Wednesday, July 18, 2012

Questions for Genetics Students

Rosie Redfield has published a (deliberately) provocative article on teaching genetics in the 21st century (Redfield, 2012). I disagree strongly with her premises and her conclusions but the issues are complex—too complex for a single posting.

Let's start by looking at one aspect of her proposal.

Rosie thinks that a 21st century course on genetics should focus on information that students can use later on. Here's how she would begin her new course ...
Box 4 gives a suggested syllabus for a 21st century genetics course. It begins with a human focus, introducing personal genomics and our natural genetic variation. Students then learn about the underlying molecular explanations—how differences in DNA sequences arise and evolve, and how they cause differences in phenotype—followed by how genetic differences are inherited and recombined.
Long-time Sandwalk readers will know that I am opposed to the idea that we should teach fundamental science by focusing on applications. They will also know that I'm not a fan of appealing to what students are interested in instead of what we experts think they should know. Catering to the prejudices students is a cop-out.

Finally, I think we should try whenever possible to wean students away from an anthropological perspective on science. They enter university with the misconception that humans are the most important thing on the planet and we should try, whenever possible, to teach them about other species.

As far as I'm concerned those are fatal objections to beginning a genetics course with an emphasis on humans and personal genomics. However, there are also practical matters.

The source of natural genetic variation is, indeed, a fundamental concept that should be part of any modern genetics course. It's not a simple concept. Students first need to know about mutation, including its source and its frequency. They then need to know about forces that remove variation, including natural selection and random genetic drift. Then they need to learn about the rates of these processes and the principles of population genetics; especially the effect of population size.

In order to understand human genetic variation, students need to appreciate complex population structure as well as simple Mendelian genetics.

Then there's the important question that occupied geneticists in the 1960s. Is standing variation maintained by balancing selection? [The Cause of Variation in a Population] (I learned recently that many evolutionary biologists still think that balancing selection is important.)

All these things need to be covered first. Then, and only then, can we introduce students to human variation and personal genomics. You don't start a course by showing students the most complex known example of genetic variation. You start with simple examples and basic concepts then work your way up to the more difficult examples.

Here's a list of questions that students need to answer, according to Rosie Redfield.
A good place to see these and other changes is headline news—Box 3 gives some high-ranking hits from a recent Google News search for “genetics”. These raise complex questions, both personal and societal, that our students will need to answer. Is genetic testing a wise thing to do? Is it a sound financial investment? Should I have full access to my genetic information? Should my insurer and my employer? Should athletes be tested for genetic modifications (“gene doping”)? Is it ethical to DNA-fingerprint all convicted criminals? All suspects? Did my genes make me gay? Are genetically modified foods safe? Are cloned animals ethical? How different are human races, and how different are we all from chimpanzees and gorillas?
Some of those are interesting questions and a thorough discussion of possible answers can go a long way toward teaching critical thinking. It would be very useful for students to debate those issues in tutorials, although I would substitute some others that had less of a human focus (e.g. why do species have sex?, are genomes full of junk DNA?, what is a gene?, where do superbugs come from?, do mutation rates vary?)

But let's keep in mind that these kinds of questions are not the purpose of the course. The purpose is to teach fundamental principles and concepts of genetics. If discussing the answers to certain controversial questions will advance the primary goal then I'm all for it. That's why exploring the purpose of sex is a good question and so is a discussion about the origin of superbugs and how to combat them.

And if you're teaching planetary astronomy then debating whether Pluto should be a planet is an excellent way to get into some serious science. I agree with Rosie that we need to move away from traditional memorize & regurgitate courses and we need to focus on student-centered learning. I just disagree with her solution.

Some of Rosie's questions are questions about ethics, not science. We debate some of them in my course but only after spending a good deal of time on the difference between science controversies and ethics controversies. (Starting this year, my co-instructor is a philosopher, Chris DiCarlo.)

In my experience, most scientists and graduate students are not capable of leading a discussion on human races, the ethics of DNA fingerprinting, or whether you should pay to get your genome sequenced. In order to do it properly, I think it takes a lot more skill than Rosie is willing to admit.

Redfield, R. (2012) "Why Do We Have to Learn This Stuff?"—A New Genetics for 21st Century Students. PLoS Biol 10(7): e1001356. [doi:10.1371/journal.pbio.1001356]


  1. I think your approach should not be grouped under student centered. All of the negatives things that you reject about Rosie's course at the beginning are put of the student centered approach. Amazingly enough considering the terminology overload in educational research, no one that I'm aware has bothered to introduce terminology beyond the misconceived teacher centered vs. student centered distinction.
    Your philosophy seems more similar to what I would call knowledge centered education. The instructor takes responsibility for defining the critical knowledge (not the student, if they knew what was important they shouldn't be there after all), and then the students are expected to show mastery of the material. At the university level this is far more than filling in bubbles on a test, and requires substantial active discussion and writing. As you frequently point out direct interaction is the only way to truly assess and attempt to stimulate critical thinking. With the big often missed fact that students actually need to know things in a fair amount of depth before they can think critically at all.

  2. I agree with Larry - the first emphasis should be on fundamentals. But the most important fundamental is still missing from the discussion: introductory genetics courses should introduce genetics as applied information science (not just applied chemistry) and demonstrate evolutionary processes in action. This can easily be done using computer simulation: one could start with simple toy models that simulate evolution of genotype and/or phenotype and show how various phenomena observed in real data appear as you add sophistication to the model. This would address the problems many people have with intuiting how random mutations and natural selection can lead to information-rich outcomes; with understanding the roles of drift vs selection; with understanding the complex interplay between genotype and phenotype; with understanding concepts such as linkage, extended phenotype, etc, etc - the possibilities are almost endless. It would also help students understand why they need a rudimentary understanding of computing and statistical modelling if they want to have an in-depth understanding of the complex processes underlying genetics.

  3. The big question is, what are our goals in teaching an introductory genetics course? Larry and Konrad's main goal appears to be to enable every student to become a geneticist. I think this goal wastes the time of the great majority of students.

    I agree with Konrad that genetics is applied information science, but a rudimentary understanding of computing and statistical modeling is only useful to people who want an in-depth understanding of the complex processes underlying genetics. Thus this is best taught in (or treated as a prerequisite for) an advanced genetics course.

    I think we're short-changing the students in an introductory course if we (a) teach them complex material they're not able to appreciate and (b) fail to give them the basic genetics principles they'll need in their non-academic lives. These principles are the true 'fundamentals' of genetics.

    1. Rosie, I'm assuming that we're talking about introductory genetics for the masses as you described in your article.

      At the university of Toronto we have two genetics courses; one for the students who will be specializing in genetics and molecular biology (BIO260H) and one for all other life science students (HMB265H). One of these (BIO260H) teaches genetics properly and the other one (HMB265H) is much like the one you want to have at UBC. One of them is considered to be a bird course by most students while the other has a reputation of being much more rigorous and harder to get a good grade.

      The goals I advocate are for ALL students, not just those who might want to become a geneticist.

      The primary goal is to teach critical thinking and how to think like a scientist. That goal applies to all science courses but we must never lose sight of it. If we are successful then those students who go on to become lawyers, politicians, physicians, and business leaders will have learned a valuable lesson.

      The goal that is specific to a genetics course is to teach the fundamental principles and concepts of genetics. That includes many of the things you talk about in your article but it must not come at the expense of distractions like personal genomics and the ethics of DNA fingerprinting.

      I don't agree with Konrad when he says that computing and statistical modeling are fundamental principles (or concepts) of genetics.

    2. @Rosie:

      "Thus this is best taught in (or treated as a prerequisite for) an advanced genetics course." - I agree. That's why I said "It would also help students understand why they need a rudimentary understanding of computing and statistical modelling if they want to have an in-depth understanding of the complex processes underlying genetics." An introductory course should help students appreciate which topics they need to know more about if they want an in-depth knowledge; it shouldn't actually teach that in-depth knowledge.

      Giving students hands-on experience with simulations that allow them to explore which assumptions are needed to make sequences evolve in an evolution-like way is hardly "complex material they're not able to appreciate" (I wasn't advocating that they write the simulations themselves :-) ). And I do think those assumptions (e.g. a source of random variation, natural selection, a complex relationship between genotype and phenotype) are fundamental to both evolution and genetics.

      @Larry: I didn't say that computing and statistical modelling are fundamental principles of genetics, I said that a rudimentary knowledge of these topics is required for an in-depth understanding of the complex processes underlying genetics. Not all genetics students will want to study those processes in depth, so not all of them need to study these topics. But all students in an introductory genetics course should be exposed to the idea that there is real (quantitative) substance to descriptions of evolution: it's not just a qualitative theory to be illustrated by discussing a few case studies, the way it operates can actually be _demonstrated_.

      In similar vein, the most fundamental concept in genetics is probably the nature of the map from genotype to phenotype. By playing with particular cases of this (say a protein-coding sequence with software that predicts folding - naturally, one would clarify that the prediction is not accurate), students can explore how some small genetic changes lead to large phenotypic changes while others do not. This ought to be a routine part of any introductory course in molecular biology.

  4. You are both so lucky. I teach (and really enjoy) an introductory cell biology and genetics course for non-majors at a US community college. Most of my students had their last biology class in the 9th grade (I ask). Their conceptions of how living systems work are so full of misperceptions that it takes an entire semester just to get them to the point where they seems to be getting what an allele really is....if you don't understand that a gene is made of DNA, that DNA consists of a sequence of nucleotides, that nucleotides are .... etc etc you can't really understand any of it

    For instance: In Fall 2011 I started doing the following in class - fill a ziplock bag with warm water, some bakers yeast and some sugar, mix, get the bubbles out and seal. I weigh it in front of the students and then lecture for a bit. Once the bag has swelled with some CO2, I have the students vote on whether the weight of the bag will change. I make very clear that no atoms can get into or out of the bag before the vote and give them time to discuss. 60-80 % of students (approximately 70 students total so far, majors and non-majors) think the weight will change (most of those think it will be lighter). These students do not believe in conservation of mass, they are not sure what matter is and they are very susceptible to believing that matter can magically appear and dissappear (which explains certain aspects of US society to me). My question at this point is: "how can they possibly understand metabolism at any level, where cells take solids, put them into solution, take them apart and turn them into gas?". I would argue the same is true for their understanding of genetics and population genetics --> they will learn the right answers for the test, but what good is that when they really don't believe it??

    I find that questions such as "how can your body make fat out of starch" to actually get some engagement in class precisely because it explains something fundamental about themselves. Giving students a basis in fundamental knowledge is very important, but in my classroom, small gains in real learning correlate with how many times the student has opened their mouth in class to make a point or ask a question - neither of which they do if they aren't interested.

    Which gets me to my second point about why you are both so lucky: in majors classes you really CAN say that your students need to know this material for later in their lives. That is not true for my non-majors students, even if I really wish it were. The fact is that you can get through your whole life with massive misperceptions of how science and biology really works (by biology, I mean living systems). All my data strongly suggests that even fundamental misunderstandings about genetics, meiosis and mitosis do not prevent these students from living happy, healthy lives (although, I owuld argue, very boring lives). As long as you perform your (non-science) job, pay your bills, follow your doctor's advice and keep eating you have a better than average chance of getting to reproductive age. I hear so many teachers proclaiming that their students "need to understand biology" for various reasons --> poppycock! My non-majors need to understand biology about as much as I need to understand how this computer I am typing on works. Our challenge is to get me interested in computers to the point where I will WANT to ask questions about how this computer works.
    I do not let this change what I do. Most of my criticisms (that I am aware of) from students is that my non-majors classes are "taught at the majors level" (snort), but I think it helps to be honest with ourselves about what our students "need" to know versus what we want them to know.

    1. It sounds like more than half our students should flunk the course because they are incapable of mastering the material.

      What's is the failure rate in your classes?

  5. Sometimes as low as 30% sometimes as high as 50%. The bigger failure is before they got to my class though.... they got out of a high school chemistry class without truly believing that matter is made of atoms. In some of my discussions about the yeast in a bag demonstration, I sense that some of them understand the fundamental relationship between matter/atoms but they do not know how to apply that understanding to this new situation. What they truly need isn't more facts, its practice "thinking about" the facts they already know. Next semester I am hoping to more formally assess if thats the problem. Incidentally - anyone interested in this whole fundamental misperception thing : visit the AAAS Project 2061 Assessment site... their questions aren't perfect, but are extremely revealing (especially when you give your students an "I don't know" option). Sigh, too many thoughts bubbling up.... must put something up in the blog before I lose them.

  6. With the exception of not promoting personal genomics, is the proposition by Redfield substantively different from what I proposed in 2009? (Dougherty, MJ. 2009. Closing the Gap: Inverting the Genetics Curriculum to Ensure an Informed Public. Amer J Hum Gen: 85, 6-12;
    My proposal suggests beginning with complex traits and variation and encompasses multiple levels of genetics instruction.

    1. Hi Mike,

      Yes, you anticipated many of my points. I've put a link to your paper as a comment at the end of mine. Sorry I hadn't seen your paper when I wrote mine - I would certainly have discussed it.

      Maybe we should have a separate blog discussion about how these proposals for change are received...

  7. Rosie, I have to disagree with your statement: "a rudimentary understanding of computing and statistical modeling is only useful to people who want an in-depth understanding of the complex processes underlying genetics. Thus this is best taught in (or treated as a prerequisite for) an advanced genetics course. "
    Students need a BASIC understanding of chi-square analysis to understand populations: ratios, Hardy-Weinburg and GWAS. They can learn a DEEPER understanding of statistics later, after they have some idea that at least it is important for drawing firm conclusions from population data.
    Regardless, I loved the article (Mike's too), and applaud your efforts!