ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS", Bangkok, May 23-25, 2001
Tutorial:
• Set the view, which contains the Michigan County
boundaries, to the State Plane Coordinates, Michigan
Southern Zone. (Modified from ESRI, Chapter 11,
Setting a map projection).
Quiz:
1) Use SPC as an example to describe the basic
elements of plane coordinate systems, (close book)
2) Specify a conformal projection for the Michigan Upper
Peninsular.
Group A (script-based) outperformed B on both questions. For
question (1), a common deduction for Group B students was
from incomplete answers. This was expected because the
menu-driven tutorial requires only selections of items from the
drop-down menu. All projection parameters are set
automatically. It generates no stimulus that may help re-enforce
the lecture. On contrast, the script-based tutorial requires
students type in the script, debug, and execute it. The script
itself provides a structural and precise procedure for specifying a
projection, which clearly supplement the lecture. The script also
provides a general procedure for specifying a project, which
helps Group A to achieve a higher score in question (2).
2. Thematic mapping
ILOs.
• Understand the basic concepts of Visual variables and
cartographic symbolization
• Be able to perform a specific thematic mapping
• Be able to interpret a map generated with a particular
classification method
• Be able to choose an appropriate type of classification
method
Tutorials:
• ESRI Chapter 9, Classifying and Displaying Themes
o Classifying features based on their
attributes
o Creating different legends
o Using different classifications
• ESRI Chapter 10, Symbolizing Themes
o Using markers and graduated symbols
o Using pens, fills, and colors
Quiz:
1) What are the major visual variables? Which visual
variables are more effective in representing different
cartographic objects (i.e., point, line, polygon) and
different types of measurements (i.e., quantitative,
qualitative)?
2) What are the major classification methods for
quantitative attributes?
3) Create a map that shows the geographic distribution
of population in Michigan, US.
The overall scores were not significantly different between the
two groups. For specific questions, however, there were some
notable differences. Group A (script-based) showed better
performance on question (2) while Group B outperformed in
question (3). We have not come to a full understanding as to
what attributed to such difference. At this point, we can offer
some speculations. During the tutorial session, Group A
students had to write the Avenue statement that specifies the
classification method, which is similar to the following:
aLegend.Interval (aTheme, aString, numClasses)
Students in this group were reminded to review the on-line
documentations on the related software objects and the
requests, when encountered problems. Many ended up
studying the on-line documents, which gives excellent
discussions on classification. For students in Group B, the
menu-driven tutorial presents no challenge what so ever hence
lacked the opportunity to actively discover supplemental
information.
For question (3), students had two obvious approaches, i.e., a
choropleth map or a dot density map. Most students were
unsure which one to use before hand so they began
experimenting with both methods. This gave Group B a greater
advantage. For Group A, some students were unable to
complete the task within the quiz period due to the complexity of
the programming process for legend construction.
3. Selection
ILOs:
• Understand relational and logical operations.
• Understand the common types of spatial relationships.
• Be able to solve application problems using a
combination of attribute and spatial selections.
Tutorials:
• ESRI 17 (proximity and adjacency)
• ESRI 18 (containment)
• ESRI19 (intersect)
Quiz:
1) What are the common spatial relationship types
between geographic features (vector)?
2) Three application problems (proximity and adjacency,
containment, intersect).
Group A students had no problem at all with question (1)
because all they need to do is find the on-line help for
“FtabRelTypeEnum” that lists all the spatial relationship types
supported by ArcView. Group B students, however, were mostly
familiar with the “Select by theme” dialog which shows the
relationship types only appropriate for the corresponding
themes. They did not have a clear view on the entire set of
spatial relationships. Therefore, answers from Group B were
mostly incomplete.
Results for question (2) did not show significant difference
between the two groups. There were some interesting
observations. Students in Group A seemed to know what to do.
The main problem was debugging the script which was modified
from the tutorial. They needed to make all the necessary
changes to fit the current data, which requires the student
understand the scripts. The students had limited experience
with programming therefore they got frustrated easily, which had
negative impact on the resulting scores. For Group B, the
problem was just the opposite. Students were not sure where to
begin. They started trial-and-error. Since the choices were
limited, students with a clear strategy were able to complete the
task within a short period of time. Subtle operational details,
however, caused errors in their answers. A common mistake
was selecting the wrong “theme”. Regardless how many time
the instructor highlighted the importance of activate the theme to
be selected, there were always substantial number of students
making the same mistake again and again.
CONCLUSION
Preliminary findings from this experiment were not conclusive. It
seems clear that the script-based tutorials help improving
conceptual understanding but may not be much more effective
than the menu-driven tutorial in learning problem solving. Below
are some additional thoughts.
1. It is obvious that the script-based approach requires a longer
learning curve. The learning outcome therefore may need to be
assessed in an extended period of time and through subsequent
courses.
2. It is also obvious we need to find out if a hybrid approach,
which combines script-based and menu-driven tutorials, would
perform better.
