Tuesday, May 18, 2010

Guest Speaker: Richard McElreath- UC Davis Antropology

Professor Richard McElreath came to speak with us yesterday at GTC. He discussed the major challenges of and techniques for teaching formal models of evolution to undergraduate and graduate students. Each point in bold represents Richard's own words, while the unbolded words are my own notes from his talk. Enjoy!

1) Anxiety --- Learning to model is like learning a foreign language, but easier; it only takes hard work, not much talent. Students need realistic expectations of how quickly (slowly) they will progress.

a. Biology graduate students are just as anxious as undergraduates concerning math, in fact graduate students in some cases are more anxious because it has been longer since they have had any serious math courses.

b. There is a culture that exists in many scientifc fields (especially biology) where People who can do mathematical models are put on pedestals and considered to be 'smart people.' This is not nessesarily true, as talent is responsible for only a small part of what gets people to that point.

c. Techniques to make students less anxious about mathematical modeling:

i. modeling is a language (like a foreign language), and people can slowly learn units (or parts) of the equations at a time. They don't need to learn everything at once.

ii. Mathematical modeling is not a talent! No one is born doing calculus. Two people discovered calculus at the same time because the math had built up to that point by predecessors. Pratice is what makes people 'good at math.' In fact there is an good example of this in athletics. A strangely large number of European soccer pros are born in January and February. This is not due to zodiac signs! They are the oldest in their elementary school class, so their athletic ability is better, and so they get socially reinforced that they are better. This makes them work harder and more likely to become pros! (I believe this story and others like it are talked about in Outliers: The Story of Success.)

iii. most people give up on understanding math (because it seems hard) and so if you don’t, you are going to be much more hirable in the future.

2) Gender-biased anxiety --- Girls now out-perform boys in math and science at all grade levels up through high-school. In Scandinavian countries, this has been true for 20 years. Also, tell the class the homeworks are "easy," because this helps to close the self-stereotyping gap between men and women.

a. Many students engage in "Self Stereotyping": They feel they don’t have to be good at something if they are part of a certain demographic that is stereotyped to be poor at this particular skill. (I've heard that if you move the part of an exam that asks about the demographic information of the exam taker to the end of the exam instead of at the beginning, students do better! I need a citation on this though.)

b. Social Dominance (Sidanious, Pratto) shows you can completely erase the gender gap (exam performance difference) by telling the students that the test is easy! If you tell students it’s hard, you increase the difference between boys and girls.

3) Philosophy of science --- All models are false, so complaining about "unrealistic" assumptions is pointless. Rather, we want useful models, and those are often very simple and unrealistic models.*

a. Models are like maps OR they are a simplified version of the real world

b. Some assumptions are made that are obviously wrong-some models are deliberately unrealistic (infinite populations, frictionless surface)

c. They are not invented just to make the math easier, it’s to get an understanding about how things work

d. Representing and Intervening (Ian Hacking): This is a book about the role of experiments in biology, does the last 300 years of philosophy of science in 30ish pages.

4) Participation and practice --- You can't learn kung fu by watching Jackie Chan movies, and you can't learn to model without doing the math and making mistakes. Students need to practice the algebra, most often, because this is their weakest skill.

{culture is different in biology, there is a need to show off the most state of the art stuff, which is not the most conceptually powerful to do}

a. Students gain a lot from simply pausing and processing during class time. Give them a moment to type things into a calculator does a lot to stimulate understanding. This is an example of how to make class interactive!

b. Homework problems should be designed to be interesting and possible (this is difficult!)

c. Students should not lose credit for poor algebra because it is most important for them to learn the process, so that's what they should be getting and losing credit on

d. Don't try to humiliate students, but they have to get up and try to solve the problem. (They need to be involved in the learning process.)

e. Try to have students work in teams so they don’t get hung up on math, each person has a different skills, and each has a different issue. By pairing them together they are more capable to attack any problem.

GTC sends a big thanks to Richard for visiting us on Monday!

* The folk Popperian philosophy of (naive) falsificationism that most scientists subscribe to just doesn't work---all theories are incomplete approximations of reality, and therefore "false" a priori. No experiment can disconfirm a theory, because the theories are born false and only aspire to be useful.

Wednesday, May 5, 2010

Making assessments work for you ... and your students

What better way to begin a discussion about assessment than with an assessment?  We began with a somewhat tongue-in-cheek quiz about the "right" way to conduct an assessment, modelling some common faults in poor assessments while also serving as a brainstorm to get some ideas out there.  We presented some background theory on assessment, drawing from the excellent text: Handelsman, J., Miller, S., & Pfund, C.  Scientific Teaching, highlighting three reasons for assessment:

  • Quantifying student performance, measure students as learners
  •  Measuring our own effectiveness as instructors
  •  Motivate students

Handelsman offers a particular focus on the third with their concept enGauge, tools that both measure performance and motivate engagement with the material, often in real-time setting rather than waiting to the final exam.  From here we moved into general discussion that visited various challenges in assessment, particularly in teaching and assessing ''process'' rather than factual knowledge.   Moss (2001)  also has an excellent discussion of this challenge.  We discussed some unusual examples of assessment, such as Lee Sheldon's XP based grading system, motivated by World Of Warcraft.