1.
Web-Based Representations of
Ill-Structured Problems: Analysis
and Implications for Practice
Nada Dabbagh
Professor, Learning Technologies
George Mason University
Virginia, U.S.A.
PBL 2016, Sao Paulo

2.
Ill-Structured Problems (ISP)

3.
Ill-Structured Problems (ISP)

4.
Examples of ISP

5.
ISP Representation
 How does one determine the degree and type of structure with
which to arrange information for an ISP, and does it matter?
 Does problem structure impact student learning?
 Does problem structure impact problem-solving?
 Does problem structure impact knowledge assembly?
 Examine how students interact with two types of ISP
representations, hierarchical and heterarchical web-based
designs, in order to better understand problem structure
 User Experience (UXD)

6.
Hierarchical versus Heterarchical
Web-Based Designs
 Hierarchical design:
 tree-like, top-down structure
 resulting in several navigation levels or layers
 content generally organized into logical sections often by
major topic area
 Heterarchical design:
 network-like, relational, conceptually indexed structure
 embedded links are sprawled throughout the content,
resulting in an unstructured navigation path with random
links jumping from one topic to another (referential access)
 Depth versus Breadth
 Topical versus Contextual

7.
ISP Used in this Research
 Design Problem/Dilemma – Medical Malpractice
 Authentic, ambiguous, multiple stakeholders, incomplete information,
emergent, elusive
 Concerning the issue of “informed consent” prior to surgery:
 The informed consent problem was used in an ID context:
 Instructional Design (ID) is a dynamic process of problem
understanding and problem solution
 ID cases are archetypal examples of ill-structured problems
 Students are responsible for designing instruction and assessment
that legally verifies that patients were fully “informed”
 ISP was used in a PBL environment

8.
Tracking Program

9.
Tracking Program

10.
First Study - Exploratory
7 participants – 2 groups
 Hierarchical (H) Design of ISP:
 Exploration was found to be more comprehensive in terms of the #links
visited/revisited, # of unique links visited, and total # of interactions generated
 Heterarchical (R) Design of ISP:
 Average time spent on a page was 1.5 times the average time spent on a page
in the H case design
 Evidence of more collaboration between group members (97% versus 86% of
the decisions were group-based)
 Group problem solving strategy was clearly one of discussion and reasoning
through the problem information
 Perceptions of structure in terms of organization of links, resource information,
and navigation, were overall more positive
 Both H&R: Perceptions of ill-structuredness, authenticity, meaningfulness,
and relevance of the learning task was equal in both groups

11.
Second Study – Case Study
14 participants – 4 groups
 Problem solutions developed in response to the heterarchical (R) design
were more cogent and viable than problem solutions developed in
response to hierarchical design
 explicit links among multiple factors in the case
 focused on building from what is known
 recommendations described in tentative terms and subject to change as
additional information becomes available
 explicit consideration of implementation and/or effects of recommendations
 Problem solutions developed in response to heterarchical (R) design
provided evidence of a heuristic problem-solving process facilitating
the identification of an expert-like solution to the case and the articulation
of learners’ understanding and application of grounded and engaging
instructional designs

12.
Third Study – Experimental Design
43 participants – 8 groups
 How does team exploration (information seeking behavior) in the
heterarchical (R) ISP design compare to team exploration in the
hierarchical (H) hypermedia case design?
 How do team-based heterarchical (R) ISP design solutions compare
with team-based hierarchical (H) ISP design solutions? Are the
heterarchical ISP solutions more expert-like, comprehensive, cogent,
and viable than the hierarchical ISP solutions?
 Did the online discourse reveal differences in participants’ problem
solving behaviors (e.g., novice vs. expert-like) between teams that
interacted with the heterarchical (R) ISP design and teams that
interacted with the hierarchical (H) ISP design?

13.
Results
 Exploration, information seeking behavior:
 team exploration in the R case design was more comprehensive (covered
more case content) and decisive (minimal uncertainty with respect to the
usefulness of the visited links to solving the case)
 average time spent on a page by the R teams was 2 minutes versus 4.75
minutes for the H teams which explains why exploration was overall more
comprehensive in the R case design
 Research suggests that user interaction with R case designs requires
more cognitive effort in terms of navigation decision making resulting in
more time spent on a page and hence less links visited overall;
however, this was not the case in this study
 The results of this study were not consistent with the results of the first
study with respect to number of links visited (comprehensiveness) and
time spent on a page

14.
Results
 Problem solving behavior/design solutions:
 Initial/draft ISP solution means were higher for the hierarchical (H)
teams, however, this trend was reversed for the final ISP solutions
 heterarchical (R) ISP interactions and solutions demonstrated more
expert-like
 Users interacting with a simplified, externalized, or highly organized
hypermedia navigation structure, as is the case in the hierarchical (H)
ISP design, may initially perceive the learning task as well-defined,
while students or users interacting with a more contextual (natural) or
unstructured hypermedia navigation structure, as is the case in the
heterarchical (R) ISP design, may initially perceive the learning task as
ill-structured or ill-defined

15.
Conclusions
 Collaborative decision making and problem solving are a staple of PBL and
interactions involving these processes are becoming largely technology
supported and mediated through online learning technologies
 The results of this series of studies that examined the effects of problem
representation of ISP or cases on student collaborative information seeking and
problem solving behaviors seem to support the tension that exists in CSCL
research (Computer Supported Collaborative Learning)
 While CSCL researchers argue that group tasks should be complex and ill
structured to maximize group interactions, recent CSCL research is increasingly
tempering this position and arguing that clearer and greater structure is needed
to scaffold collaborative tasks and problems
 More research on UX (User Experience) as this relates to the design of
technology supported ISP representations and CSCL in PBL

16.
For more information please contact
Dr. Nada Dabbagh
at
ndabbagh@gmu.edu