’ section of the Contadina case. The comparable group was also given a
access to the same task environment but not to the case library. Instead, they had
access to the fact sheets. The control group received access to the same task
environment but instead of the case library or fact sheets, they received additional
general information about food product development not relevant to the issues being
directly addressed by the Contadina case, the case library or facts sheets.
12.
Dr. Molten, a VP of marketing in a small food product company, has a unique strategy for
determining market potential and size, and percentage of sales. This strategy comes in handy for
this small company since it does not have many resources for elaborate market research
methods. Every time the company needs data for forecasting purposes for a product (sales,
volume, etc.), he sends out to randomly selected supermarkets in the city of Atlanta for junior
marketing personnel to measure the area (“footprint”) dedicated to similar or related products.
With this information Dr. Molten predicts future performance of products existing only on the
drawing board. Many products whose performance was predicted with the “footprint” method
have become actual successes after launch.
a. Explain why this “footprint” strategy seems to work.
b. Do you think this is a good approach for conducting market research? Why or why not?
c. Instead of the “footprint” strategy, suggest at least one other alternative and explain why it
should work for this company.
Fig. 5. Short-answer test item as used in pre-test
Case libraries 109
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
For the next three weeks, all three groups continued to work on previously
scheduled class activities. All participants were asked and consistently prompted in
class and through email to work individually on the ‘Market Potential’ section of the
case without sharing the assigned material with other students. Faculty refrained
from discussing any aspect of this section of the case during this period, referring the
learners instead to the study material.
After these three weeks, all participants completed the post-test and a short-
answer examination. These post-treatment instruments were identical in format to the
pre-treatment assessment but used different content.
Results
Descriptive statistics for the multiple choice and short answer tests are presented in
Tables 1 and 2. Analysis of variance (ANOVA) showed no significant difference
between the group performance on the multiple-choice pre-test (F2,41 = 0.51,
p > 0.05), indicating that all groups possessed equivalent levels of ill-structured
problem-solving skill at the outset of the study. After studying the case for three
weeks, the post-test scores were compared. There was a significant difference
between the group performances on the multiple-choice post-test (F2,41 = 8.76,
p = 0.001). A Tukey post-hoc test showed that the case library group out-performed
both the control and comparable groups, but there was no difference between the
control and fact sheet groups. Students who accessed the case library while solving
the ill-structured problem obtained higher scores (M = 5.43) on the multiple-choice
post-test than the groups who accessed fact sheets (M = 3.25) and random text
(M = 3.64).
Scores on the short-answer pre-test were also statistically compared. The ANOVA also
showed that there was no significant difference in short answer performances
between the groups (F2,41 = 0.612, p > 0.05). Following problem analysis, the
performance of the groups on the short-answer post-test was compared. Unlike the
multiple-choice assessment, there was no significant difference between the group
performances on the short-answer post-test (F2,41 = 0.165, p > 0.05). In fact, the post-
test scores for all groups decreased by a few points. A total of 65 min was allocated
for completion of the post-test. A substantial minority of the students failed to
complete the exam, and most students complained about the difficulty of the exam.
At least five of the students simply quit prior to the 65 minutes, all of which indicate
that the difficulty of the short answer test resulted in test fatigue. Students were not
inured to completing this type of examination requiring higher-order thinking skills.
Table 1. Multiple-choice tests descriptive results
Std. Std. CV % CV
nmean Dev. Error Min. Max. (as %) (as %)
POST- 1: Random text 14 3.6429 1.4469 0.3867 1.00 6.00 39.7 8.2
TEST 2: Fact sheets 16 3.2500 1.3416 0.3354 1.00 6.00 41.3 18.7
3: Stories 14 5.4286 1.6968 0.4535 2.00 7.00 31.3 – 39.5
Total 44 4.0682 1.7441 0.2629 1.00 7.00 42.9 4.4
PRE- 1: Random text 14 3.6429 1.3363 0.3571 1.00 6.00 36.7
TEST 2: Fact sheets 16 3.1250 1.0878 0.2720 1.00 5.00 34.8
3: Stories 14 3.3571 1.7368 0.4642 0.00 6.00 51.7
Total 44 3.3636 1.3825 0.2084 0.00 6.00 41.1
110 J. Hernandez-Serrano & D.H. Jonassen
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
Next the qualitative data collected from the interviews with the six students
following the study was analysed. Most of the interviews comprised stories about the
students’ experiences while studying and analysing the problem and the supportive
materials. An abstracted replay of the processes of using stories while studying the
material and working on the post-test was used. In order to accomplish that, the two
novice learners who were assigned to the case library group were asked to re-enact
the processes they employed using the cases. The other four novice learners who did
not have access to the stories during the course of the study (from the control and
comparable groups) were shown stories during the in-depth interviews, given time to
study them carefully, and asked to connect the stories to the issues being raised in the
Contadina case. During this process, probing questions were asked and discussion
was stimulated to gather impressions. Six novice learners were shown fact sheets and
were asked to contrast those with the corresponding stories and to apply both fact
sheets and stories to specific pre-test and post-test items. Using a q-sort method,
students’ comments were clustered into five different categories. The results of these
intense conversations are presented by category.
Strategic Factors Students’ participation in the study created a series of problems,
including how in-depth to go over material, worries about the test, the amount of
time allocated for the different tasks, and difficulties in understanding the material
well enough. For Cynthia (fact sheet), when asked about her preference between
stories and fact sheets, she sees value in a story but not for her:
I would rather read the one from the textbook [fact sheet] because it gets directly to
the point. But I don’t know if I would remember it as well.’ Stories were an inefficient
way for her to get ‘the point.
On the other hand, for Bob (story group), this apparent inefficiency is a small price
to pay given the rewards of better understanding:
Even though it [the story] might take longer to read, when I think for myself, at least I
have an easier time getting something. that is written as a story as opposed to
something that is written like a textbook [fact sheet].
For Joan, assigned to the experimental group (stories), the problem was under-
standing the stories and how she could best prepare for the post-test:
I had a problem with finding like a real narrow. what it was talking about. I would just
put the main. what I thought the main gist was, what the most important part of the
article [story] was underneath the company name so I could remember it for when I
went to go take the test.
Students who received stories were uncertain about how they would apply what they
learned from the stories to the test.
Grounding (anchoring) phenomenon Students were shown stories during the
Table 2. Short-answer tests descriptive results
Std. Std. CV % CV
nmean Dev. Error Min. Max. (as %) (as %)
POST- 1: Random text 14 27.5000 6.3442 1.6956 18.92 39.17 23.1 37.5
TEST 2: Fact sheets 16 28.6667 6.3287 1.5822 14.75 39.25 22.1 22.1
3: Stories 14 27.7976 4.5240 1.2091 18.17 33.25 16.3 – 23.8
Total 44 28.0189 5.7085 0.8606 14.75 39.25 20.4 10.9
PRE- 1: Random text 14 31.6667 5.3289 1.4242 22.50 42.25 16.8
TEST 2: Fact sheets 16 33.2604 6.0047 1.5012 17.58 43.75 18.1
3: Stories 14 30.8929 6.6058 1.7655 16.75 43.50 21.4
Total 44 32.0000 5.9483 0.8967 16.75 43.75 18.4
Case libraries 111
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
interviews and gathered their impressions about how they experienced meaning
through stories. For some of these novice learners, stories supported meaning making
by providing grounding and concreteness, that is, stories are a means of maintaining
a continuous flow of meaning. Regarding the narrative style, Samuel said:
It gives you grounds to understand on. it gives you something tangible to look at and
to actually work with rather than just a jumble of words.
For Bob (stories group), placing information in context gives it meaning:
it would be easier to digest and understand better information when you are given a
direct context to put it in. When it’s explained in something that already happened
and just tell that story I think it makes it easier.
For Cynthia (fact sheet group), stories are a means to relate a problem situation to
concrete solutions as experienced by others:
they [the students] can get a lot of those kind of cases [stories] because it is actually
something you can relate to and I think something like this probably would help with
that case [the Contadina case used in the task environment].
For Joan (stories group), the stories provided concrete descriptions of how real-life
companies do things:
I thought the articles [stories] were definitely interesting and really described the
whole process that the company had to go through to implement or do whatever they
were trying to do.
Role of Context For novice learners, the difference between experiencing the
understanding inherent in the stories will depend on whether they can easily
recognise familiar contextual elements, and whether the stories deal with issues that
are relevant to their immediate problems. If the story deals with issues directly
applicable to the problem, connection to the story is expected to be direct, reducing
the amount of effort necessary to understand and apply the story to the problem. For
Joan (stories group), the closer the story’s context is to her context, the better the
chances of understanding and remembering. However, as Joan was quick to point out
when referring to the relevance of stories used in the study, ‘It is not something that I
draw on in any of my classes.’ Even if the problem is not directly applicable to the
problem, Bob (stories group) believes that just placing information in space and time
makes it easier to experience understanding when compared to plain facts: ‘When it’s
explained in something that already happened and just tell that story, I think it
makes it easier.’ For Samuel, thinking through stories is a valuable and powerful
way of thinking that is compatible with his explaining strategies: ‘I use a lot of
metaphors and examples or experiences that I have had in trying to relate them to
what I am teaching a person or what I am trying to explain to a person that
happened. That is just the way I am. It works better for me.’
Comprehension of Stories Understanding, telling a or even attempting to remember a
story requires a conscious effort to think in narrative form. For these novices, this
invariably seemed to require that a connection to the situation be made. Some
learners (Bob, stories group), ‘found it kind of easy to make the connections. I
understood how each one, there was a Coca-Cola one and Pokémon I believe, I
understood how they related to the text [Contadina case].’ For him, it was a
different way of thinking than traditional classroom methods: ‘In an academic
setting maybe I think through telling a story and giving these facts. actual reference
points, makes it easier to understand initially.’ For others, however, the experience
of making meaning through a story is more difficult because of an inadequate
connection strategy and contextual distance. This was the case with Joan (story
112 J. Hernandez-Serrano & D.H. Jonassen
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
group) with one of the stories in which she gave up after being unable to see how the
story connected to the Contadina case: ‘I don’t think I really picked up on that.’
Outcomes from stories Pursuing a strategy of thinking through a story led to different
outcomes. For Bob, it is about reminding, committing to memory and adapting
(Schank, 1999): ‘when something is talked about in a context and actually how it
was done it is easier to remember and recall and answer.’ Bob continues by
explicating one of the primary rationales for case-based reasoning: ‘I think it almost
gives you experiences that you never really experienced but now you have seen a
certain set of problems and you know exactly how it was solved. It is kind of like
gaining experience without really being in the industry.’ Learners experience
meaning through stories in order to understand the treatment material. The most
important experience of the learners was the sense of concreteness that the stories
offered. Samuel said: ‘It gives you a real something tangible to hold on to rather
than just a definition or an explanation of what market potential is because you
could talk about market potential until you are blue in the face but if somebody
can’t see how that relates to what is happening in the real world, when they get out
there they are just going to be clueless.’
Discussion
This study focused on the role of stories as represented in case libraries in
comprehending and solving a complex ill-structured problem in an undergraduate
course on food product development. Learners were presented with a complex case
to understand. Assessments included multiple-choice and constructed-response tests
of problem solving and in-depth interviews. The improvement of the story group
performance from pre-test to post-test on the multiple-choice tests indicates that the
stories in the case library improved the choices that the students within this group
made. Given a test design where there was not one ‘correct’ answer but rather a
preferred answer (i.e. the choice that alluded to one or more lessons contained in the
stories in the case library), the students who read the stories decided in accordance
with a story’s lesson. These results were similar to earlier research (Brown, 1992;
Kearny & de Young, 1995) that stories increase problem-solving skills, address
misconceptions, and contributes to the changing of attitudes while solving well-
structured problems. Though no research has compared performance, it seems that
the introduction of case libraries is more important to solving ill-structured problems
than well-structured problems. Why? Dunkle et al. (1995) concluded that perform-
ance in solving well-defined problems is independent of performance on ill-defined
tasks, with ill-defined problems engaging a different set of epistemic beliefs. Shinn et
al. (in press ) found that solving ill-structured problems in a simulation called on
different skills than well-structured problems, including metacognition and
argumentation, and Jonassen & Kwon (2001) showed that communication patterns in
teams differed when solving well-structured and ill-structured problems.
Unfortunately, the constructed-response questions were not consistent with the
multiple-choice responses. Like the multiple-choice results, there were no
differences in pretest performance. Unlike the multiple-choice tests, there were no
differences in post-test performance either. Case libraries may have benefited
students in some ways, however, more work remains before it can be assured that
they improve ill-structured problem solving in general. That the performance of all
Case libraries 113
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
groups decreased on the post-test performance can only be explained by fatigue.
While both the multiple-choice and short-answer tests both assessed higher-order
thinking required for problem solving, the short-answer tests was perceived to be
significantly more difficult because the students had to construct responses. Only
65 min were allotted for completion of the post-test. None of the students completed
the constructed-response items. Many simply gave up trying, complaining that the
task was impossible. Retrospectively, the post-test should have been shortened or
more time provided for its completion.
The in-depth interviews provided some important perspectives on how students
use case libraries to understand problems. Clearly, many factors can influence
learning strategies. In this case, both the task and the methods were different than
those normally encountered in undergraduate classrooms. For example, Joan was
more inclined to use her personal experience in making a choice even though two
weeks after studying the material she could remember some of the stories: ‘I think it
does have a lot to do with just my experience or what I know of the product or
products better like that product.’ For Bob the determining factor for his choice in
one post test item was something he learned in school before participating in this
study: ‘Because I saw that they had done BASES research and we were taught when
we first learned about this that you should always.’ Case-based reasoners would
argue that students in the case library group were reminded of a case library story
after reading a post test item and that this in turn prompted them to make a choice
consistent with the story lesson (Kolodner et al., 1996; Schank, 1999).
However, varying descriptions of how stories were used or could be used were
related by the interviewees. Using stories as adjunct aids represents a novel approach
to learning, one that inherently makes more sense for some students while confusing
others. It does not seem that these differences are mediated entirely by individual
differences, however, individual differences invariably intervene in any learning.
Students unfamiliar with both the nature of the instruction (stories) or the outcome
(problem solving) do not possess scripts for how to apply the stories to more
meaningful learning activities. These students have not developed any connection
strategies. The kind of analogical reasoning necessary for connecting stories to
problems is rarely practised in formal education, where recalling information
engages a different set of learning strategies. Despite that, a number of the students
comprehended the importance of stories. Even Cynthia, who was more committed to
preparing for the test using facts sheets, recognised the salience of some stories by
recalling the ‘story of the stupid manager who didn’t look at his research and who
was trying to compete with competition, the competition was coming out. Didn’t do
research. I will remember that one rather than the other ones.’ Future research will
provide training sessions on how use stories while engaged in meaningful learning.
Although the use of stories appears intuitive in everyday contexts, it contradicts
student learning in formal education contexts, so practice is necessary.
References
Brown, D.E. (1992) Using examples and analogies to remediate misconceptions in physics:
Factors influencing conceptual change. Journal of Research in Science Teaching, 29, 1,
17–34.
Bruner, J. (1986) Actual Minds, Possible Worlds. Harvard University Press, Cambridge.
114 J. Hernandez-Serrano & D.H. Jonassen
2003 Blackwell Publishing Ltd, Journal of Computer Assisted Learning, 19, 103-114
Cho, K.L. & Jonassen, D.H. (2002) The effects of argumentation scaffolds on argumentation
and problem solving. Educational Technology: Research and Development, 50, 3, 5-22.
Collins, A. (1991) Cognitive Apprenticeship and Instructional Technology. In Educational
Values and Cognitive Instruction: Implications for Reform. (eds. L. Idol & B.F. Jones),
pp. 121–138. Lawrence Erlbaum Associates, Hillsdale, NJ.
Dunkle, M.E., Schraw, G. & Bendixen, L.D. (1995 April) Cognitive processes in well-defined
and ill-defined problem solving. Paper presented at the annual meeting of the American
Educational Research Association, San Francisco, CA.
Edelson, D.C. (1993) Learning from Stories: Indexing and Reminding in a Socratic Case-
Based Teaching System for Elementary School Biology. (Doctoral Dissertation,
Northwestern University).
Ferguson, W., Bareiss, R., Birnbaum, L. & Osgood, R. (1991) ASK Systems: An Approach to
the Realization of Story-Based Teachers. Journal of the Learning Sciences, 2, 1, 95–134.
Jonassen, D.H. (1997) Instructional Design Models for Well-Structured and Ill-Structured
Problem-Solving Learning Outcomes. Educational Technology, Research and
Development, 45, 1, 656–694.
Jonassen, D.H. (2000) Toward a design theory of problem solving. Educational Technology:
Research and Development, 48, 4, 63–85.
Jonassen, D.H. & Hernandez-Serrano, J. (2002) Case-based reasoning and instructional
design. Educational Technology: Research and Development.
Jonassen, D.H. & Kwon, H.I. (2001) Communication patterns in computer-mediated vs. face-
to-face group problem solving. Educational Technology: Research and Development, 49,
10, 35–52.
Kass, A., Burke, R., Blevis, E. & Williamson, M. (1993) Constructing Learning Environ-
ments for Complex Social Skills. Journal of the Learning Sciences, 3, 4, 387–427.
Kearney, A.R. & De Young, R. (1995) A Knowledge-Based Intervention for Promoting
Carpooling. Environment and Behavior, 27, 5, 650–678.
Kolodner, J.L. (1997) Educational Implications of Analogy — A View from Case Based
Reasoning. American Psychologist, 52, 1, 57–66.
Kolodner, J.L., Hmelo, C.E. & Narayanan, N.H. (1996) Problem-Based Learning Meets Case-
Based Reasoning. In Proceedings of the International Conference on the Learning
Sciences. (eds. D.C. Edelson & E.A. Domeshek) pp. 180-187. Northwestern University,
Evanston, IL.
Lave, J. (1988) Cognition in Practice: Mind, Mathematics and Culture in Everyday Life.
Cambridge University Press, New York.
Polkinghorne, D. (1988) Narrative Knowing and the Human Sciences. State University of
New York Press, Albany.
Randall, W.L. (1999) Narrative Intelligence and the Novelty of Our Lives, Journal of Aging
Studies., 13, 1, 11–28.
Riesbeck, C.K. & Schank, R.C. (1991) From training to teaching: techniques for case-based
ITS. In Intelligent Tutoring Systems: Evolutions in Design. (eds. H. Burns & J.W. Parlett)
pp. 177–193. Lawrence Erlbaum Associates, Hillsdale, NJ.
Schank, R.C., Fano, A., Jona, M. & Bell, B. (1993) The Design of Goal-Based Scenarios.
Journal of the Learning Sciences, 3, 4, 305–346.
Schank, R.C. (1999) Dynamic Memory Revisited. Cambridge University Press, Cambridge.
Schön, D.A. (1993) The Reflective Practitioner – How Professionals Think in Action. Basic
Books, New York.
Shinn, N., Jonassen, D.H. & McGee, S. (in press) Predictors of well-structured and ill-
structured problem solving in an astronomy simulation. Journal of Research in Science
Teaching.
Sinnott, J.D. (1989) A model for solution of ill-structured problems: Implications for
everyday and abstract problem solving. In Everyday Problem Solving: Theory and
Applications (ed. J.D. Sinnott) pp. 72–99. Praeger, New York.