Journal of Science Education and Technology, Vol. 16, No.

4, August 2007 (
2007)

DOI: 10.1007/s10956-007-9054-6

Inquiry-Learning with WebLab: Undergraduate Attitudes

and Experiences

Judith Fischer,1,3 Rudolph Mitchell,1 and Jesus del Alamo2

The Microelectronics WebLab at MIT allows students to do actual (not simulated) laboratory
research on state-of-the art equipment through the Internet. This study assesses the use of
WebLab in a junior-level course on microelectronic devices and circuits in 2004–05 and 2005–
06. In quantitative surveys and qualitative interviews, students and faculty reported that
WebLab was effective as an instrument of learning, and grew more so with refinements of the
program. WebLab allowed undergraduates to learn at their own pace and on their own
schedules. It enabled them to use different processes of learning (intuitive, visual, abstract),
and it gave them an opportunity to link individual and collaborative effort in creative com-
binations. Online laboratories on this model have broad applications in the experimental
sciences and in other research-oriented disciplines.

KEY WORDS: WebLab; microelectronics; engineering education; online laboratories; inquiry learning;
assessment

INTRODUCTION: THE DEVELOPMENT OF in undergraduate and graduate microelectronics

WEBLAB courses at MIT, and also in overseas programs. Since
1998, more than 3600 students on four continents
This inquiry began not with a theory, or a model, have worked with it in university-level courses (MIT
or a fixed hypothesis, but rather with a set of open- iCampus iLabs website, 2006).
ended questions about student attitudes toward their WebLab consists of laboratory instruments for
own learning experiences with WebLab in an micro- the characterization of microelectronic devices, to-
electronics course at MIT during the Fall and Spring gether with computer hardware and software that
semesters of 2004–05 and the Fall of 2005 (Fischer, make this equipment accessible to users through the
2005a, b; Mitchell, 2005a, b). World Wide Web (Figure 1).
Opportunities for laboratory research by engi- The system is based on iLabs Shared Architec-
neering students have been limited by the cost of ture, a design that separates online labs into modules
equipment, limitations of space and other factors. connected by a Web service broker (Hardison et al.,
WebLab was developed through the MIT iCampus 2005). It can accommodate a large number of users
project as a response to this problem. It has been used through a queuing system for access to the server and
test equipment. A switching matrix allows remote
selection of one of eight possible devices. The process
of remote control, combined with a local database,
1
Teaching and Learning Laboratory, MIT, Building 5-122, permits flexible and efficient management of multiple
77 Massachusetts Ave, Cambridge, MA, 02139, USA users (Figure 2).
2
Department of Electrical Engineering and Computer Science,
MIT, Cambridge, MA, USA
Students can use WebLab at a time and place of
3
To whom correspondence should be addressed; their choice. Accessing WebLab remotely, students
E-mail: jfisch@[Link] perform real-time experiments and collect data about

337
1059-0145/07/0800-0337/0
2007 Springer ScienceþBusiness Media, LLC
338 Fischer et al.

Fig. 1. WebLab hardware.

Fig. 2. iLab shared architecture.

the actual operation of real microelectronic devices. Inquiry-learning has been a major theme in the
They can also compare their results with theoretical history and philosophy of education since the work
models presented in the course (del Alamo et al. 2003). of John Dewey (1859–1952) and Jean Piaget (1896–
1980). It is flourishing today in many forms and
INQUIRY-LEARNING THROUGH WEBLAB under many names, such as discovery learning,
problem-solving approaches, and constructivism.
A major purpose of WebLab is to promote Always, it centers on the active engagement of the
authentic, experiential, hands-on, inquiry-based learner in an open process of questioning and dis-
learning. It does so through genuinely open-ended covery. An expanding literature has abundantly
laboratory experiments, which benefit learners in documented the value of this approach. Authentic
ways that rarely happen in lectures, or simulated problem-solving experiences are especially helpful in
research. In a more open process of inquiry and science and engineering, and are also useful in most
discovery, it also helps students to learn about the empirical disciplines (Albert, 2000; Anderson, 2002;
learning process itself (Jeppson, 2004; Bransford et al, Bruner, 1987; Felder and Brent 2003; Henry, 1996;
2002; Dunbar, 2001; Henry, 1996). Dewey, 1938; Feisel and Rosa, 2005; Martin-Hansen,
Inquiry-Learning with WebLab 339

2002; Minstrell, 2000; National Research Council, THE USES OF WEBLAB IN A MICROELEC-
2000; Piaget, 1928; Prince, 2004). TRONIC COURSE
WebLab is very much in this tradition. Students
can use it to devise their own experiments, within the This project assesses undergraduate perceptions
framework of a course syllabus or a curricular of the effectiveness of WebLab in a junior-level
objective. In its microelectronic applications, it al- course in Microelectronic Devices and Circuits at
lows students to construct their own test vectors, MIT during three semesters; the spring and fall of
measure current-voltage characteristics of a device, 2004–05 and the fall of 2005–06. The purpose of the
and extract specific parameters. They can also com- course was to teach students how to use incremental
pare their own experimental data with theoretical and large-signal techniques to analyze and design
models. transistor circuits and their components. It covered
WebLab has been designed to challenge the following topics: modeling of microelectronic
students. They are asked to design test-vectors that devices, basic microelectronic circuit analysis and
respond to assignments. Students must also learn to design, physical electronics of semiconductor junc-
deal with issues of measurement such as range, tion and MOS devices, relation of electrical behavior
point-distribution, and compliance. Another form to internal physical processes, development of circuit
of learning can be gained from the process of models, and understanding of their uses and limita-
debugging errors in submitted test vectors. The tions. The course syllabus and other information
graphical component of the WebLab interface about the 2005–06 fall term Microelectronics course
enables students to select variables and graph re- can be found at MIT OpenCourse Ware (2005).
sults on either a linear or logarithmic scale. They WebLab assignments in the course asked stu-
can also download quantitative data, analyze it with dents to work with three microelectronic devices.
standard software packages, and compare their Two were common transistors: a MOSFET (metal-
results with theoretical models. That process in turn oxide semiconductor field-effect transistor) and a BJT
encourages reflection about non-idealities of the (bipolar junction transistor). The third was a PN
devices and shortcomings of the models. The flexi- Diode (an electronic equivalent to a check valve). In
bility of the WebLab interface allows students to typical WebLab experiments, students were asked to:
follow their curiosity beyond initial assignments, to • collect data for a given microelectronic device, on its out-
explore other questions in an independent way put, transfer or backgate characteristics
(Figure 3). • graph observed device characteristics

Fig. 3. Screen shots of the graphical user interface of the MIT Microelectronics WebLab.
340 Fischer et al.

• download and analyze data on the studentÕs personal com- interface found to be for the collection of data? How
puter did students analyze the data? How did they assess
• extract device parameters from the data
their learning experiences? How did they feel about
• program a model of device operation based on equations
presented in lecture further applications of WebLab?
• compare real measurements with theoretical model To test the reliability of each scale, coefficient
• measure other characteristics of interest to the student alphas and item-total correlations were calculated,
following methods discussed by Streiner and Nor-
ASSESSING WEBLAB: RESEARCH DESIGN man, with reliability procedures in SPSS 10.0 for
AND METHODOLOGY MacIntosh (SPSS, 1999; Streiner and Norman, 1995;
Norman and Schmidt, 1992; Schmidt et al. 1999).
The purpose of this project was to study the Alpha correlations on all four scales were robust,
process of WebLab learning as perceived by students ranging in value from .80 to .90. Results were tabu-
at MIT. All were undergraduates enrolled in the same lated and compared by semester, to assess change in
microelectronics course within the Department of perceived effectiveness of WebLab. We did not ana-
Electrical Engineering and Computer Science. The lyze data by major, gender, age, or ethnicity. Our
research had two major components: quantitative interest was in the experience of individual learning,
surveys and in-depth qualitative interviews with a rather than in the performance of groups.
small number of individual students.
WebLab Interviews
WebLab Surveys
The second part of the assessment consisted of
The surveys were developed in the fall semester in-depth interviews with a small number of students
of 2004, and used in the spring semester of 2005, and in each semester (Table I). An interview protocol was
again in the fall semester of 2005. At the last lecture developed in the fall semester of 2004, on the basis of
in each semester, all students who attended were conversations with faculty and pilot-interviews with
asked to complete the survey in writing. Every stu- nine students. In the spring of 2005, 27 students were
dent who was present at the last lecture did so: 37 of selected at random and invited to be interviewed on a
63 students enrolled in the spring semester of 2005; 33 voluntary basis. Twelve agreed to participate. The
of 63 enrolled in the fall of 2005. interviews were repeated in the fall semester of 2005,
The spring survey included 44 questions about when interview-requests went to 51 students (again
the use of WebLab, the collection and analysis of selected at random). Of that number, 22 agreed to
data, and general assessments of Web-Lab experi- participate, and all were interviewed.
ence. On the basis of further study in the summer of The interviews were held in the offices of the
2005, the fall survey repeated all of those questions, Teaching and Learning Laboratory at MIT. Each
and added 13 others on time-commitment and work session was conducted individually, and recorded
habits, for a total of 57 survey-questions. Each digitally by iPod, with the permission of the student.
question framed a statement and asked students to The median length of interviews was approximately
register degrees of agreement or disagreement on a 30 min. The range was from 25 min to more than an
seven-point Likert scale. hour.
For purposes of analysis, the survey answers Interviews were guided by a set of 39 questions
were scaled in four groups around the following on individual patterns of use of WebLab, strengths
general issues: How effective was the WebLab and weaknesses of hardware and documentation, the

Table I. Student Interviews

When interviewed

Course enrollment # Students interviewed Total # interviews Mid-semester End of semester After end of semester

Fall 2004 63 9 9 9
Spring 2005 90 17 19 9 10
Fall 2005 65 22 22 10 12
Total 218 48 50 19 22 9
Inquiry-Learning with WebLab 341

use of models and empirical questions processes of assignment to the last week in the term left students
data collection and analysis, the use of graphs, the with a feeling of frustration at end of the course.
utility of WebLab in meeting course objectives, op- These problems had an impact on quantitative eval-
tions for individual and collaborative study, and uations in the WebLab Survey.
assessments of WebLab as a learning experience. Based on assessments of WebLab in the 2004 fall
These questions established the initial frame of each and 2005 spring semesters, many changes were
interview, but did not constrain it. The conversations introduced in the summer of 2005. The documenta-
ranged far beyond them in open, creative and free tion was overhauled and the manual was rewritten.
flowing discussions. The object was not to obtain On-line tutorials were developed, and a class tutorial
quantitative data, which had been the purpose of the session was added at the beginning of the semester.
surveys. The interviews were conceived as open- FAQs (frequently asked questions) were added and
ended, heuristic inquiries into individual attitudes, posted on the website. A number of small bugs were
experiences, and judgments. One goal was to deepen identified and fixed. Large WebLab assignments were
understanding of empirical results in the surveys. divided into a larger number of more specific
Another was to explore creative possibilities for the assignments, which were distributed with care
refinement of WebLab in the present, and the devel- through the semester. Demonstrations of WebLab
opment of WebLab in the future. and device characteristics were added to lectures (del
Alamo, informal conversations [2005–2006]; MIT
STUDENT RESPONSES: PROBLEMS OpenCourseWare, 2005).
OF ACCESS TO WEBLAB These changes improved the operation and
effectiveness of WebLab. During the third semester in
In surveys and interviews during the first and the fall of 2005, students reported that WebLab ran
second semesters during the fall and spring of 2004– more smoothly. Documentation was more clear, and
05, undergraduates had mixed opinions about the also more accessible. Students reported fewer tech-
accessibility of WebLab. They described it in general nical problems, and they were progressively more
as ‘‘user friendly,’’ ‘‘cool,’’ ‘‘neat,’’ or ‘‘fun.’’ These positive about the experience. No one who was
judgments were tempered by reports of specific interviewed in the 2005 Fall semester reported diffi-
problems with the WebLab interface. In the spring culty in accessing WebLab. Many found the program
semester of 2005, students found the program docu- so clear they used the documentation rarely, if at all.
mentation to be insufficient to their needs. Some One student said, ‘‘When itÕs time to do the WebLab,
couldnÕt find it. Others thought it was flawed. They you just connect it [and begin.].’’ Surveys indicted
complained that the server failed in periods of very that these positive assessments increased with time.
heavy demand, especially during the spring semester A related change occurred in the number of
of 2005. A rescheduling of the second WebLab assignments (Table II). WebLab can be used at many

Table II. WebLab Assignments in Three Semesters

Week due Content WebLab % of assignment WebLab % of final course grade

Fall Semester 2004

7 MOSFET characterization 100 10
Total 10
Spring Semester 2005
2 PN diode characterization 100 5
13 Device characterization and circuit design project 50 2.5
Total 7.5
Fall Semester 2005
2 PN diode characterization 40 1.1
3 PN diode characterization (cont.) 40 1.1
4 PN diode characterization (cont.) 10 1.3
6 MOSFET characterization 50 1.4
7 MOSFET characterization (cont.) 45 1.3
10 BJT characterization 50 1.4
Total 6.5
342 Fischer et al.

12 that controlling the WebLab remotely from a PC did

Percent of
10 final grade not interfere with their learning. Bugs in the WebLab
8 based on were viewed as less frustrating (Table III).
WebLab
6 Students liked being able to choose when and
Assignments
4 using where to use WebLab. They preferred to work pri-
2 WebLab marily on their own computers. Students also used
0 computer clusters on campus. Only one student in the
Fall 2004 Spring Fall 2005 fall semester reported that he worked all of the time
2005 with another student in a computer cluster. Other
Fig. 4. Number of assignments using WebLab and percent of fi-
students reported that they worked with other stu-
nal grade based on WebLab. dents at computer clusters, especially when confused
by part of the assignment, or wanting to verify their
results.
Working remotely was not difficult for them. In
different levels of frequency and intensity. In the first each semester students commented about the
semester studied here (fall 2004), students completed importance of being able to work with real devices
only one WebLab project near the middle of the used by professional engineers, rather than using
semester (the seventh of thirteen weeks). In the sec- computer simulations. They found data collection
ond semester (spring 2005) there were two projects, efficient and easy, and commented about the time it
one near the beginning, and the other near the end of saved them in their busy schedules. One student
the semester. During the third semester (fall 2005) noted that ‘‘WebLab did in minutes what would take
students used WebLab six times, nearly two-thirds of hours by hand.’’
the homework assignments. Students in both courses found WebLab to be a
In successive semesters, the use of WebLab also new sort of hands-on lab experience. One student
increased as a proportion of homework assignments said, ‘‘ItÕs a neat way to do experiments without
and projects, and declined as a percent of final grades having to toil over the wiring and complications.’’
(Figure 4). Students noted that WebLab was quite different
The impact of technical improvements appears from other laboratory work. One noted the impor-
in student responses to the WebLab Survey tance of actual inquiry experience. He wrote, ‘‘We-
(Table III). Rankings on WebLab user-friendliness, bLab is a good opportunity to teach people how to
clarity, and coherence of data presentation were do good analysis, which I think is important because
substantially higher in the fall 2005 semester than in you donÕt get that experience with regular problem
the previous semester. Students agreed more strongly sets.’’

Table III. Self-Reported Student Experience with the Use of the WebLab

Spring 2005 Fall 2005

Degree of agreement or disagreement to the following statements about WebLab Mean SD n Mean SD n %Change
a
WebLab is user friendly 3.78 1.51 36 5.18 0.95 33 +37.04
WebLab documentation is clear and effectiveb 2.95 1.74 36 4.45 1.39 33 +50.85
The graphical interface enhances the learning experiencec 4.08 1.64 36 5.76 1.06 33 +41.17
The graphical interface presents data in a clear and coherent mannerd 4.39 1.61 36 5.58 0.90 33 +27.10
Controlling WebLab remotely from a PC does not interfere with my learning 4.76 1.66 36 6.03 1.16 33 +26.68
The bugs in the WebLab make the lab a frustrating experiencee 3.95 1.52 36 3.09 1.68 33 âˆÕ12.30

Student survey responses, ranked on a 7-point Likert scale, ranging from ‘‘1’’ (= strongly disagree) to ‘‘7’’ (= strongly agree). Intervening
values have no descriptors.
a
In the spring of 2005, this question was worded differently: ‘‘I found the WebLab user friendly.’’
b
In the spring of 2005, this question was worded differently: ‘‘The WebLab documentation was well designed.’’
c
In the spring of 2005, this question was worded differently: ‘‘The graphical interfaced provided an effective experience.’’
d
In the spring of 2005, this question was worded differently: ‘‘The graphical interface presented data in a coherent manner.’’
e
In the spring of 2005, this question was worded differently: ‘‘I found the bugs in the WebLab frustrating.’’
Inquiry-Learning with WebLab 343

Table IV. Self-Reported Student Learning Behavior

Fall 2005

Mean SD n

How well do the following phrases describe the mental operation used while gathering data when running the WebLab Experiments?
Visualizing how devices work 4.33 1.27 33
Reasoning about the behavior of devices 4.91 1.28 33
Applying concepts discussed in lecture or readings 4.82 1.38 33
Thinking about relationships among the variables 5.12 1.34 33
How well do the following phrases describe the mental operations used completing the data analysis?
Visualizing how devices work 4.56 1.56 32
Using intuition to understand devices 4.78 1.52 32
Formulating concepts about the behavior of devices 4.81 1.38 32
Thinking about relationships among variables 5.41 1.32 32
Thinking about how the lab experiences relate to material previously learned 4.88 1.34 32

Student survey responses, ranked on a 7-point Likert scale, ranging from ‘‘1’’ (= poorly) to ‘‘7’’ (= extremely well). Intervening values have
no descriptors

USING WEBLAB: DATA GATHERING One noted that others could ‘‘help catch mistakes
AND ANALYSIS that youÕd made,’’ or ‘‘explain something you didnÕt
get the first time around.’’ Another considered ‘‘col-
WebLab assignments typically included two laborative work absolutely mandatory, with a very
tasks: data gathering and data analysis (Table IV). clear correlation to doing well in a course.’’ A third
The students reported that collection of data was agreed that ‘‘two heads are better than one.’’
simple and easy. They worked alone on this step, and In both interviews and surveys, students ob-
could do it quickly, although students occasionally served that WebLab provided opportunities to de-
had long waits for data retrieval, typically on the velop skills of critical and creative thinking. One
night before the assignment was due. ‘‘ItÕs pretty student said, ‘‘I definitely took stuff away from We-
quick to get the right data, and itÕs Ôoutput-edÕ in a bLab I couldnÕt have taken away from just problems.
nice convenient form,’’ said one student. In other problems in the assignment it seemed like
Data analysis was the next stage. Students ap- you found the right formula to match the problem,
plied concepts discussed in lecture, formulated con- plug in the numbers and got the solution. With We-
cepts about the behavior of the devices, examined bLab you actually thought about the problem and
relationships among variables, and considered how how to set it up.’’ Another student commented,
lab experiences related to material that they had pre- ‘‘Even when I was working on one of the problems
viously learned. They found that the task of analysis not requiring WebLab, I could approach the problem
was straightforward, and were helped by clear in a more logical way, because I already had a picture
instructions, but those who were pressed for time of the sort of graph to expect.’’
complained that the analytic process was ‘‘tedious.’’ The graphs generated by the program were
Even as most students worked alone at their thought to be especially helpful as aids to analysis
computers for the collection of data, they also col- (Figure 5). Everyone liked the WebLab graphing
laborated with others on particular tasks, such as feature, calling it ‘‘absolutely necessary,’’ ‘‘the best
sharing questions, exchanging ideas and exploring part,’’ and ‘‘very valuable.’’ Several students com-
possible solutions. One student said, ‘‘Once IÕd col- mented that ‘‘a picture is worth a thousand words,’’
lected the data, I worked with others trying to figure and one added, ‘‘And thatÕs very true, even in things
out what to do with it.’’ Another commented that that are so mathematical.’’ Others thought the graphs
‘‘after being stuck for a while, IÕd ask one of my helped them to learn if they were on the right track or
friends.’’ A third noted, ‘‘Occasionally IÕd look at not, and to see if their data was in the right range.
someone elseÕs work just to make sure the graphs Students used their graphs in an interesting
looked right…sanity check, mostly.’’ variety of ways. For one student graphs helped to
In these various ways, students agreed the ‘‘capture the image in my mind.’’ Another student
importance of collaboration to their own learning. said, ‘‘Graphing was useful because you could
344 Fischer et al.

Fig. 5. Comparison of measured and modeled current-voltage characteristics of a MOSFET as provided by a student.

visualize difference between components. The best useful than strictly learning material anyway.’’ Others
part was that you could visualize those relation- used this part of the program differently. ‘‘WebLab
ships.’’ One student reported that to him ‘‘holding graphs help me,’’ commented another student. ‘‘I try
things in my head visually is second nature.’’ Others to think about it conceptually—what would change if
worked in different ways. One described himself as a I did this. I would change a parameter and make a
‘‘numbers person rather than a visual learner.’’ He new graph to see how the variable changes.’’
continued, ‘‘If I had a choice between graphs and
numbers I would always pick a formula.’’ At the LEARNING WITH WEBLAB: UNDERSTANDING
same time he noted that ‘‘visual images help develop CONCEPTS IN MICROELECTRONICS
an intuitive sense.’’
The graphs also helped students understand the In the language of the course syllabus, students
behavior of various devices in yet other ways. One were expected by the end of the term to be able to:
said that ‘‘WebLab showed the behavior of real de- • Explain and apply basic concepts of semiconductor physics
vices and the difference between the real devices and relevant to devices.
the ideal model.’’ Another thought that ‘‘the graphs • Describe, explain, and analyze the operation of important
help you get a feel about how a device is going to semiconductor devices in terms of their physical structure.
• Explain describe, and use physics-based models of varying
behave.’’ A third noted that ‘‘A lot of learning hap- levels of complexity, select models appropriate to a specific
pens because you see how things fit together and what need, and apply those models to analyze multi-component
needs to be adjusted.’’ circuits.
The students spoke at length about the impor- • Analyze and design microelectronic circuits for linear
tance of WebLabÕs graphing capabilities in the amplifier and digital applications.
• Confront integrated device and/or circuit design problems,
development of intuitive understanding of micro- identify the design issues, and develop [Link],
electronics. For some, visual images became an students were confident that they had reached these goals.
intuitive shorthand for important relationships and They felt that they had gained an understanding of basic
ideas. One student mentioned that intuition was microelectronic principles. They believed that they had
stressed in the course. He said, ‘‘WebLab graphic mastered the operation of semiconductor devices, and their
underlying physics. They also felt they had enlarged their
images helped develop an intuitive sense more than intuitive understanding of microelectronics. WebLab aug-
any other part of the course.’’ Another noted that mented what students gained from lectures and the text-
graphic images helped him to ‘‘think more intui- book, providing another perspective on the learning
tively,’’ while a third said that WebLab ‘‘solidifies an process. The students spoke about the ways in which Web-
intuitive understanding of whatÕs happening.’’ One Lab helped them increase their understanding of basic
microelectronic concepts. They commented that WebLab
student wasnÕt sure that WebLab helped him ‘‘learn ‘‘helped to conceptualize things,’’ encouraged them to
the material per se, but at the very least itÕs good for ‘‘build concepts in the mind,’’ and made it possible ‘‘to
gaining intuition. Often times I think thatÕs more confirm concepts learned in class.’’
Inquiry-Learning with WebLab 345

This perception was reinforced by another fac- strength of these judgments increased from one
tor. One student added, ‘‘WebLab builds confidence, semester to the next, with the improvements that were
and even enthusiasm for learning and understanding made in the WebLab program.
the subject.’’ This growing sense of confidence from
hands-on research was itself an important part of the OVERALL STUDENT ASSESSMENTS
learning process. OF WEBLAB EXPERIENCE
That sort of confidence permitted students to
approach their work in a more open and creative WebLab Surveys in the spring of 2004–05 and
way. Some took a playful approach, and were de- the fall of 2005 included sections in which students
lighted in learning from their own inquiries. One were asked to rate their experience of WebLab in
student said, ‘‘It was kind of cool just to play around general terms. Here again, the results showed a clear
with actual numbers, actual data.’’ Another added, pattern of improvement from one semester to the
‘‘What you can do is play around with inputs and next (Table VI).
figure out why they often differ so much.’’ He went Students found the revised programs to be more
on to note, ‘‘It gives you a good overall view. ItÕs ‘‘effective,’’ more ‘‘interesting,’’ and more ‘‘stimulat-
more intuitive than just the numbers.’’ ing.’’ Larger gains appeared in the strength of per-
Quantitative results of the WebLab Survey con- ceptions that WebLab was more ‘‘enjoyable,’’ more
firmed what students said in interviews. Students ‘‘meaningful,’’ and more ‘‘exciting.’’ The linkage of
strongly agreed that WebLab helped them under- these responses suggested that as WebLab became
stand the behavior of the devices, how to characterize more effective, students were in their own words,
a device) and also helped them understand the dif- more ‘‘excited’’ by work that was more ‘‘meaningful’’
ferences between theory and application. They also and more ‘‘enjoyable.’’
agreed that WebLab helped them develop a ‘‘feel’’ for At the same time, negative judgments grew
how microelectronic devices work, and deepened weaker. As the WebLab system worked better, stu-
their understanding of related lectures and readings dents were less likely to find the experience ‘‘tedious,’’
in the course (Table V). ‘‘frustrating,’’ ‘‘stressful’’ and ‘‘discouraging.’’ As this
These responses grew stronger from one semester happened they were less likely to blame themselves
to the next. In both semesters, students believed that for learning problems. In the first semester, when the
the use of WebLab, in combination with the more WebLab was not working as well as it did later,
traditional methods of lectures, texts and tests, had students remarked in their interviews that they were
enlarged and enriched their opportunities for learn- overwhelmed by the current workload, and failed to
ing. They also felt that they had come to understand give themselves enough time to do a good job, and
the behavior of semiconductor devices, and had were too tired to learn much from their work with
mastered the equations that can be used to compare WebLab. Structural problems in WebLab created a
with theoretical models. They commented that We- sense of failure in students, and they blamed them-
bLab ‘‘helped to conceptualize things,’’ encouraged selves.
them to ‘‘build concepts in the mind,’’ and made it After the problems were corrected, students
possible ‘‘to confirm concepts learned in class.’’ The believed that they had done better job in planning

Table V. Self-Reported Student Perceptions of the Effects of WebLab on Understanding

Spring 2005 Fall 2005

How well the WebLab experience affected understanding in particular areas Mean SD n Mean SD n %Change

Behavior of the devices 4.30 1.73 37 5.42 1.17 33 +26.05

How to characterize a device 5.45 1.15 33
Differences between theory & application 4.22 1.49 37 5.55 1.03 33 +31.10
Developing intuitive sense on how devices worka 3.81 1.66 37 5.00 1.20 33 +31.23
Related lectures and assigned readings 3.62 1.52 37 4.82 0.88 33 +33.15

Student survey responses, ranked on a 7-point Likert scale, ranging from ‘‘1’’ (= poorly) to ‘‘7’’ (= extremely well). Intervening values have
no descriptors.
a
In the spring of 2005, this question was worded differently: ‘‘Developing an intuitive feel of how to model.’’
346 Fischer et al.

Table VI. Self-Reported Student Opinions about the Overall WebLab Experience

Spring 2005 Fall 2005

Mean SD n Mean SD n %Change

Degree of Agreement with descriptors

of the overall WebLab experience
Meaningful 3.82 1.59 34 5.18 1.10 33 +35.61
Effective 4.12 1.57 34 4.79 1.22 33 +16.26
Interesting 3.56 1.38 34 4.67 1.36 33 +31.18
Stimulating 3.43 1.38 34 4.09 1.20 32 +19.24
Enjoyable 3.03 1.29 34 4.00 1.17 33 +32.02
Exciting 3.00 1.41 34 3.63 1.29 32 +55.67
Tedious 5.69 1.39 34 4.00 1.60 33 âˆÕ29.70
Frustrating 5.51 1.27 34 3.56 1.63 32 âˆÕ35.40
Stressful 5.43 1.34 34 3.38 1.64 32 âˆÕ37.76
Discouraging 4.85 1.58 34 2.75 1.37 32 âˆÕ43.30
Degree of agreement with statements about the overall
WebLab experience
Overall, WebLab was a positive learning experiencea 3.54 1.50 35 5.48 1.15 33 +54.80
Including WebLab examples in lecture was effective 2.97 1.34 35 5.64 1.34 33 +52.66
I did not give myself enough time to do a good job on the project 4.43 1.87 35 3.56 1.82 33 âˆÕ19.60
I was too tired to learn much from the WebLab project. 4.51 1.74 35 3.94 1.59 33 âˆÕ12.64
A key objective for the WebLab project 5.83 1.32 35 4.74 1.85 33 âˆÕ18.70
was simply to get it done because I was so
overwhelmed by the current workload.

a
In the spring of 2005, this question was worded differently: ‘‘I found the WebLab project an effective learning experience.’’

their work, and had given themselves enough time WebLab and judged it to be a valuable and useful
to get it done, and succeeded in doing so without experience for students. Most observed that WebLab
being overwhelmed or exhausted. This result is very could be used by students when and where they
instructive, and also very hopeful in its evidence that wanted. One instructor thought that WebLab could
correctable flaws in the learning process can do seri- be introduced in places where hands-on labs would be
ous injury not only to what is learned, but also to the difficult or impossible. Another mentioned how
self-esteem of the learner, and to the possibility of important it was for students to be able to work
learning in the future. remotely with confidence.
Faculty-members also thought another strength
INFORMAL JUDGMENTS BY FACULTY of the program was to be found in its efficient use of
AND TEACHING ASSISTANTS time and effort by students and instructors alike. One
instructor described WebLab as ‘‘80 percent of the
One-on-one discussions were also held with three experience for 20 percent of the effort.’’ Another
faculty members and four teaching assistants who observed that WebLab ‘‘takes the hassle out of a
had used WebLab in their teaching. These conver- lab.’’ Several commented on time wasted in labs be-
sations were very informal and wide ranging, on cause of faulty or broken equipment. With WebLab,
questions about faculty experiences, approaches to they found that students had more time to think
Web Lab, and perceptions of its strengths and about learning; faculty had more time to think about
weaknesses. The conversations were held in their teaching.
offices and were not recorded. The numbers were too WebLab was seen as a highly effective way for
small to be quantified, but the discussions were very students to have direct experience with real devices; a
interesting and suggestive. chance to play around and explore them; to make
In general, assessments of WebLab by faculty actual measurements in experimental settings, and to
and teaching assistants were very similar to those of characterize devices. According to one instructor,
undergraduates. All teachers were positive about ‘‘you canÕt actually characterize devices in any other
Inquiry-Learning with WebLab 347

way.’’ Another observed that students all began from Another instructor noted that fear itself may have
a level playing field, measured the same thing, and been the biggest barrier to learning.
went about the procedures in the same way. Instructors also commented that some students
From the faculty perspective WebLab had an seemed entirely centered on getting the right answer,
important impact on student-learning. ‘‘Students and did not think about what they were learning. A
learn best by doing,’’ was one instructorÕs comment. few students were very annoyed when they could not
Another noted that students mastered a new com- figure out quickly what to do. Instead of thinking
puter program, learned to characterize devices, about a solution by themselves, they immediately
compare them to theoretical models. According to asked instructors for answers to each specific prob-
one instructor, most students succeeded in learning lem. To some instructors, this suggested that some
what they needed to learn with WebLab. Some stu- students were working for grades rather than for
dents got it very quickly, a few never did. Others were understanding. Overall, Faculty observed that We-
sure that students had gained a greater understanding bLab is useful and important in what it offers
of basic concepts, although the evidence was some- teachers as well as students. This inquiry indicates
what indirect. that WebLab worked well for both groups.
WebLab graphic and visual images were men- Overall, students and faculty agreed that the
tioned as a particularly important way of working more WebLab was used, the better it functioned, and
with ideas. One instructor observed that graphs were the more effective it became as a way of helping
more important than numbers. Another thought students to learn. It enabled them to develop their
WebLab probably helped studentsÕ intuitive under- own learning processes within a rigorous empirical
standing, but cautioned that the evidence was limited. discipline. And its success in these ways helped stu-
A third commented that ‘‘it takes experience to get an dents to become more confident and more positive
intuitive feel,’’ but he thought that in time students about their own learning. In all of these strengths,
would develop a more robust intuitive understanding WebLab can be applied throughout the natural
of microelectronics. sciences and other experimental disciplines. Few
Instructors also commented that some students innovations are as promising as WebLab for inquiry-
had a limited understanding of WebLab. Several said based learning.
they were not sure that all the students believed they
were testing real devices. Students occasionally asked,
‘‘Are we actually testing real devices? Is this real REFERENCES
data?’’ Several were not sure that all students
understood what WebLab offered to them in terms of Albert, B. (2000). Some thoughts of a scientist on inquiry. In
J. Minstrell, & E. H. Zeevan ([Link] into inquiry
developing an understanding of the ‘‘big ideas’’ in learning and teaching in science (pp. 3–13). Washington:
microelectronics. One said that students might American Association for the Advancement of Science.
appreciate WebLab more if they knew how frustrat- Anderson, R. D. (2002). Reforming science teaching: What
research says about inquiry. Journal of Science Teacher
ing it would be to do the same thing without We- Education 13(1): 1–12.
bLab. Another added that students often donÕt realize Bransford, J. D., Brown, A. L., and Cocking, R. R. (Eds.) (2002).
until later what they have gained from WebLab. How people learn: Brain, mind, experience and school. Com-
mission on Behavioral and Social Science and Education,
Instructors also commented that studentsÕ opinions National Research Council. Washington: National Academy
on WebLab varied; some liked it, others didnÕt. One Press.
thought that WebLab fitted the learning styles of Bruner, J. S. (1987). Actual minds, possible worlds, Harvard
University Press, Cambridge, MA.
some students better than others. del Alamo, J. A., Hardison, J., Mishuris, G., Brooks, L., McLean,
They reported that some students got stuck on C., Chang, V., et al. (2002). Educational experiments with an
minor problems, or made stupid mistakes, or didnÕt online microelectronics characterization laboratory. In Inter-
national conference on engineering education, 2002, Manchester
read the instructions carefully. One instructor ob- (UK), August, 2002.
served that these problems could be corrected quickly del Alamo, J. A., Chang, V., Hardison, J., Zych, J. D., and Hui, L.
by briefly working alongside a student on a com- (2003). An online microelectronics device characterization
laboratory with a circuit-like user interface. In International
puter. Several mentioned that some students were conference on engineering education 2002, Valencia (Spain),
fearful that their mistakes would blow out a device, pp. 153–162.
or cause the program to shut down. One instructor Dewey, J. (1938). Experience and education, McMillan, New York.
Dunbar, K. (2001). The analogical paradox. In D. Gentner,
observed that students also worried about peripheral K. J. Holyoak, & B. N. Kokinov ([Link] analogical mind
problems, which to them werenÕt peripheral at all. Cambridge, MA: MIT Press.
348 Fischer et al.

Feisel, L. D., and Rosa, A. J. (2005). The role of the laboratory in MIT iCampus iLab website, retrieved July 8, 2006 from http://
undergraduate engineering education. Journal of Engineering [Link]/icampus/.
Education 94(1): 121–130. MIT OpenCourseWare, Electrical engineering and computer sci-
Felder, R. M., and Brent, R. (2003). Learning by doing. Chemical ence, microelectronic devices and circuits, Fall 2005. Retrieved
Engineering Education 37: 282–283. July 8, 2006 from [Link]
Fischer, J. H. (2005a). MIT iCampus WebLab: A report on Engineering-and-Computer-Science/6012Fall-2005/Course-
interviews and responses by students enrolled in 6.012 (Micro- Home/[Link].
electronic Devices and Circuits), Fall and Spring Semester, Mitchell, R. (2005). A survey study of the MIT iCampus WebLab
2004–2005. Document from the Electrical Engineering and used remotely by electrical engineering students in a lab of device
Computer Science Department, MIT. characterization and circuit design, Document from the Elec-
Fischer, J. H. (2005b). A report on interviews and responses by trical Engineering and Computer Science Department, MIT.
students enrolled in 6.012 (Microelectronic Devices and Cir- Mitchell, R. (2005). A survey study of the 6.012 WebLab experience,
cuits), Fall Semester, 2005. Document from the Electrical Document from the Electrical Engineering and Computer
Engineering and Computer Science Department, MIT. Science Department, MIT.
Hardison J., Zych D., Del Alamo J. A., Harward V. J., Lerman S. National Research Council. (2000). Inquiry and the national science
R., Wang S. M., et al. (2005). The microelectronics WebLab education standards: A guide for teaching and learning,
6.0—An implementation using web services and the ILab National Academy Press, Washington.
shared architecture. In International conference on engineering Norman, G. R., and Schmidt, H. G. (1992). The psychological
education and research 2005, Tainan, Taiwan. basis of problem-based learning: A review of the evidence.
Henry, J. (1996). Running laboratory experiments via the World Academic Medicine 67: 557–565.
Wide Web. ASEE annual conference, session 3513, 1996. Piaget, J. (1928). ChildÕs conception of the world, Routledge &
Available from National Instruments Academic Resources, Keegan Paul, London.
CD-ROM 2001 Edition. Prince, M. J. (2004). Does active learning work? A review of the
Jeppson, K., Lundren, P., del Alamo, J. A., Hardison, J. L., and research. Journal of Engineering Education 93: 223–232.
Zych, D. (2004). Sharing online laboratories and their com- Schmidt, H. G., Norman, G. R., and Boshuizen, H. P. (1999).
ponents: A new learning experience. In 5th European workshops A cognitive perspective on medical expertise: Theory and
on microelectronics education, Lausanne, Switzerland, April implications. Academic Medicine 65: 611.
15–15, 2004. SPSS. (1999). SPSS Base 10.0 userÕs guide, SPSS, Chicago.
Martin-Hansen, L. (2002). Defining inquiry: Exploring the many Streiner, D., and Norman, G. (1995). Health measurement scales
types of inquiry in the science classroom. The Science Teacher (2nd ed.). Oxford University Press, Oxford.
69(2): 34–37.
Minstrell, J., and Van Zee E. H. (Eds.) (2000). Inquiring into inquiry
learning and teaching in science. Washington: American
Association for the Advancement of Science.