AC 2012-3291: FROM DEMONSTRATION TO OPEN-ENDED LABS: RE-

VITALIZING A MEASUREMENTS AND ANALYSIS COURSE

Dr. Bridget M. Smyser, Northeastern University

Bridget M. Smyser is an Assistant Academic Specialist and Director of Laboratories, Department of

Mechanical and Industrial Engineering .

Kevin McCue, Northeastern University

American
c Society for Engineering Education, 2012
 From Demonstration to Open Ended: Revitalizing a
Measurements and Analysis Course

Abstract

The course entitled Measurements and Analysis with Thermal Science Application is a required
course for Mechanical Engineering students at Northeastern University. The existing
experiments have become dated and in many cases have devolved to demonstrations by the
teaching assistants, with little hands on experimentation by the students. This has resulted in
severely decreased student satisfaction with the labs. An extensive redesign was performed to
develop hands-on, open ended lab experiments that allowed students increased control over the
outcome of the experiments. Pre-lab homework assignments require students to develop lab
procedures, research sensor specifications, and develop virtual instruments in National
Instruments’ LabView. A term project required student groups to design and execute a
measurement experiment, presenting their findings in oral and written reports. Results show that
class attendance was higher than previous offerings while student grades were similar. The new
lab experiments were rated as more interesting and engaging by the students than comparable lab
experiments in the previous course. The new term projects demonstrated in depth knowledge of a
wider range of course topics and showed a great deal of sophistication and ingenuity on the part
of the students.

Introduction

Courses in measurement and experimental data analysis are staples in many mechanical
engineering departments. i,ii,iii,iv In the mechanical engineering department at Northeastern
University, Measurements and Analysis has been a required course for juniors for many years.
At some point in the past, lab equipment was purchased or constructed for the purpose of
teaching students how to measure fundamental engineering variables such as strain, temperature,
pressure, flow rate, drag forces and rotational frequency, as well as to give students practice in
statistical data analysis and exposure to measurement system response to varying inputs. It is to
the credit of the initial instructor that these experiments were robust enough to survive for
approximately twenty years without any significant upgrades. When the technology became
available the experiments were supplemented to include the use of National Instruments
LabView for data acquisition, but otherwise the lab handouts and the lab experiments remained
largely unchanged. The labs were generally easy to administer and grade, and the constancy from
term to term was very economical in terms of instructor time.

In the last several years, however, these labs have become increasingly dated, and student
complaints have been on the rise. Students at this university have two to three six month co-op
experiences during their five year program, and are thus exposed to real world measurement
situations early and often. They began to see these labs as boring at best and completely
irrelevant at worst. There were frequent complaints that “No one measures pressure that way
anymore.” and “This equipment is older than me!” and senior exit surveys showed that many
students felt it was time to update the experiments.

An added difficulty is that the number of students in the mechanical engineering department has
been growing steadily over the last several years. The lab equipment that worked fine for two lab
sections of twelve students each per week now has to stand up to six lab sections of up to sixteen
students each. In some cases this has been dealt with by constantly babying old equipment or
upgrading small devices such as multimeters or oscilloscopes. In other cases, the solution was to
turn to demonstrations as an alternative to letting the students handle the equipment, in order to
preserve it for all the sections. The current lab space is adequate, but small, and no lab space is
available for large amounts of new equipment. Departmental budget constraints have also limited
the ability to replace equipment on a large scale. Other institutions have also noted that lab
courses can stagnate due to the high cost in both instructor time and money associated with
changing apparatus. v

The old equipment and labs are only part of the overall problem. As other institutions have seen,
repeating the same standard labs year after year can lead to widespread plagiarism and the
passing down of labs from one class to the next.v,vi In addition, the literature is full of examples
of the increased learning benefits that come from open ended, design type problems. vii Fixed,
static experiments, while easy to administer and grade, limit student creativity and interest.
ABET guidelines require students to learn skills in design of experiments, which is difficult for
students to learn when they are limited to either demonstrations or ‘cookbook’ type labs. viii
Students benefit from being able to make mistakes and learn from them, and from being able to
influence the outcome of their experiments. All of these factors led to the need for a concerted
effort to change the way this class was taught.

The goals of this course redesign were as follows:

 Eliminate demonstration labs entirely in favor of hands-on lab experiences

 Eliminate textbook homework problems in favor of focused pre-lab assignments
 Supplement traditional lectures with in-class demos, problems worked individually and in
groups, and other active learning techniques
 Progressively give students more control over the design and execution of the lab
experiments over the course of the term
 Institute an experimental design project that required students to design, execute, and
analyze the results from a measurement experiment of their choice

Specific problems with the previous course

As far back as 2006, senior exit surveys indicated a desire for more open ended problems and
hands-on labs.ix In more recent years, there have been more and more student comments
expressing an intense dislike for demonstration type labs. Demonstration type labs seem to have
evolved primarily to deal with increasing enrollment, which has gone up dramatically in recent
years. Unfortunately, it has been difficult for the infrastructure to keep up with the increase.
Since a broken piece of equipment in the first of six lab sections makes it impossible to teach the
remaining sections, the emphasis had shifted in many cases to protecting the equipment above all
else. In the Measurements and Analysis class, for example, the pressure experiment was by and
large performed by the TA, with the students watching and writing down data. In another
example, the strain lab required several inconvenient work-arounds to allow students to adjust
the zero setting on a Wheatstone bridge. In real life this would be done by making a manual
adjustment to the data acquisition module. In the lab, however, it was found that the modules
were easily broken, and so students were not allowed to adjust them, but rather had to manipulate
the data afterwards to account for the inability to zero the circuit. This made the concepts more
difficult to understand, and students tended to lose sight of the point of the experiment and focus
on the clunky, complex work-arounds. Because of the need to shield the equipment, students
were often bored and frustrated in the lab sections.

As has been reported in other institutions, lab reports can be passed down from one semester to
the next. Although software does exist for the purpose of comparing documents for similarity,
this requires documents to be submitted in electronic form. For an 80 student class, this ends up
being an administrative burden. Moreover, since the labs have been essentially unchanged for
more than a decade, it may not be possible to root out all past lab reports. The previous course
instructor attempted to address this possibility by having two midterm exams and one final exam.
This required students to demonstrate that they understood the key concepts of the course,
effectively punishing those who had copied lab reports without understanding. One of the
problems with this approach is that some of the concepts that are very important outcomes, such
as the ability to design experiments and do extensive statistical data analysis, are difficult to
assess in an exam setting. The number of topics covered in the course was large enough that it
was not possible to cover all topics in any sort of depth on an exam. Exams were therefore a
rather blunt and unsatisfactory assessment tool for this type of course.

At this institution, lab sections are generally taught by graduate teaching assistants. TA duties
range from supervising students to demonstrating equipment to grading lab reports.
Demonstration type labs both helped and hindered TAs in their ability to educate the students in
their sections. Demonstration labs were relatively easy to teach to the TAs. Because only the TA
was handling the equipment, the number of possible errors or equipment malfunctions was
generally low, which saved the TA from having to help students troubleshoot four to six lab
stations. However, these labs required the TA to give mini-lectures and explanations of what was
happening. For some international graduate students this was a challenge due to ESL issues. The
grading of lab reports could also be problematic for some non-native speaking TAs, leading to
grading differences between sections. The lab handouts in this class were designed to make
grading easier by providing a numbered list of questions that students answered in their lab
reports. This tended to make the labs much more like extended homework problems, with a
limited range of possible correct answers. If data did not come out correctly in lab, students were
given correct data to do the analysis on so that the answers would come out as expected. This
encouraged students to focus more on the ‘right answer’ than on the process of doing the
measurement and critically evaluating the results.

Another repeated complaint from the student body was that the course appeared too late in the
curriculum to allow them to benefit from it. Since it is a junior level course, the students had
already been exposed to real life measurement applications during their co-op experience. By
comparison, the rigidly structured experiments and older equipment seemed hopelessly out of
date. Although the class does provide needed information on data analysis and measurement
techniques that can be used in the senior capstone design class, the students often had trouble
translating what they learned in this course to other situations. Students might learn, for example,
how thermocouples work and be able to demonstrate the laws of thermocouples. However, they
would repeatedly get to the capstone design class and be unsure how to choose the right
thermocouple for the job, or would not understand when a more accurate measurement device
was needed. The combination of all these factors led to an extensive effort to redesign the course,
both in terms of lab equipment used and assessment schemes.

Revised course

The new course covered essentially the same material as the previous course, but delivered in a
hands-on, open ended style. Lectures were supplemented with in class lab demos, group
discussion, unannounced quizzes, and other activities designed to reinforce class concepts.
Formal, scheduled exams were eliminated entirely as an assessment method. In-class graded
assignments, worked both individually and in groups, now make up 20% of the grade. The
homework, worth 20% of the grade, consists of problems and activities directly related to the
upcoming lab experiment. Students were assigned to lab groups by the instructor, and were
required to work in these groups for both the lab and term project. The number of labs was
reduced to seven, with the lab reports making up 35% of the final grade. The term project, now
25% of the grade, was completely altered. Instead of researching an existing measurement
application, student groups were expected to propose, design, and execute an experiment using
available lab equipment and present their results in an oral and written report.

In-Class Activities

Demonstration labs have a certain amount of value when it comes to illustrating techniques and
giving a preview of how lab experiments will work. The older lab equipment for this course was
able to be repurposed in several cases for use as in-class demos during lecture. For example, the
cantilever beam arrangement originally used for strain analysis could be hooked to a portable
strain gauge indicator box and used to demonstrate, in real time, the effect of altering the
arrangement of the gauges in the bridge circuit. Students were then asked to do problems using
the strain gauge data gathered in lecture in order to predict what the strain should be in various
configurations. Students were also able to see firsthand the minor mistakes made by the
instructor in handling the instruments and thus were better able to troubleshoot problems in lab.
In another demo, various temperature measurement devices were used together with an ice bath
and a boiling water bath to illustrate the principles of calibration and thermocouple laws and to
compare the response of various sensors.

Some of the lab experiences which had seemed tedious to the students as a separate experiment
were used in an altered form as an in-class activity. For data analysis, the previous experiment
which involving measuring sets of forty resistors in the lab was eliminated. Instead, smaller
numbers of resistors were measured in small groups in class and all the data combined. This data
could then be manipulated in real time by the instructor and the students to illustrate statistical
data analysis concepts. Because the measurements were spread out over a larger group, the
students found it much less tedious and still were able to benefit from the generation and
manipulation of a large data set.

It was recognized that individual student understanding of concepts needs to be proven for
accountability purposes. To this end, several key concepts were tested via unannounced quizzes.
Some of the quizzes were short answer or multiple choice, and others required working more
complicated mathematical problems. These quizzes served as both an incentive to come to
lecture, as well as a check to ensure students were not ‘coasting’ in their lab groups.

Finally, a number of group activities were performed in lecture with an eye toward improving
understanding and preparing for lab activities. Lab groups might be given class time to discuss
their data collection scheme for an upcoming lab and come to a common consensus on what data
should be gathered. Complex multistep data manipulation could be practiced and the difficulties
worked out in groups before coming to lab, which served to both prepare for lab and reinforce
lecture concepts. Since many measurement decisions and sensor choices can be judgment calls,
the student groups would have to come up with a solution and be able to justify and explain their
choices. These essential lab skills are very difficult to distill into a multiple choice question, but
lend themselves well to group discussions.

Lab experiments and Pre-lab exercises

The first step in the redesign of the labs was to have LabView instruction occur during the very
first lab session, as opposed to waiting until part way through the term. Students were expected
to write LabView virtual instruments for several of the labs, and for their term project if needed,
and it was felt that introducing the software as early as possible would facilitate this. Some of the
earlier labs were eliminated or their elements distributed across other labs to cover similar
material in a different fashion. The new labs included: Pressure, Temperature, Strain, Mechanical
Power, Heat Transfer, Wind Tunnel Testing, and Flow Measurement/Hydropower. The goal was
to have the labs become progressively more open ended and unstructured, giving students
increasing control over the outcomes as they learned techniques and concepts.

A key tool in preparing students for lab was the pre-lab homework assignment associated with
each experiment. The typical pre-lab would contain:

 From one to three numerical calculations similar to what would be required in lab (i.e.
calculating the theoretical flow rate, predicting heat transfer)
 Several comprehension questions covering key theoretical concepts (i.e. explaining the
laws of thermocouple behavior, identifying variables)
 The development of a data table for the lab. Students were required to read the lab
handout, decide what data should be collected and turn in a blank, labeled table indicating
what data they planned to collect during lab.
 In some cases, students were also asked to select the sensors they would use and justify
their selection, or write a procedure for the lab.

These pre-lab assignments were graded and returned prior to the related lab in order to give
students feedback to keep them on track.

Some of the labs required only slight modification to make the experience more hands-on and
less ‘canned’. The previous temperature experiment required the students to measure four
temperature environments with several instruments and construct a four-point calibration curve,
followed by using nonstandard thermocouples and thermocouples with nonstandard reference
junction temperatures to make the same measurements. In the new experiment, students were
presented with four known temperature environments and one unknown temperature
environment in the form of a fixed temperature bath which the lab technician had set before
covering up the readout. The students were required to choose 3 temperature measurement
devices out of a possible 5, create calibration curves for them, and then determine the
temperature of their unknown bath. As an added incentive, a small bonus was given to the group
that got closest to their actual unknown temperature. The lab groups had to choose the devices
and justify their choices based on expected accuracy, precision, and response time, as determined
by examining the manufacturer’s specifications. In addition, the students had to choose at least
one device that could be read by LabView and write a VI to capture that data. In another case,
the flow measurement/hydrodynamic power lab, the existing Pelton wheel demonstration
equipment was used, but the experiment changed. Each lab group was given an hour with the
equipment, instruction on its use, and a list of goals for the experiment. The lab groups were
required to write the procedure, decide what data to collect, and determine what calculations
were necessary to achieve the desired goal. This involved no extra expenditure in equipment, but
completely changed the lab from one in which students showed up and wrote down numbers
blindly to one in which they were in control of the experiment from start to finish. The wind
tunnel experimentation lab was already well liked by the students, but additional sensors were
added. These additional sensors were placed by the students, who needed to decide on the
optimal location for the sensors and explain and justify their choices.

In some cases, equipment that was less automated was substituted in order to provide a more
student controlled experience. In the new strain lab, students were challenged to turn a piece of
aluminum, instrumented with strain gauges, into a scale by relating the strain readings to the
weight added to the aluminum plate. To avoid the problem of students breaking the DAQ
modules, the strain gauges were read by an indicator box instead. This equipment, while older
and requiring manual data recording, was extremely robust and could stand up to repeated
manipulation by the students. It also gave obvious erroneous readings when the circuit was wired
incorrectly, forcing students to troubleshoot and leading to a much deeper understanding of the
relationship between the physical location of the gauges and the location in the bridge circuit.
Students were asked to weigh an unknown object (a backpack, a group member, a tool box, etc.)
with their ‘scale’ which was then verified on a bathroom scale. This forced them to consider
which arrangement of strain gauges gave the best response and was also highly interesting to the
students. For the pressure lab, the large stand-alone pressure demonstration board was replaced
with four smaller setups involving regulators to control house air. These regulators were verified
by two separate sensors – one digital, one analog. The students were required to calibrate the
sensors and also to write a LabView program to log the digital data. As this was the first lab,
there was more guidance given, but the students were all much more physically involved in
connecting and disconnecting sensors, writing the LabView VI, and troubleshooting the physical
system, rather than just recording data.

The previous rotational frequency lab had measured rotational frequency of a spinning shaft, but
was extremely uninteresting to the students. A completely new lab was substituted which related
rotational frequency to power produced using a bicycle powered generator. Each lab set up
contained a bicycle powered generator connected to an oscilloscope. Students were required to
generate power curves by measuring voltage, current, and rotational frequency using the
oscilloscope, a multimeter, and a phototachometer. The students quickly learned that pedaling at
a constant rate took some practice, and that the accuracies of the various sensors and measuring
devices differed. Students were also required to read the extensive manufacturer’s specifications
for all the devices and determine expected uncertainty as well as confidence intervals for their
data. Because there was a genuine need to gather the rotational frequency data, in order to
determine the efficiency of the generator and produce power curves, the students saw the data
gathering as relevant, rather than a pointless exercise.

This course has long had an emphasis on verifying theoretical calculations through actual
measurement. To this end, the heat transfer lab allowed students to verify a problem from their
heat transfer textbook experimentally. In the pre-lab assignment, the students were asked to
predict the time it would take for a hot dog to heat to a certain temperature in boiling water. In
the lab, they were then asked to measure this for comparison, and also to determine the heat
transfer coefficient of their thermocouples. Most of the students were enrolled in heat transfer at
the same time as this course, and had already had one exam on these concepts. The timely
verification of the theoretical calculations learned in the other course really brought the concepts
home to the students and tied the entire curriculum together in a way that was much appreciated
by the students.

Term project

The new term project was inspired by a similar project in the MIT Measurement and
Instrumentation course.iv The MIT course had a project entitled ‘Go Forth and Measure’ in
which students were given the opportunity to measure something of interest to them. The term
project in the current course had a similar theme. Students working in their lab groups were
required to propose a measurement experiment, design the procedure, perform the experiment,
and present the results in a written and an oral report.

The project consisted of three milestones. The first milestone was a proposal in which students
had to provide a clear statement of the problem they were interested in, evidence of background
research on similar problems with references, and a preliminary procedure. These were vetted by
the course instructor and the lab technician in order to ensure that the equipment was available
and that the procedure could be performed safely. Students were also given guidance if their
problem was too narrow or broad.

The second milestone was a project update. The students were required to produce a much more
detailed, refined procedure together with a safe operating procedure detailing what safety
concerns applied to their project. They were also required to list the equipment they would need,
discuss any preliminary data, and explain their plan for statistical data analysis. This information
was used to schedule student groups in open lab times or to arrange for groups to borrow
equipment for measurement outside of lab. Student topics were varied and showed a great deal of
creativity, as shown in Figure 1 below.
Sample of Student Topics for Term Projects
Measuring decibel levels around the T (subway)
Determining which vendor’s coffee cups kept coffee hot the longest
Measuring the pressure produced by adding Mentos to Diet Coke
Determining the critical radius for heat flux through an insulated copper pipe
Quantifying the effect of drafting on bicycle velocity
Determining which fan produces the largest wind volume per kW and per dollar
Producing a human comfort map of the library based on temperature and air flow
Determining the maximum load for a longboard (skateboard)
Figure 1: Selection of student term project topics from Fall 2011

The third deliverable was a fourteen minute group oral report and a formal written report. In
addition to stating the problem, procedure, and results, students were asked to discuss how they
would improve the experiment if they were to do it again. Students were also required to perform
uncertainty analysis and to give confidence intervals for their data, as well as perform any
necessary statistical analysis.

This project required a certain amount of scheduling and infrastructure on the part of the course
personnel. By eliminating one of the previous lab experiments and moving that activity into the
lecture hall, lab time was made available for student groups to come in and use available lab
equipment for their experiments. Some devices such as anemometers, decibel meters,
multimeters, and thermometers could also be checked out by groups who needed to make
outdoor measurements or otherwise did not need to work in the lab. Some groups needed to be
scheduled so that limited equipment could be shared, but this ended up being a small issue.
Students were given a list of available equipment at the beginning of term, but some devices
were purchased on a case by case basis to support particular projects, with the idea that these
devices would then be available for future terms. All of this was accomplished without a large
financial outlay by the department.

Assessment

A common concern in switching to open-ended assignments is the issue of fair grading. This was
a large concern in the current course as all three of the TAs were non-native speakers, and one
was assigned to the course for the first time. This problem was alleviated by the development of
comprehensive rubrics, an example of which is shown in Appendix A. These rubrics gave very
clear indications of what information was being sought and exactly how points should be
distributed. TAs were asked to attach copies of the rubric to each graded lab report with the
appropriate cells circled in order to inform the students of exactly where the various points came
from. Similar rubrics were used by the instructor to grade the homework, the oral reports, and the
written report. In-class assignments were generally graded in a more traditional fashion, with 1-5
points for a correct answer, depending on the difficulty of the question.
The lab report rubrics proved to be a success. Reports graded by each of the three graduate
students tended to have similar averages. In cases where grades were disputed by the students,
the rubric served as the final arbiter of what was required, and allowed for easy remedy of the
situation. The lab rubrics were posted on the course website prior to the lab due date, which
allowed students to know exactly how they would be graded. On one occasion when the rubric
was not posted as early as previous rubrics, a large number of students sent emails asking for it,
indicating its value as a tool to promote better report writing.

Results

Table 1 below shows a comparison of the grades earned by three offerings of the course. All
three offerings were taught by the same instructor and lab technician, and two of the three TAs
were the same for all three terms. The Fall 2010 semester was taught in essentially the same form
as it had been for the previous ten years, due to a new instructor coming in with only eight days
to prepare. During the Spring 2011 term the course was also essentially the same, with some
updates and improvements to the lab handouts. The complete redesign of the course occurred in
Fall 2011. ANOVA analysis shows that there was no statistical difference between the three
terms with regards to average grades. It is important to note that many of the low grades in the
Fall 2011 course were due to unexcused absences resulting in missing in-class participation
points, as well as lab reports that were penalized for being turned in late, rather than for incorrect
analysis. This data seems to indicate that the student performance was not adversely impacted by
the change in the course. Indeed, it is gratifying to notice that even without the benefit of access
to past lab reports, the students achieved high grades on both individual and group assignments.
Table 1: Comparison of grades for three offerings of Measurements and Analysis

Term Average Lab Average Average Average

Average
Grade Project Homework Final Exam
Overall
Grade Grade Grade
Course
Grade
Fall 2010 92 88 89 90 91
Spring 2011 94 89 77 86 90
Fall 2011 93 90 87 93* 88
*This grade was determined by the average grades on all the individual in-class assignments that
covered the same topics as typically covered by a final exam.

The term projects were always intended to be a culmination of the course efforts, and a chance
for students to demonstrate knowledge of course concepts as applied to a real world problem. In
order to evaluate how well the projects achieved this goal, the final presentation slides for all the
groups in Spring 2011 and Fall 2011 were examined to determine the number of course topics
discussed in the talks, as well as the number of times each topic was mentioned. These results are
shown in Table 2 below. During Spring 2011 nine distinct course topics were mentioned. Due to
the framing of the project assignment, which required the groups to research an existing
measurement problem, most groups focused on sensor options. However, despite being required
to discuss calibration needs and data analysis needs, very few groups actually discussed these
topics. Some groups did specific calculations relating to their individual project (such as heat
transfer or HVAC related calculations), but more than half of the groups did not. The topics that
were covered were discussed competently and demonstrated understanding, but many difficult
topics were neglected entirely. In contrast, during the Fall 2011 term at least 15 distinct course
topics were discussed by the various student groups. More than half of the groups discussed
possible measurement errors in detail, as well as providing specific calculations. Moreover the
calculations were much more involved and sophisticated, and were used to determine real
answers, rather than as an intellectual exercise. Statistical data analysis and uncertainty analysis,
both difficult topics, were not discussed by any groups in Spring 2011, but were discussed by at
least a third of the groups in Fall 2011. Overall a much broader range of topics were covered in
greater depth during the open ended student designed experiments than in the instructor assigned
research projects previously used in the course.
Table 2: Course topics discussed in final projects, Spring 2011 vs. Fall 2011

Spring 2011 Fall 2011

Course Topic # Instances % of Course Topic # Instances % of
Projects Projects
Sensor Sensor
options/comparing options/comparing
sensors 14 93 sensors 9 39
Specific Specific
calculations 6 40 calculations 14 61
Calibration 4 27 Calibration 7 30
Strain 1 7 Strain 1 4
Pressure 3 20 Pressure 2 9
Force 2 13 Temperature 12 52
Measurement
Temperature 1 7 Errors 14 61
Measurement Statistical Data
Errors 1 7 Analysis 9 39
Accelerometers 1 7 Uncertainty 7 30
Heat transfer 3 13
Variable
identification 2 9
1st order systems 1 4
Flow rate 3 13
Electric power 1 4
Accelerometers 1 4
Another measurable difference between the two terms was the degree of student satisfaction with
and engagement with the lab experiments. Students in Spring 2011 were surveyed to determine
their attitudes toward the first five lab experiments: rotational frequency, data analysis, pressure
measurement, strain, and temperature measurement. Students in Fall 2011 were administered the
same survey after the first four lab experiments: pressure, temperature, strain, and mechanical
power. These labs were chosen because they were the most comparable to the previous labs in
terms of concepts taught. Since the mechanical power lab involved measuring rotational
frequency and using oscilloscopes, as did the original rotational frequency lab, it was thought
that these two labs would be reasonably comparable. The following statements were rated on a 5
point Likert scale:

1. This lab helped me learn the material

2. This lab was interesting and engaging
3. This lab was frustrating and confusing
4. This lab was supported by the lecture
5. I can imagine applying some of this information to other problems

The results of this survey are shown in Table 3 below. All of the labs were seen as more
interesting and engaging in the newly renovated course. The temperature lab in particular more
than doubled its score in this regard. The new pressure lab was much more effective in helping
the students learn and was seen as less frustrating and confusing by a large margin. The new
temperature lab improved in all measures. Strain is a difficult concept for the students, and the
strain lab suffered from some failures in the indicator boxes, leading to some amount of
frustration. However, even with the equipment problems the students found the lab much more
interesting than the previous incarnation. The mechanical power/rotation lab was more difficult
to compare, as the new lab was much more involved than the old lab. Some of the frustration in
the mechanical power lab was due to miscommunication on the part of the TAs, resulting in
some groups not gathering required data. Nevertheless the students found it more engaging, if
slightly more frustrating, and fewer people disagreed with the statement that the lab helped them
to learn the material. Collectively, there was a substantial improvement in student satisfaction
with the lab experiments, despite the fact that the new labs required more work and more original
thought on the students’ part.
Table 3: Results of student lab survey

Spring 2011 Fall 2011

% %
Strongly %Disagree/ Strongly %Disagree/
Agree/ % Strongly Agree/ % Strongly
Agree Neutral Disagree Agree Neutral Disagree
Pressure Lab
Helped me learn 59 25 15 86 12 2
Interesting and engaging 15 42 42 49 40 11
Frustrating and confusing 17 36 47 9 12 79
Supported by lecture 97 3 0 91 7 2
Can apply to other problems 65 28 7 70 25 5
Temperature Lab
Helped me learn 68 25 7 87 7 6
Interesting and engaging 26 53 21 67 30 4
Frustrating and confusing 22 42 36 10 15 75
Supported by lecture 86 14 0 96 4 0
Can apply to other problems 73 25 2 78 18 4
Strain Lab
Helped me learn 57 33 10 67 28 5
Interesting and engaging 25 47 27 46 40 14
Frustrating and confusing 41 29 31 46 30 25
Supported by lecture 83 17 0 84 13 4
Can apply to other problems 66 32 2 53 40 7
Rotational Frequency/Mechanical Power Lab
Helped me learn 76 17 7 66 29 5
Interesting and engaging 54 36 10 88 9 4
Frustrating and confusing 7 25 68 18 23 59
Supported by lecture 93 7 0 74 23 4
Can apply to other problems 74 21 5 70 21 9

Discussion

The grades in this course have historically been rather high. This can be attributed to the fact that
at this point in their academic careers students have seen many of these topics, either in other
classes or on co-op. However, it is interesting to note that even with greater demands placed on
the students to generate their own lab procedures and data gathering schemes the grades earned
were still relatively high.

If one looks at the presentations produced for Spring 2011 compared to Fall 2011, the Spring
presentations collectively cover 9 distinct course topics in 15 presentations. The Fall
presentations collectively cover 15 course topics in 23 presentations. The Spring talks discussed
weighing of sensor options and some specific calculations, but barely mentioned calibration, and
didn’t mention errors and uncertainty at all. The Fall talks had many instances of in-depth
statistical data analysis, discussion of measurement error, uncertainty analysis, calibration, in
addition to heat transfer, specific calculations, comparing sensor options, and a wide variety of
techniques. In Fall 2011, every single group did actual measurement, and with few exceptions
the experiments were well planned, involved detailed data analysis, clear explanation of errors
and the need for calibration, detailed, specific calculations that involved outside research and
concepts learned in other classes, and clearly demonstrated knowledge of course concepts.

Class attendance is a large problem, especially at the early morning lectures. By making in-class
participation graded and mandatory, fewer students came to class less than twice a week. Those
that did rapidly found that their grade decreased by a letter grade or more. In the past, students
were able to show up for the labs and the exams, and turn in homework, and get a reasonable
grade with very little interaction. However, this made individual assessment somewhat
problematic in that a large proportion of the grade (55%) was based on work that could very
easily be subject to unauthorized collaboration. The exams helped somewhat with this, but the
course concepts that could be easily turned into exam questions tended to be limited, numerical
calculations that were unable to explore some of the more difficult concepts in any depth.
Because a final might have to cover heat transfer calculations, drag coefficient calculations, fluid
flow, uncertainty, and data analysis, there was insufficient time to cover any one topic in depth.
The in-class problems, however, could be more difficult, and a large amount of time could be
spent puzzling out one problem. The mixture of individual and group assignments ensured both
individual accountability and peer-learning which made difficult concepts more manageable. The
in-class activities also included demonstrations, hands on measurement activities, real-time data
calculation, and other tasks designed to make students think and not just plug in numbers.

The most startling change is in the student perception of the course. Students typically take this
course in their junior year, after having had at least one and often two co-op experiences.
Because of the timing, students often expressed the opinion that they had already learned what
they needed to know on co-op and had nothing to gain from this course. Moreover, since the
course had equipment that was in many cases severely outdated, the prevailing view was that
there was no relevance to current real-life measurement practices. Students in the redesigned
course were very enthusiastic about the term projects, with several groups coming to three open
lab sessions of two to three hours each in order to complete their experiments. The projects were
especially beneficial in getting students to understand the practical difficulties in measuring
engineering variables. Several groups planned experiments, only to have to revise their
experiments extensively once they realized that their planned approach either would not work, or
would take much too long to complete, or would result in data files that were too large to easily
process. This type of information is next to impossible to convey in a lecture, but is invaluable
for students who want to make measurements in real world situations.
Conclusions

The course redesign was a resounding success. Students learned the material in greater depth,
applied it to a broad range of situations, and gained practical experience in experimental design.
This was accomplished with a relatively modest amount of financial outlay. The key purchases
were new, low cost sensors, bicycle generators, and new oscilloscopes. The redesign did require
a large amount of time on the part of the instructor and the lab technician. However, the new
experiments are easily modified, and will require less time to administer and maintain in
subsequent course offerings. Additional funding is being sought to continue to upgrade old
equipment and provide more options for students to use during their term projects. Future classes
will continue to be surveyed to ensure that the gains made in the new offering continue as the
new labs mature and to prevent stagnation from occurring. It is hoped that the concept of giving
more control to the students will be valuable to other instructors, both at this institution and
elsewhere.

i
University of California Berkeley, Department of Mechanical Engineering, Syllabus for ME 107A,
http://www.me.berkeley.edu/ABET/2005/courses/107Aweb.shtml, last accessed 5/10/2011
ii
University of Texas-Pan American, Department of Mechanical Engineering, Syllabus for MECE 3320,
http://crown.panam.edu/measurements/syllabus.html, last accessed 5/10/2011
iii
University of Texas at San Antonio, Department of Mechanical Engineering, Syllabus for ME3113,
http://engineering.utsa.edu/~mechanical/curriculum/syllabi/ME3113%20ABET%20Syllabus.pdf, last accessed
12/20/2011
iv
Massachusetts Institute of Technology, Department of Mechanical Engineering, Syllabus for 2.671,
https://wikis.mit.edu/confluence/display/2DOT671/2.671+Home, last accessed 9/7/2010
v
Middelberg, A., “Laboratory Projects – Should Students Do Them or Design Them?”, Chemical Engineering
Education, vol 29, no. w, 1995, p. 34.
vi
Macias-Machin, A., Guotai Zhang, and Octave Levenspiel, “The Unstructured Student-Designed Research Type
of Laboratory Experiment”, Chemical Engineering Education, vol. 24, no. 2, 1990, pp. 78-79.
vii
Dym, C.L., A.M. Agogino, O. Eris, D.D. Frey, and L.J. Leifer, “Engineering design thinking, teaching and
learning”, Journal of Engineering Education, vol. 94, no. 1, 2005, 103-120
viii
ABET, 2009-2010 Criteria for Accrediting Engineering Programs, 2008
ix
Northeastern University, ABET Self Study Report for Mechanical Engineering, 2007
Appendix A: Sample grading rubric for Temperature Lab

Item Description Points Give 6 points if Give 3 points if Give 0 points if

Discussed the purpose of the experiment, any They covered all three
Introduction relevant theory, and the significance of the results. 6 of these points They missed one point They missed two points
Described the equipment/sensors used. Discussed
manufacturers' specs, Described form of data
Equipment collection. Included VI screenshot and description They provided all the They missed two or
Description of what measured. 6 requested information They missed one item more items
They described all the They repeated the lab
steps they performed, handout or left out any
Described step by step what they had to do to in a clear, easy to major portions of the
Procedure make the measurements 6 follow style procedure Section missing
Discussed errors,
failures, and Discussed the errors, but
Explained any errors, unexpected results, or troubleshooting clearly, were hard to follow or
equipment failure. Discussed troubleshooting or explained that vague, or if the errors
Discussion necessary 6 everything had worked. were improbable Section missing
Results
Converted the thermistor date from resistance to Converted the data, but
temperature, using the correct part number and Converted the data used the wrong column of Did not convert the
Q1 table. 6 correctly the table (part 44004) data.
Produced the proper
Created a calibration plot with three curves, one plot, with labels,
for each device they used, with the platinum RTD equations, R2 values,
data as the standard. Each curve was fit to a second three curves, and the
order polynomial with the equation presented and correct data on each Missed 3 or more of
Q2 the R2 value given 6 axis (RTD on x axis) Missed 1-2 of these items these items
Used the correct
equation, did the
correct derivative, only
did this for Used the correct Used the wrong
Used the equation for the second order polynomial thermocouples, and equation, but either made equation, or did it for
for the thermocouples, took the first derivative, clearly presented the a math error or did not something other than a
Q3 and used that to determine the sensitivity at 75 °C. 6 sensitivity at 75 C present the answer clearly thermocouple.
Produced the proper
plot, with labels, three
curves, properly
Determined the deviation from the standard, calculated deviations
plotted the deviation as a function of temperature, and the correct data on
with three curves on the plot, connected by each axis (temperature Missed 3 or more of
Q4 smooth lines 6 on x axis) Missed 1-2 of these items these items
Both data sets are
Corrected their data for the unknown bath using a provided, properly
calibration curve developed for each device. Both labeled, and properly They had any math errors,
corrected and uncorrected data should be provided corrected based on the or labels were missing or
Q5 in a labeled table. 6 calibration curve confusing Either data set missing
Developed a table of the results for the unknown as Missed 3 or more of
determined from the three different devices, along Had all the required these items (labels,
with expected accuracy and standard deviation. (It elements in a clearly accuracy, results, or
Q6 is okay if they combined this table with Q5) 6 labeled table Missed 1-2 of these items st.dev)
Analysis Give 5 points if Give 3 points if Give 0 points if
Discussed the accuracy of the three devices,
naming one as most accurate and one as least They clearly discussed Were unclear in their
accurate. Based their decision on the deviation all the devices and gave discussion or did not use
Q1 data and plot 5 evidence from the data the evidence Answer missing
Discussed the precision of the devices, naming one
as most precise and one as least precise, and They clearly discussed Were unclear in their
explained if they did not feel they had enough data all the devices and gave discussion or did not use
Q2 to make this determination. 5 evidence from the data the evidence Answer missing
Provided the plot, well
Provided a plot of the thermistor data, fit a linear labeled with the proper
trendline, and determined the β value, explaining trendline, explained
what it means and what errors could lead to any errors, and
nonlinearity (IF they did not use the thermistor, determined the β value Missed 3 or more of
Q3 this is a 'free' 5 points) 5 and what it means Missed 1-2 of these items these items
Provided a discussion of the limitations of a liquid Explained the benefits
in glass thermometer, the benefits, and a and drawbacks and
discussion of the greatest possible accuracy from discussed the Missed 1-2 of these items
this device. (IF they did not use the LIG maximum possible or were unclear in their
Q4 thermometer, this is a free 5 points) 5 accuracy discussion Answer missing
Compared the two thermocouples in terms of
sensitivity, accuracy, or other behavior that they Clearly compared the
noticed. Discussed which would be more accurate thermocouples with all Missed 1-2 of these items
at this range (If they only used one thermocouple, the correct items, using or were unclear in their
Q5 this is a free 5 points) 5 evidence from the data discussion Answer missing
Described the possible sources of error, considering
the thermocouple itself, the connection to the Multiple sources of The explanation was More than one
DAQ, the internal voltage meter, and other error considered, all of unclear, or 1 of the errors implausible error or
Q6 plausible sources 5 which are plausible was implausible answer missing
Provided the best answer for the unknown
temperature, discussed their range of error, and Provided their answer, They had too many
explained how many decimal places are significant. to a reasonable decimal places, discussed
(Check to see how close they are to the known number of decimal things unclearly, or were More than 10 degrees
value for their temperature bath number. Group places, and explained more than 4 degrees off off, or did not discuss
Q7 that gets closest gets 5 bonus points) 5 their range of error from the known value error
 Explained what changes they could used to Gave reasonable ideas
improve the accuracy of the experiment.Explained and discussed whether
whether or not they would use certain devices certain devices were
Q8 again, and gave good reasons for their answer. 5 worth using again Missed one of these Missed both of these
100
If they did not bring a LabView VI, subtract 5
points from each person's report. It doesn't
matter if the VI worked the first time, but they
had to bring one.