ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS”, Bangkok, May 23-25, 2001 
147 
EFFECTIVENESS OF MENU-DRIVEN VS. SCRIPT-BASED GIS TUTORIAL SYSTEMS 
Bin LI 
Department of Geography 
Central Michigan University 
Mount Pleasant, Ml 48858 
USA 
(517) 774-8886 
bin.li@cmich.edu 
KEYWORDS: GIS, learning, tutorials systems 
ABSTRACT 
GIS tutorial systems are largely based on graphic user interface (GUI) consisting of windows, menus, and buttons. These menu-driven 
tutorials hide the complex details of the computational processes and eliminate the need to memorize command syntax, which 
supposedly enables students to focus on learning the concepts and developing problem-solving skills. Recent studies in cognitive 
psychology, however, revealed that the menu-driven tutorials for such subjects as mathematics and physics were not as effective as the 
traditional command-based tutorials. This paper reports the preliminary results of a project that evaluate the effectiveness of menu- 
driven tutorial systems as compared to script-based systems. We found that while both tutorial systems yielded similar overall test 
scores, script-based system showed improvement on learning conceptual materials. 
INTRODCUTION 
Geographic Information Systems, or GIS, is a subject that has 
undergone rapid development in recent years. GIS provides a 
conceptual framework and associated methodology that enable 
effective representations of spatial problems. It has become an 
important component in the curriculum of geography and related 
disciplines such as geology, biology, and operational research. 
Similar to other technically oriented subjects, GIS has 
increasingly been taught with tutorials for a particular software 
package. These tutorials are based on the graphic user 
interface (GUI) that mainly constitutes menus and buttons. 
Menu selection hence becomes the main mode of operations. It 
is the dominant form of GIS tutorials. 
Personal observations and extensive research in cognitive 
psychology indicate that menu selection systems indeed may 
have significant drawbacks as learning mechanisms. Corbett, et 
al., for example, found that menu-driven data entry inhibits 
learning, compared to type-in code entry, even though students 
in the menu condition tend to like their tutor more (1991). 
Norman, on the other hand, detailed the general comparison of 
menu and command interface from the point of view of cognitive 
control and flow (1991). No such studies, however, have been 
conducted in the domain of geography and GIS. 
GIS, in the form of computer software, is a computational 
abstraction of geographic space. It represents spatial objects as 
geometric entities and their relations as spatial functions. GIS 
attempts to model the real world processes as digital data and 
flow of computational instructions. Solving problems using GIS 
require the understanding of the entire computational process. 
The advent of graphic user interface, predominantly with menus 
and buttons, makes transparent the low level operations and 
supposedly allows the user to focus on problem solving instead 
of remembering command names and syntax. While this makes 
software easier to use, the high level abstraction by menus and 
buttons blocks the user from operational details. For students of 
GIS who need to understand the concepts and logics of a 
complex computational model, menu driven systems may 
actually serve as a cognitive obstacle. And yet, we don't know if 
this is the case and to what extent menu driven systems hinders 
the learning of GIS concepts and logics. 
We believe a comparison between menu-driven and script- 
based GIS tutorials may shed some lights to the problem. 
Similar to command interface, script has well defined syntax. 
Each statement must be concise and each sequence must 
follow the operational logic. Compared with menu-driven 
interfaces, scripting provides a completely different learning 
experience. Would it be more effective to use script-based 
tutorials? This paper reports some of the preliminary findings 
from a comparison between two sets of ArcView GIS tutorials, 
with one menu-driven, the other script-based. 
The following sections include a description on the methodology, 
case studies, and a discussion. 
METHODOLOGY 
The objective of the project is to find out if script-based tutorials 
are more effective than menu-driven ones in learning GIS at the 
introductory level. Below are some of the methodological 
considerations: 
1. How should the tutorials be developed? 
2. What procedures to follow to conduct the experiment? 
3. How will the results be analyzed? 
Development of tutorials 
There are a number of considerations in developing the tutorials. 
We first have to make some assumptions about the background 
of the students. In this study, the only prerequisite for the course 
is an introductory cartography class. Programming experience 
is not required. The second consideration is the software 
platform to use. Such platform must be widely used and 
accommodate both menu-driven and scripting interfaces. The 
obvious choice is ESRI’s ArcView GIS (ESRI, 2000). ArcView 
makes its entire application programming interface available 
through the scripting language Avenue. ArcView also has by far 
the most popular menu driven tutorial system, "Getting to Know 
ArcView GIS”. The apparent approach, therefore, is to select 
the appropriate sections in the existing menu-driven tutorial then 
to develop the corresponding script versions. 
Development of the tutorial follows these procedures: 
• Formulating learning objectives 
• Identifying learning topics 
• Decomposing learning topics into tutorial topics 
• Design and implementation 
• Testing 
• Documentation 
Formulating learning objectives