slides

Report
Designing assessments for an edX course:
a multi-stage approach
Simona Socrate
Principal Research Scientist, ISN
Senior Lecturer, MechE, MIT
Some underlying “philosophical” questions
•
Purpose of assessment exercises:
–
–
–
–
–
•
reinforce comprehension of “lecture” material?
promote critical thinking /problem solving skills?
provide “connecions” to experiential learning?
validate pedagogical approaches for the teacher?
measure/certify mastery of material (for grading purposes)?
Practical considerations:
– what assessment approaches are feasible in a mooc?
– cost/benefit of enhanced interface complexity: is it warranted?
– what (sustainable) student workload meets course objectves?
•
Is what we are measuring relevant to the core
pedagogical objectives of the course?
Some challenges and opportunities of the mooc model
• Challenges
– Recast all assessment exercises to work with an automatic
grader
– Assess problem solving “methodology” with an automatic
grader (sketches? procedure?)
– Minimize operational barriers and sources of frustration
• Opportunities
– Large scale study on the effectiveness of alternate
assessing approaches ( quantitative vs. conceptual
exercises)
– Emergence (the forum and the mind of the group)
2.01x A multi-stage approach
• Stage 1: first release
– Recast all assessment exercises developed in the residential MIT 2.01
course to work with an automatic grader
•
•
•
•
•
•
multiple choice
numerical answers
numerical answers with unit selection
symbolic answers
MATLAB problem (x-y plotting and solution of linear systems of equations)
graphic selection
– Separate assessment exercises into
•
•
•
•
Examples (inlined with lectures)
Recitations (with instructor videos)
Homework Problem (weekly, counting towards student grade)
Quizzes (two quizzes measuring overall retention/comprehension, counting
towards student grade)
– Provide some plotting functionality and introduction to numerical
methods by embedding MATLAB into some of the exercises
Problem2 [15 Pts total] You need to indicate units in your answers for full credit
The composite beam AB of length
L=2 m,
is fixed at B (x=L) and is composed
of a round cylindrical core of constant
radius
R0= 1cm
bonded inside a sleeve of thickness R0
(outer radius 2 R0= 2cm).
q0 (x/L)
y
EO
RO
2RO
3EO
B
x
A
L
The beam is loaded by a downward
linearly varying distributed load per unit length:
!!
!!! ! ,
with q0 = 2.76 KN/m
as indicated. The material moduli are,
for the core: EC = 70 GPa = E0
and for the sleeve: ES= 210 GPa =3 E0
(1) [2 pt] Obtain a symbolic expression for the bending moment M (x) in terms of L and q0.
(2) [3 pt] Obtain a symbolic expression for the effective section stiffness of the beam (EI)eff in
terms of R0 and E0 . Calculate the value of (EI)eff .
(3) [2 pt] Obtain a symbolic expression for the curvature at the neutral axis, 1/r(x), in terms of
L , q0 , R0 and E0 . Calculate the magnitude (absolute value) of the largest curvature along the
beam and indicate its location.
(4) [3 pt] Obtain a symbolic expression for the slope of the beam, q(x), in terms of L , q0 , R0 and
E0 . Calculate the magnitude (absolute value) of the beam slope at the free end (x=0).
(5) [1 pt] Set up the integral to calculate the vertical displacement of the beam, v(x), with
appropriate bounds and boundary conditions. You do not need to solve the integral.
(6) [4 pt] Calculate the maximum tensile stresses in the core and in the sleeve. Sketch the stress
profile, sn(y), on the corresponding cross section of the composite beam.
2.01x A multi-stage approach
• Stage 1: first release
– Recast all assessment exercises developed in the residential MIT 2.01
course to work with an automatic grader
•
•
•
•
•
•
multiple choice
numerical answers
numerical answers with unit selection
symbolic answers
MATLAB problem (x-y plotting and solution of linear systems of equations)
graphic selection
– Separate assessment exercises into
•
•
•
•
Examples (inlined with lectures)
Recitations (with instructor videos)
Homework Problem (weekly, counting towards student grade)
Quizzes (two quizzes measuring overall retention/comprehension, counting
towards student grade)
– Provide some plotting functionality and introduction to numerical
methods by embedding MATLAB into some of the exercises
2.01x A multi-stage approach
• Stage 1: Lessons Learned
– Be very (VERY) careful with problem statements: rethink
problems in terms of the grader capability
– Pick and choose where you introduce complexity: make sure
the additional barrier for the student is NECESSARY for
pedagogical purposes
– Online students are (for the most part) understanding and
appreciative. Be honest and apologize/correct as needed
– Accept the fact that the mooc teacher is on a (steep!)
learning curve as well: perfect is the enemy of good
2.01x A multi-stage approach
• Stage 2: second release
– Introduce Conceptual Problems
– Extend MATLAB functionality to enable “design” problems
• Stage 3: third release
– Introduce a graphical interface (similar to the circuit interface
in 8.02x) for structural elements
– Embed a numerical (Finite Element) structural solver.
– Enable direct peer feedback?
Do NOT forget the “philosophical” questions
•
Purpose of assessment exercises:
–
–
–
–
–
•
reinforce comprehension of “lecture” material?
promote critical thinking /problem solving skills?
provide “connecions” to experiential learning?
validate pedagogical approaches for the teacher?
measure/certify mastery of material (for grading purposes)?
Practical considerations:
– what assessment approaches are feasible in a mooc?
– cost/benefit of enhanced interface complexity: is it warranted?
– what (sustainable) student workload meets course objectves?
•
Is what we are measuring relevant to the core
pedagogical objectives of the course?

similar documents