Teaching personal genomics - the commercial issues

(Apologies for the long interval of dead air...)

I'm working on the video lectures for my upcoming MOOC Useful Genetics, and I'm stalled at Module 5, on personal genomics.  I know from our preliminary survey that this is something many students will be especially interested in. 

One factor that makes these videos different from the rest is that I'll be discussing services that are provided mostly by for-profit companies.  When I taught this in my face-to-face class, I found that students were particularly sensitive to this -- when I said positive things about 23andMe's interface, or marveled about how inexpensive genetic testing has become, they worried that I might have a commercial interest in the providers.

Another issue is copyright.  I don't think I can use images from commercial providers' pages without their permission, but obtaining such permissions is likely to be a big pain in the butt.

So I guess the solution is to avoid mention of specific providers, and just discuss the general types of services that are available.  Students will want to know how these analyses work, what kinds of information they provide, why consumers might choose to use such services, how to choose a particular provider, and how to interpret the information they obtain. 

The students can discuss specific providers and their services in the forums.  My job is to provide the information and issues that will help frame these discussions.  I think I'll want to explicitly explain this approach in the overview video for this module.

First lecture video

Yikes!  I just recorded my first draft of a lecture video, and it's 25 minutes long!  This is twice as long as I want my videos to be.

I don't feel that I covered a lot of material, but I'll see how long the next few are before panicking...

I have a plan (for developing the lecture videos)

Yesterday most of the people involved with UBC's Coursera afforts met to discuss how things were going.  One result of this is that I've changed my plan for preparing the ~60 lecture videos. 

The general plan has been that I'll record the videos in my office, using the built-in camera on my laptop, and a good external microphone and lighting provided by CTLT.  We had set this up using my main office computer (a new MacBook Pro), but switching back and forth between recording and everything else I need to do with that computer was going to be a big hassle.  So yesterday I re-set it up a 5-year-old MacBook Pro that I don't use any more (battery dead, a bit unreliable).

I had been assuming that I'd just start recording some, and then rerecording them as needed to improve the content and the presentation.  But now I think I'll instead try to quickly develop the content (mostly PowerPoint slides) and record them all as voice-over-annotated-PowerPoint, with no video of me (camera off).  Once I've done them all I'll go back through them, deciding what needs changing, what self-test questions to add, and when and where video of me should be added, and then I'll rerecord them with the camera on.

This is good in several ways.  First, it removes the issue of my appearance from the initial development and recording of the lectures - that's the part I was finding most stressful.  Second, the revisions won't start until I have the full set of lectures done - that will help ensure that the whole set is coherent.  Third, it maximizes the amount of recording experience I'll bring to the final recordings.  Fourth, it means that we'll quickly have a full set of lecture videos to help us develop all of the associated material (resource links and the various assessments).  Finally, because the camera will be off, I can move the whole recording setup to a little desk on the other side of the room rather than cramming it onto one side of my main desk.

If I set a goal of doing at least one voice-only video per day, starting today, I should have a full set by early February, maybe sooner.  Revisions will take time, but the re-recording should go fast, because I can probably do a full week's worth in a single day.

MOOCs and universities - a tragedy of the commons in reverse?

(This is an idealistic and probably not-very-well-thought-out post.  Comments and critiques are welcome.)

Ecologists often refer to the 'tragedy of the commons', the destruction of a shared resource due to uncontrolled overexploitation by individuals.  The concept was popularized by Garrett Hardin; he described it here. 

"The tragedy of the commons develops in this way. Picture a pasture open to all. It is to be expected that each herdsman will try to keep as many cattle as possible on the commons. Such an arrangement may work reasonably satisfactorily for centuries because tribal wars, poaching, and disease keep the numbers of both man and beast well below the carrying capacity of the land. Finally, however, comes the day of reckoning, that is, the day when the long-desired goal of social stability becomes a reality. At this point, the inherent logic of the commons remorselessly generates tragedy.
As a rational being, each herdsman seeks to maximize his gain. Explicitly or implicitly, more or less consciously, he asks, "What is the utility to me of adding one more animal to my herd?" This utility has one negative and one positive component.
1. The positive component is a function of the increment of one animal. Since the herdsman receives all the proceeds from the sale of the additional animal, the positive utility is nearly + 1.
2. The negative component is a function of the additional overgrazing created by one more animal. Since, however, the effects of overgrazing are shared by all the herdsmen, the negative utility for any particular decision-­making herdsman is only a fraction of - 1.
Adding together the component partial utilities, the rational herdsman concludes that the only sensible course for him to pursue is to add another animal to his herd. And another.... But this is the conclusion reached by each and every rational herdsman sharing a commons. Therein is the tragedy. Each man is locked into a system that compels him to increase his herd without limit -- in a world that is limited. Ruin is the destination toward which all men rush, each pursuing his own best interest in a society that believes in the freedom of the commons. Freedom in a commons brings ruin to all."

The outcome is destruction of the commons by overgrazing, and the unsavory solution is land-privatization, with individual landowners fencing off the land they graze.  Each privately fenced pasture has only a limited carrying capacity, but private ownership provides the incentive to ensure that its capacity is not exceeded, thus preventing its degradation by overgrazing.  The same problem is now being seen in the world's oceans, with local restrictions on fishing seen as the only practical solution to the destruction of shared fish stocks by overfishing.

I'm now wondering if the introduction of MOOCs into the higher education ecosystem might have the opposite effect - converting what used to be private resources for the elite into a shared commons of learning.

Under the current system of higher education, each university is a private stock of learning, with access available only to those who pay the tuition and meet the entrance requirements.  This has been inevitable because education can only be provided face-to-face to small local populations of students, and because the costs of this delivery are borne by the individual universities.  Most attempts at distance-learning had limited success, still needing substantial personal interaction and entailing high costs.  In ecological terms, the carrying capacity of each university is small, and tuition and enrollment limits prevent overuse.

Viewed simplistically, the major cost components of university teaching are (i) course development; (ii) course delivery, and (iii) assessment of student learning.  Course development remains expensive, but the other components have changed.  The internet greatly reduced the cost of distributing information, but the development of web-based 'learning management systems' such as Moodle and Blackboard is really what changed the economics.  These systems allow instructors to efficiently deliver information to defined groups of students, and to track the activities of the students.  Most important is their ability to carry out automated testing, where instructors design test questions that are automatically graded by the learning management system, These questions allow students to easily test their own understanding, and allow instructors to carry out formal (graded) assessment of learning progress without the need for human grading.  A further advance is the availability of software that supports peer-grading, first training students to assess the written work of fellow student, and then managing the distribution of work and recording of results.  These automated methods of evaluation reduce or eliminate the need for instructors and teaching assistants to grade student work.

Learning management systems make MOOCs possible, and MOOCs change the economics of higher education.  Once educational resources have been developed, the internet allows them to be delivered to large numbers of learners with no geographical limits and at very little cost.  Learning management systems allow students to be tracked, guided and assessed, again at a cost that's not only low but largely independent of the number of students.

Much of higher education can now exist as a global resource (very well-mixed, not patchy and local).  The 'carrying capacity' of a single MOOC (the number of students that can be taught) is very high, maybe even unlimited.  The cost to a university of providing a single MOOC is quite small, relative to its total budget.  If each of the approximately 10,000 major universities around the world were to provide just one MOOC, the learning commons would likely cover just about everything that is now taught in undergraduate programs.

Like the individual herdsmen in the tragedy of the commons, the individual universities would pay an individual cost (course development for their MOOC) and gain individual short-term benefits of their MOOC (reputation, public relations).  They would also pay a long-term cost, that of reducing the market for formal university education on which their success depends.  Like the cost of a degraded common pasture to all herdsman, this cost is experienced by ALL universities, not just those that provide MOOCs, so it is not an effective deterrent to MOOC production.

The critical difference from the tragedy of the commons is the effect on the commons (the global population of learners and MOOCs): rather than being degraded by this shared use, the commons would be enriched.

In the very long term this process might drive many universities out of business.  That might be its own kind of tragedy, but as long as the modest cost of providing individual MOOCs could be met in some other way, it would be a triumph for higher education.

Recording videos for Coursera - the technology

We (me and the guys of the Centre for Teaching and Learning Technology) have started working out the best conditions for recording all the lecture videos. We have a lot of factors to consider - for most of these Coursera provides detailed advice, which we're modifying to suit our circumstances.  This is all very new to me, so what I've written below will probably sound very naive to anyone with video experience.

Planning the Useful Genetics weekly modules

Part 1.  Genotype and phenotype

We should have one learning objective for each video (one video for each learning objective)

1.  How much humans are the same genetically, and how much we differ.  DNA, genes, chromosomes and genomes are all both physical entities and informational entities.  One video for each of these (with usages and representations)?.  Ploidy and the basic cycle of sexual reproduction. Populations, races and out of Africa. SNPs?

2.  How DNA molecules become different. Comparing DNA sequences.  Polymerases.  Mutations happen and are passed on to the next generation.  DNA repair.  Mutation rates and frequencies.  70 new point mutations in each baby, most from Daddy.  Do we need to worry about mutagens (yes for cancer, no for babies)?  Start considering how DNA differences affect what genes do (this will also teach more about genes).  Genes and proteins - what proteins do.  The genetic code.

3.  Lots about how DNA differences (mutations and polymorphisms) affect proteins and protein functions (or not).  Heterozygosity issues.  'Mendelian' and 'quantitative' effects (not these terms). Chromosome differences, aneuploidies.  Gene families.  Homologous genes in other species let us use animals as 'models' for human diseases.

4.  Predicting phenotypes from genotypes.  Sex chromosomes, X inactivation.   Genes and cancer.  Thinking about risks and probabilities.  Genes and behaviour.

5.  Personal genomics.  What can we know (will we be able to know) about our genotypes?  Gene-typing.  SNP-typing.  Exome sequencing.  Genome sequencing.  Transgenic organisms.  Genetic modifications (GMOs). Gene therapy.  Forensic DNA identification. 


Part 2.  Inheritance

6.  Can I start with something catchy?  The mechanics of inheritance.  DNA structure again.  Chromosome structure again.  Mitosis: the problem and the solution.  Meiosis uses this solution but now problem with new solution.  Mating"  gametes don't know their genotypes, random encounters.  Following genotypes through.  Physical molecules and information again.

7.  Consequences of the mechanism of inheritance.  Almost all variation was present in a parent.  Probabilities.  Interacting with risks and with other genes.  Errors in the mechanism (cause translocations, aneuploidies etc.)

8.  Linked genes, genetic maps, sex-linked inheritance. Chromosome rearrangements and fertility.  Paternity and relationship testing.  Inbreeding/selfing.

9.  Heritability.  Twin studies.  Environment and chance play big roles.  The 'missing heritability': contributions of gene interactions.  GWAS. 

10.   Epigenetics.  Mitochondrial genes.  Mosaicism.  Fetal DNA in mothers.  Other cool stuff we can now understand.

Seriously beginning to prepare for Useful Genetics

My Useful Genetics MOOC now has more than 8000 students signed up.  It won't be offered until May (6 months form now), but there's so much preparation that I'm already in danger of panic unless I get things under way now.  Last week I meet with the CTLT (Centre for Teaching and Learning Technology) team, and then with the instructional designer specifically working on my course.

Avoiding the 'exercise bike' problem with MOOCs

Over the past few days I've been reading what seems like hundreds of articles and blog posts about MOOCs.  This is mostly because I've discovered a number of sites that aggregate these articles in convenient ways.  I've given up trying to remember everything I read about MOOCs - I'm just letting the flood wash over me and seeing what might stick.

But I want to think a bit more about one article (or is it a blog?): MOOCs and exercise bikes: more in common than you'd think.  Although some writers see the high attrition rate of MOOCs to be evidence of failure, I've been taking more of a toe-in-the-waters view - the barriers to signing up for a MOOC are so low that of course lots of enrollees will subsequently decide not to continue. 

This article suggests a different perspective, that of the well-meaning learner who somehow loses motivation.  Just like with that exercise bike, they feel bad about dropping out, and really wish they could have continued.  Sometimes they will have stopped for a solid reason (bike equivalent - sprained ankle), but for many it was just lack of motivation.  They know that they're missing a lot by not keeping up with the work, but their motivation fades and they're left with another failed attempt at learning.

So how can I build features into Useful Genetics that will help students stick with the course and get the full benefits of the course and the personal reinforcement of being a successful learner?  

One part of the solution is course-specific - building relevance into every week's work.  For Useful Genetics, week 1 is likely to be highly motivating (how people differ), but the next few weeks material may be very dry (gene expression, how heredity works), and I can see a lot of attrition happening here unless I make a special effort to prevent that.

Another other part of the solution is more general.  What features of courses make them easier to stick with to completion? I haven't seen much discussion of this yet.  Maybe this is one of the things that course-analytics can help with.  (If any readers know of studies, please post them in the comments.)

The exercise-bike article mentions the motivational benefits of being part of a group.  I don't think this motivation can come from the discussion forums; there are too many participants.  Face-to-face study groups are great, and I can encourage students to form them, but these won't be an option for most people.  But there might be a way to have people form interest-group-based online study groups, for genetic diseases or dog breeding or political concerns or whatever.  Perhaps, once I see the feedback from the 'Why are you taking this course' part of the initial survey, I can encourage the formation of many small discussion groups focused on the specific motivations students describe.

Doing for math what I want to do for genetics

Keith Devlin is teaching a Coursera course titled Introduction to Mathematical Thinking, and he's blogging about the experience here.

In his latest post he discusses the relationship between what his course aims to teach and what is usually taught in post-secondary mathematics courses.  To paraphrase slightly, he contrasts the formalism of pure mathematics ("chess on steroids") with the role that abstract, pure reasoning plays in dealing with the more messy issues of the real world.  Few students can really appreciate the former, but they all can benefit from the latter, so that's what his course teaches.

This is a lot like what I hope to do with genetics, since I want to replace much of the formalism of Mendelian analysis with reasoning how genetic effects play out in the world our students live in.

He's planning to use calibrated peer review for his final exam.  I'll be very interested to see how this works in Coursera because I want to make extensive use of it in Useful Genetics.  I've used the standard version of CPR in my BIOL 234 genetics course (see here and here), but Coursera describes their version as 'beta' so I don't know how good or solid it is.

Thinking about Peter Sloep's comments

Peter Sloep has some thoughtful comments in his Networked Learning Scoop-it on my MOOC-opalypse post.  Here's his comments in purple, and my responses in black:

The line of argument followed in this essay is a familiar one: MOOCs are there to stay, there are all these apocalyptic predictions about the disappearance of colleges and universities as we know them, hence, to stay employed, I'd better make the student experience worth their money. Apart from the observation that it is a bit ironic that only now that jobs are on the line we start thinking about giving value for money, there are many problems with this kind of argument.

I was joking about being concerned about my job; I'll probably be over-the-hill by the time academia feels the big impacts of MOOCs.

First, Rosie argues from the assumption that MOOC-courses are bound to improve shortly. I am not so sure, I actually think that on the average, when more people jump on the MOOC bandwagon, the quality will go down. Yes, the better courses may be tweaked to offer a better learning experience, for instance by replacing the fora with more intelligent ones that help the learner find sensible stuff amongst the massive number of not-so-useful entries.

If most of the people who jump on the MOOC bandwagon are only doing so because it's trendy, we might see a decrease in average course quality.  I don't think that's likely - I expect most institutions will try to produce good courses, and as more courses are out there, competition will motivate improvement.  But even if the average course quality is no better,  having more courses will mean we have more good courses.

Second, Rosie guesses that flipped classrooms, which provide tutoring next to a (free) MOOC, won't convince the students. That depends, I would say. Read Jonathan Marks' contribution, sitting next to this one. Also, she believes "nobody knows enough about how learning works to do a credible job of this". That simply isn't true. There is a long tradition of research on distance education which explains how to do this online and much research on learning in face-to-face settings other than classrooms and lecture halls which offers valuable insights (see my blog on Katie Vale's presentation, below). It is true, though, that this research has often been ignored by people used to and happy with ordinary lecturing.

I stand by the 'nobody knows enough...' statement. There's a fair bit of research and some valuable insights (including those that motivate flipped classrooms), but not nearly as much as we need.

Third, Rosie then concludes that "[...] one advantage a university gains by offering Coursera courses is that the enormous numbers of students and the online record-keeping make it possible to collect unprecedented amounts of data about student learning. But in practice most of the data will be worthless unless we carefully design our courses as learning experiments.' Under the label of learning analytics such data collection is already taking place and delivering insights. And, yes, it does make sense to carefully design courses as learning experiments. That is precisely what Harvard is doing with its EdX platform (again, see Kathie Vale). I would hope many more colleges start to do so, designing other learning environments than the default lecture hall and learn from the experience.

I couldn't find the Kathie Vale link, nor anything by Googling her.  I read the Wikipedia entry on Learning Analytics, which reinforced my impression that this is primarily a set of tools we can use in our learning experiments.  Learning analytics can be applied to 'found' data (e.g. any Coursera course) but is going to be most valuable in the context of carefully designed experiments.
 
In summary, I don't believe the apocalyptic predictions about MOOCs for one minute. The educational landscape, shaped by learning needs and wants on the one hand and forms and environments for learning on the other, is too vast and rugged to be surveyd to the full by a search party led by commercial MOOC providers alone. However, it is a good thing we start to question the traditional, much trodden roads to learning. If that is what they manage to achieve, we should thank them for that. (peter sloep, @pbsloep)

I don't really think that the rise of MOOCs will lead to the collapse of universities.  Not because universities deserve to be preserved in their present form, but because the whole structure of higher education is so very very conservative that even apocalyptic forces will cause only slow incremental changes.  But I'll save this for another post.
   

Preparing for the MOOC-ocalypse

MO-OCalypse?  MOOC-apocalypse? (Oops, apocalypse is one of those words that, if you look too closely, always appears wrongly spelled.)  

A UBC colleague who's also going to be producing a Coursera course got me thinking about the future of the university.

He starts with two reasonable assumptions:  First, the diversity and quality of Coursera-like courses is going to increase rapidly over the next few years.  Second, universities/faculty members/students are discovering that face-to-face lecturing in large classes is not the best use of student or faculty time and effort, and they will move toward 'flipped' classes where students use class videos and other online resources to learn the course content and then use classroom time for problem solving and interactive learning.

Creating the online resources for a flipped course is a big investment of technical resources and instructor time.  So, for both instructors and administrators, it will make sense to instead use the resources of any appropriate Coursera courses.  Contemplating this for very long leads one to various philosophical considerations, such as "Since Coursera courses are free, why would students pay to go to university?" and then "Yikes, what will become of my job??!!!"

For a university education to be perceived as worth the tuition, it won't be enough to supplement the free Coursera material with scheduled classroom peer-teaching experiences and a tutorial taught by a graduate student.  The university needs to develop integrated programs with hands-on and face-to-face experiences that are seen as worth the cost.

Unfortunately, nobody knows enough about how learning works to do a credible job of this.  So if the university is to avoid selling programs with little demonstrated value, it needs to gather the information that will let it create genuine value.

Ironically, the best way to prepare for this MOOC-opalypse may be to become part of the problem by teaching a MOOC.  In principle, one advantage a university gains by offering Coursera courses or other MOOCs is that the enormous numbers of students and the online record-keeping make it possible to collect unprecedented amounts of data about student learning.  But in practice most of the data will be worthless unless we carefully design our courses as learning experiments.  That sentence makes it sound like designing a course to be a learning experiment is something I know how to do.  It's not.  And I'm not likely to have the time to do this even if I had the expertise.

On the other hand, my course is best-positioned to become an experiment, since it's the least developed of the three UBC Coursera offerings.  UBC has offered Climate Literacy as a fully online Continuing Studies course (non-credit) for several years, and I think Introduction to Systematic Program Design is going to be an online version of CPSC 110.   Although Useful Genetics will build on what I've taught in BIOL 234 - Fundamentals of Genetics, it's basically a new course.  But if we're going to use Useful Genetics as an experiment in online learning we need to start now, because it will be too late once I've developed all the components.

So I'm emailing UBC's Centre for Teaching and Learning Technology (CTLT) to ask if they have a support person assigned to work on course-evaluation development for the Coursera courses.

Later: CTLT responded that this will be discussed at a meeting they're organizing with the Coursera instructors.  I think this means "Not yet, but maybe..."

First week

 

Well, my Useful Genetics course has been up on Coursera for a week, and 3500 people have signed up. Although that's not dazzling by Coursera standards, if signups continue at this rate there will be about 50,000 students by the time I send out the 'Useful Genetics starts soon' email in April.  I expect about half of the enrollees (is that a word?) will then say 'What was I thinking - I've no time for this!'.

Yesterday I met with the instructional technology people to discuss options for recording the many short videos we'll need.  The first decision needs to be between recording in my office or recording at UBC's Telestudios down the hill.  The latter would give better video quality and save me having to learn how to do it myself, but I'd be giving up the ability to make last-minute changes and to easily update videos from one session of the course to another. 

And this morning I had coffee with instructors for UBC's other two Coursera courses, Climate Literacy and Systematic Program Design.  We're coming from very different perspectives and experiences (both academic and pedagogical), so pooling our information and questions will be very valuable.

Useful Genetics!

Next spring I'm going to be teaching a 'massively open online course' (a MOOC) titled Useful Genetics.  You can read all about it here: https://www.coursera.org/course/usefulgenetics.

The motivation for this course arises from the opinion piece I published a few months ago in PLOS Biology - "Why do we have to learn this stuff?" a new genetics for 21st century students.  My goal will be to teach an academically rigorous genetics course that cuts out the no-longer-relevant stuff and emphasizes the parts useful to non-scientists.

It will be sufficiently different from conventional genetics courses that I think I'll have to develop most of it from scratch, including learning how to make videos.  Luckily I'll have the support of UBC's awesome Centre for Teaching and Learning Technology.

The current icon and promo video were made in a rush (UBC didn't get everyone on side until the last minute.  The video is OK for now, but the icon is slick and content-free:

I hope to soon replace it with something like this:

Should we really give Mendel the boot?

My Perspectives article about reforming genetics education is up at PLoS Biology (Why do we have to learn this stuff? A new genetics for 21st century students).  There's lots of chatter on Twitter, but I gather that nobody likes using the commenting system on the PLoS Biology site.

I really want to get feedback on the article, so I'm hoping people will post their reactions and ideas here.

To get things started, I'll reiterate my main point:

The first goal of a modern basic genetics course should be to provide students with an understanding of genetic principles and processes that will be useful in their non-academic lives. 

If we agree on this, then we can discuss how best to accomplish this goal (what will be useful and how should it be taught).  If not, let's discuss what the main goal should be.

Giving Mendel the boot

This is a teaser for my opinion piece on how the teaching of genetics should be changed, which has now been accepted by PLoS Biology.  It should be out soon, so below I'm just going to put the title and the blurb:

"Why do we have to learn this stuff?"  

A new genetics for 21st century students.

Our students will go out into an astonishing new world of engineered genes and personal genomics, so why is the standard genetics syllabus stuck in the 1950s?

A modern genetics problem

This problem was on the final exam of our new Fundamentals of Genetics course.  It's an example of what I'd like our students to be able to do.

(10 points)  The ideogram above shows a normal child’s genome, with her chromosomes coloured by 23andMe to show the results of genotyping her DNA and the DNAs of her maternal grandparents.  Blue segments indicate blocks of alleles shared with her maternal grandmother, and white segments indicate blocks of alleles shared with her maternal grandfather.  Hatched segments could not be analyzed because they have too few SNPs.

a. (1 point)  What genetic process is responsible for these blocks of alleles?

b. (2 points)  When and where did this process occur?

c. (2 points)  What property of the child’s maternal chromosomes 11 and 14 is unexpected?  Why is this property unexpected?

d. (4 points)  Suggest two different kinds of events that could explain this unexpected property.  Give rough estimates of the probabilities of the events you propose.

e. (1 point)  The black triangles above some chromosomes show the locations of SNPs linked to effects on nose shape.  What do these predict about the child’s appearance?

What genetics should all our students learn? ("Stop, we're teaching the wrong stuff!")

Several years ago I was asked to take charge of developing a new second-year 'fundamentals of genetics' course, to replace our program's long-standing third-year course (a legacy from David Suzuki and Tony Griffiths).  So I put together a committee of genetics instructors (profs, sessionals, a TA), and we developed a new set of learning objectives and an ordered list of topics to be covered (a syllabus).  The committee then disbanded , leaving me to implement its work, first as a small pilot class (last winter) and then as a regular course (just finished).

We thought we had been quite radical, because we'd made a very big change in how our course would teach the two big concepts students needed to master - how genotype determines phenotype and how genetic information is inherited.  Traditional genetics courses start with Mendel, and, following in Mendel's footsteps, use analysis of crosses to reveal all the basic concepts of classical genetics; this is Suzuki's 'Genetic Analysis' approach.  Our new syllabus began not with Mendel but with three weeks about how genotype determines phenotype (no crosses yet), followed by two weeks just about how inheritance works (leaving phenotypes out entirely)  Only then would it introduce Mendelian genetics, and then use the standard genetic analysis framework to teach the more complex concepts.

It wasn't until I started to teach the pilot section that I realized we'd been much too conservative.  We'd simply assumed that the goal was to teach students the standard 'classical genetics' concepts.  But what we should have done is first thought long and hard about what students should be learning in a modern 'fundamentals of genetics' course.  That is, what genetics facts and concepts will our students actually use, not just in later courses but in the rest of their lives?

Way back, the answer was that students needed to learn genetic analysis, for two reasons:  First, analysis of how phenotypes are inherited in crosses used to be the most powerful tool for understanding how organisms work.  Even if students weren't going to go on to do this analysis themselves, as biologists they needed to understand how it was done.  And following in the footsteps of the great geneticists was thought to be the best way to learn it.  Second, genetic analysis is hard, and learning to do it trains the mind in rigorous thinking.  Genetics students' experience at solving complex genetic problems was expected to make them better at solving all kinds of problems, in everyday life as well as academia.

Although genetics has changed dramatically, this motivation has largely been left unquestioned.  Although I didn't buy the 'following in the footsteps' part, I accepted the rest.  But the importance of classical genetic analysis to biology is shrinking day by day, displaced by powerful molecular methods.  Worse, improved understanding of students' learning suggests that most genetics students pass their exams using pattern-matching rather than the general problem-solving skills we thought they were developing.

So, what should today's biology students take away from a 'fundamentals of genetics' course?  What will they use in later courses?  What will they use in the rest of their lives?  Are there other concepts that every educated person know about?

So here's a partial list of learning objectives for a modern course in the fundamentals of genetics.  Yes, I know these aren't all phrased as actions students should be able to do, they aren't in a sensible order, the list is incomplete, and the syntax isn't even consistent.  PLEASE give me suggestions for improvement in the comments.

• Students should be able to detect basic errors in news coverage of genetics stories.
• Students should be able to understand why a genetic test or sequencing aids medical diagnosis and treatment.
• They should understand how genetic differences affect health risks.
• Which genetic principles apply to all organisms.
• The extent to which the differences between individuals (humans and other species) are due to differences in their genes.
• How the phenotypes of diploid organisms are affected by interactions between different versions of a single genes, and between different versions of different genes.
• How offspring inherit genetic information from their parents (how meiosis and mating work).
• How genes and genomes change over the generations and over evolutionary time.
• At a simple level, how control of gene expression leads to differentiated phenotypes (a special case of gene interactions).
• They should be able to think about ethical and societal issues arising from genetics.

Twitter in the classroom?

The big 'Fundamentals of Genetics' course starts on Wednesday, and I'm going to try letting students ask questions in class with Twitter.  Of course they'll still also be able to ask their questions the old-fashioned way, by raising their hands, but Twitter has some nice features.

I'll tell students that, if they have a question about what I'm saying, they can post it to Twitter with the hashtag #biol234.  When it's time to pause for questions, I'll display the #biol234 Twitter feed on the screen for everyone to see.  Maybe I'll give us all a minute to read the top questions, and then I'll answer them, integrating answers to different questions where this makes sense.  And then I'll ask for verbal questions.

Students in the class can follow the #biol234 feed on their smartphones and laptops, and can 'retweet' questions that they think important.   Questions that are retweeted will rise to the top of the feed list.  The lecture room has two screens, so I plan to use one for the powerpoint slides from my laptop and a second for internet content from the built-in podium computer.  (This screen will be blanked when I'm don't want students to attend to it.)  One web tab will be the Twitter feed, ideally set so only the top 5 or so questions are visible. 

Other features and concerns?

Students can also use Twitter to answer simple questions posed by other students.

Students who want to contribute will need to have Twitter accounts as well as smartphones or laptops.  This is good - I don't want questions to be posted anonymously, as this can lead to silliness and unpleasantness.

Students won't be disadvantaged by not participating.  If they don't bring laptops or smartphones to class, or just don't want to use them for this, they'll still see the Twitter feed and and my responses.

Won't students who follow the #biol234 feed on their smartphones/laptops be distracted?  Well, they'll be distracted from watching me, but at least they'll be thinking about the material.

One thing I really like about this is that it will help shift the focus from answers to questions.

If this works well I'll need to shorten the presentation parts of my classes, to allow more time for the questions, but this is something I'd want to do anyway.

I don't know anything about Twitter apps, but I suspect that the Twitter web site isn't the best interface for what I want to do.   I'll probably ask the students for suggestions, but I'd appreciate any suggestions from readers.

Teaching evaluations

Results of student survey: no need to have a focus group

I've analyzed the preliminary results of my student survey.  It provides some ideas of ways the course could be improved, but my focus-group experts agree that it doesn't raise any issues deserving focus-group investigation.

What they said:

Agree/disagree (~Likert scale):

1. I had the necessary background for the course.  Most agreed
2. The readings and reading quizzes prepared me for the lectures. Neutral
3. The iClicker questions were not challenging enough. Most disagreed
4. The Genetics in the News slides took too much time away from course material. Neutral
5. The homework increased my comprehension of the lecture material. Most weakly agreed
6. The tutorials helped me learn to solve genetics problems. Most agreed
7. Having two mini-midterms and a midterm was too much testing. Most disagreed
8. The course grade was based on too many different components. Most disagreed
9. The workload was much higher than for other courses. Neutral
10. I feel prepared to deal with genetics issues that may arise in my life. Most agreed

 Written Answer Questions:

1. Should any topics be cut from the course material?  Most said no.
2. Were any topics missing from the course that you wish had been covered? Most said no.
3. A pizza-lunch focus group will be held later this month; all students are welcome to attend. Please mention below any specific issues that should be raised then.  Below is what they said:

• Workload, discrepancy between difficulty of lecture material and what was tested (in tutorials, reading quizzes, midterms, etc.)
• The tempo of the class. The first half seems like a review, and all the new stuff are in the second part. 
• Methods of assessing learning in this course. 
• No specific issues. 
• How much we liked the format of the lectures -the methods in which we tried to prepare for exams
• Mini-Midterm format. I think that the midterm was a fair examination however, the second mini midterm had a multiple choice question that had about 9 choices and was worth about 6 marks. I felt I did good on the rest of the exam but still didn't get a great mark because of 1 MC question.
• How to study for the final. Every test has been a different format, what to expect. 
• I would like to suggest ways to make homework more helpful in preparing us for the exams. Also, maybe investment
• into custom booklets with some notes and problems sets like Bio 201. 
• They are too little guidance in this course
• Tested materials --> what to expect in midterms/exams weren't very clear 
• Overall structure of how the course will be run next year. Textbook assignment and readings. Better formatting for the meiosis/mitosis content from the beginning of the year - personally I am still fuzzy, even though the concepts were stressed to be very important. 
• I think going over online homework and reading quiz questions in class would help. Or perhaps explanations for the answers could be posted online because there are still questions that I don't understand. I also think the amount of work this course requires should be re-evaluated. The amount of reading is quite heavy and having two quizzes (homework and reading) PLUS peerwise PLUS tutorial each week is a lot.
• How this course and its changes (234 vs. 334) related to other courses, such as Biol 335. 
• How to study genetics
• I think that the easiness of this course should be covered. I felt that this course reviewed a lot of material and didn't cover that much new material.

Ranking the course components:

Many components of this course contribute to the final grade. Please try to rank them according to how valuable you found them, taking into account your learning gains and the amount of time you invested in them. For example, an activity that took a lot of your time but resulted in little learning would score low.

• Tutorials  High
• Peerwise questions  Low
• SNP report  Low
• Calibrated Peer Review  Low
• Online homework  High
• Reading quizzes  No consensus
• Studying for midterms  No consensus
• Attending lectures  High

Only 21 of the 38 students have completed the survey so far.  That's certainly enough to go on, but I'll reanalyze the responses after the final exam marks have been posted (that's the last time the students will give any thought to the course).