Title: An Industry Leader's View of Engineering Education
1An Industry Leader's View of Engineering Education
- Steve Kirschstk_at_propel.com
2Agenda
- My qualifications
- Ideas on improving
- pre-school
- design and process
- feedback
3Why should you listen to me?
- I have no credibility in the field of education
- but the EECS auditorium at MIT is named after me
- I represent the customer
- A fresh perspective
4Engineering a better education
- 3 ways to make a better product
- Better parts (raw materials)
- Better product design and manufacturing process
- Use feedback for quality assurance (QA) and for
ideas on how to continuously improve - Same is true of education
- Better educated students coming into the system
- Improve what we teach, how we teach it, and who
teaches it - Ask students 0-15 years later for feedback and
actually use it to change the previous 2 steps
5Improving the quality of the incoming students
695 TIMSS results42 countries
7Texas had state/local control imagine what a
focus on testing can do for the rest of the
country!
8The vision
- All students in the USA receive the best K-12
education in the world
TIMSS
12th grade
4th grade
9How do we do that?
- Establish clear, measurable goals
- Example For K-12, it could be NAEP or TIMSS
improvements - Change what we do
- Copy what works
- Stop doing stuff that doesnt work
- Measure and adjust
- This isnt rocket science. Its just that we
dont do it very well.
10Changes required
- Understand why we have failed in the past
- Avoid those mistakes
- Realize that K-12 is a federal problem not a
local problem the states have had their chance
for over 200 years, yet no state stands out of
the pack (NAEP). State/local control can make
things dramatically worse, e.g., Texas results
on TASP. - Understand why other countries have succeeded
and adopt best practices of top performing
countries. Copying first to get to parity,
innovating later. - Have the leadership and courage to do things
dramatically differently than we are today - Create a vision, clear goals, and a believable
strategy based on what has proven to work
11Why isnt there a checklist for education that we
actually use?
- In aviation, a pilot uses an extensive proven
checklist to ensure a safe flight - Why not offer a substantial on-going cash
incentive to enable schools to pass a proven
checklist of statistically proven replicable
requirements that ensures a quality education?
12Why dont we run K-12 schools like airlines?
- A. Airlines governed by national safety standards
(not set by airlines!) - B. All pilots are qualified and certified to fly
the plane type (no unqualified substitutes
allowed!) - C. On-going pilot training is required
- D. Pilots who dont perform can be fired
- E. Planes that dont meet code cant be used
- F. Pilots free to determine how to fly the plane,
but not the destination - G. Require pilots to go through a proven safety
checklist (that ensures a successful flight)
before takeoff
13How bad are things today?
- There are gt50 state standards for what students
are to learn. Everyone else has 1 national
standard. - Alignment (standards, curriculum, materials,
assessments) is therefore virtually impossible - Imagine if Bill Gates were forced to write
Windows for each state? - We are set up to fail.
- There is no good reason we cant have national
standards. That is step 1. - NAEP is perfect proof of de facto national
standards. It can be done.
14Changing how we teach
15Why the sudden discontinuity after 12th grade?
16Methods to improve the process
- Set clear, measurable goals
- Understand very clearly what you are trying to
achieve - Copy what works
- Copying is always the most efficient way to get
to world class performance - Innovate later
- Innovation is required for progress AFTER you
are 1
17Metric ideas
- TIMSS-like standardized assessment tests
- Self-perceived customer satisfaction
- Peer/supervisor rated satisfaction
- employed after 2 years
- who were fired in first 10 years
- Average salary improvement after 4 years
- who got jobs after graduation
18Wrong way NCLB
- of schools that NCLB was tested on before being
rolled out nationally - ZERO
- Amount of improvement that can be expected from
NCLB - Nobody has a clue. Could be negative.
- Even worse we knew Bushs model (TASS) failed
- We should NEVER be experimenting on our kids at
mass scale like this!!
19Why not treat education policy like drugs?
- Efficacy
- Require proof of efficacy, i.e., it has to work
- Safety
- Require testing on a diverse population to look
for unintended consequences (negative side
effects) - Scalability
- Prove that it can be rolled out in scale and
still work (e.g. GMP manufacturing standards)
20Right way NCEEs America Choice
- Research best practices
- Spend 11 years studying best practices in other
countries - Prototype
- Create a prototype approach based on the
learnings but adapted for the US market - Test and validate
- Test it at few schools
- Get a third party to validate results by
comparing to other schools of similar profile - Scale
- Roll out to more schools
- Re-test
- Measure again to make sure it scaled with same
results and without any unintended consequences
21Changing what we teach
- Give people what they need to be successful
22The Kirsch dietIs it good for you?
- Genetically, we are all virtually identical. So
shouldnt this pie chart work for everyone? You
can just pick the foods you want, but EVERYONE is
required to stick to the s
23MIT
- 360 units to graduate
- 48 units humanities (pick from traditional
subjects) - x units core subjects
- y units elective
- z units General Institute requirements
- They review this every 50 years, whether they
need to or not - Why arent there multiple pie charts? Does one
pie really fit all? - executive
- Manager
- entrepreneur
- researcher
- engineer
24MIT
- In 2004, they discovered that communication is
more than technical writing
25Why arent we teaching soft skills?
- Presentation skills
- Teaching skills
- Leadership
- Teamwork
- Negotiation skills
- Nonverbal communication
- Decision-making
- Company politics
- Giving feedback
- Receiving feedback
- Sales skills
- Attire
- Philanthropy
- Behavioral change
- Hiring and firing people
- Interviewing skills
- Social skills, e.g., how to get a date
- Basic project management skills
- Managing complex projects
- Organization skills
- Coaching/mentorship
- Working a room
- Running a meeting
- Conflict resolution
- Public policy
I would love to take such a course - Tony Eng,
MIT
26Soft skills
- If there is one change you can make right now to
improve engineering education, this is it - Few offerings are comprehensive
- Even fewer universities (none?) require this to
graduate - Dont try to integrate it into existing course
copy what works - These skills have been taught very successfully
in standalone single-topic seminars
27MIT hasnt got it yet (after 20 years of
prodding)
- 6.UAT/6.ThTÂ Preparation for Undergraduate
Advanced Project/Masters of Engineering Thesis - Upon completion of the course, students will have
learned how to - propose and define research problems and think
about solutions - critically evaluate technical presentations
- architect technical presentations
- present technical material in oral and written
forms to different audiences at different levels
of detail - give and receive constructive feedback
- write progress reports
28Why not also teach basic survival skills?
- Basic investment skills
- Do you know why a stock goes up when they
announce bad news? - Do you know why most people who trade stocks
frequently lose money? - Personal relationships
29What else should we teach?
- The answer is in the feedback section!
30Changing who is doing the teaching
31Faculty changes
- Require a 6 week training course for new faculty
- West Point does that
- Require people who teach engineering to have
real world experience in industry - Example take off a semester and get a full time
job - Change incentives to reward collaboration,
teamwork, and teaching - Not just individual research!
32Measure and adjust
- Creating a feedback system for quality assurance
and continuous improvement
33Education is an open-loop op amp
Jobexperience
Collegeeducation
Students in
Students out
34Example of using feedback Why executives fail
- Interpersonal skills (CFO, VP Engr)
- Lack of teamwork
- Lack of confidence in the team to work together
(VP Sales) - Inability to execute
- Use their knowledge and skills to lead a team to
success (VP Marketing)
35Whats common?
- They failed on the soft skills
- Nobody failed due to a lack of understanding of
the basic subject matter in their core area of
expertise - Everyone failed due to an inability to translate
their knowledge into action
36The Paul Cook success model
- Q How do you spend your time?
- A On people problems.
- So feedback should measure
- what skills are the most used
- what skills they could use to be more effective
- what skills were required for success
- what skills led to failure
37My experience at MIT was lopsided
- Technical skills
- BEFOREltltAFTER
- People skills
- BEFORE AFTER
38Why isnt there a feedback system in place?
- What are our goals?
- Do we have metrics aligned to those goals?
- Do we know what they are?
- Is it used as a basis for compensation?
- Is it used as a basis for adjusting the process
and the design?
39Methods for driving change
- Change the change process
- Internally driven changes at universities are
typically done on geological time scales - To change, copy what has worked at universities
who have solved the lack of change problem - Start simply Pick one or two metrics for
driving improvement - ABET can drive changes
- Establish standardized customer satisfaction
metrics - Phase in a few key checklist items Are new
faculty members required to learn how to teach? - Phase in a few key optional items Do students
demonstrate mastery of soft skills? - etc.
- CASEE could create and market a single J.D.
Power metric for rating engineering schools on
customer satisfaction - Could be coupled with a prize for most
improved - A single combined metric is easier to focus on
- Can be component metrics
40Influence Public Policy
- e.g., your vote on Nov. 2
41Significant negative impacts from Bush decisions
- Cutting federal grants NSF, etc.
- Putting our kids at a permanent disadvantage
- Deregulating mercury emissions from power plants
has led to the possibility of permanent brain
damage in 15 of the children born in the US
today - The requirement to remove 95 of the mercury
emissions would have costed 1 of the cost of the
plant - Ignoring/distorting science
- Cutting funding on his own education bill from
what Congress proposed - Believing unbelievable test results (TASS) then
using that as a basis for national policy - Unfunded mandates
42Summary
- Improving education is a lot like building a
better computer - Quality components Demand the highest quality
components from our suppliers help them improve - Customer driven design Adjust the design of the
product to fit what people want to buy (a
successful career) - Quality volume manufacturing Improve the
manufacturing process by copying best practices
and making sure we have clear manufacturing specs
(goals) and that the products meet the spec
(testing) at high volume (scalability) - Feedback Take a look at the product after its
been in use for a few years to ensure quality and
provide ideas for improvement
43Summary
- Nothing in this talk is new
- Its just that we dont do it (very well)
44Leading indicators of success
- A few clear, measurable goals established and
used - People skills being required to graduate
- Changes are prototyped and proven to work before
scaling (no more NCLB) - Adoption of national standards (K-12)
- Importing best practices from other countries
- Federal incentives to adopt strategies that are
proven to work - New faculty are required to learn how to teach
- Changes take a year instead of a century
- Changes driven by customer satisfaction metric(s)
45Latest copy of this talk
- On my website at
- www.skirsch.com