This year I have passed a small milestone with my teaching: after running my first advanced solids tutorial, I have now taught every topic in the non-options inorganic courses at Oxford. I thought it would be valuable to reflect on what it means to teach such a large volume of material, and on how it’s changed me: what does it mean to teach such a broad coverage to the level of depth required?

Description of coverage

Inorganic Chemistry tutorials are small groups of perhaps 2-4 students, in which problems done before the tutorial are discussed. Tutors teach every topic in the syllabus (regardless of their personal expertise), though swapping advanced tutorials between colleges is quite common (“I can do your organometallics if you can do my solid state”). It is perhaps easiest to list the tutorials and classes I run to show how the volume of material I’m describing is not an exaggeration.

First Year

• Models of Electrons and Zeff

• Trends in the Periodic Table

• Ionic Structures

• Ionic Energetics

• Acid-Base Solution Equilibria and VSEPR

• Electrochemical Solution Equilibria

• Molecular Orbital Theory

• Chemistry of the D-block

• Chemistry of the Main Group

Second Year

• Group Theory 

• Bonding in Molecules

• Coordination Chemistry

• Organometallics

• Inorganic NMR

• Chemistry of the F-block

• Diffraction

• Electronic Properties of Solids

• Non-Metal Chemistry

Third Year

• Compounds of Interest

• Essays on the Chemistry of the Elements

• Magnetism and Electronic Spectroscopy

• Reaction Mechanisms

• Modern Main Group Chemistry

• Organometallic Mechanisms and Catalysis

• Bioinorganic Chemistry

• Solid State Chemistry

• Essays on Concepts and Phenomena

• Synoptic Data

With the exception of the third-year solids course, I taught all of these topics from the first year. It’s a lot of coverage!

Description of depth

The depth of coverage is at the discretion of the tutor. I currently try to align depth with (i) difficult features in the topic; and (ii) the tenor of exam questions (which tutors don’t set).

For example, I set a question about the mathematics of the structure factor for an FCC lattice because I want students to understand that the permitted {hkl} conditions (“all odd or all even” etc) aren’t magic. This question gets at a particular piece of the fundamental science, but is more-or-less useless for exams. On the other hand, I make sure to do “route map” questions in the first year main group work because they are the most common format of assessment. This is (in my view) scientifically inauthentic, but prepares students for the exam.

It’s a hard balance, and I’m not sure I always get it right. Some of my tutorials still feel too dry to me, perhaps because I err too much on the side of assessment preparation. This bias towards assessment sits in some tension with the University’s description of the pedagogy, which makes no mention of aligning tutorials with exams:

The purpose of a tutorial is: to develop an individual student’s capacity to think in depth about a subject area, and to operate with growing confidence within its techniques and methodologies, with the expectation that the process will promote increased understanding of the subject for both tutor and student. 

Tutorials are not intended to be a mechanism for delivering subject knowledge. It is expected that students will develop this through departmental teaching and guided independent study. Tutorials provide the opportunity to explore issues, to develop critical thinking, and to present ideas and analysis both orally and in writing. 

I’ve found the balance hardest to strike when the assessment forces me to pick between learning and performing as outcomes. For example: bioinorganic chemistry is a topic I enjoy teaching, but it is assessed predominantly through essay questions in the context of a tight-for-time problem solving exam. I give the tutorial and we discuss things like how the destabilisation of the metal centre in the entatic state is the heart of the kinetic elegance of enzyme chemistry, then we talk about how it’s a topic which will make time management in the exam very difficult. I am trying both to cover the material and also discourage students from using it.

Have I been a good tutor there or a bad one? I don’t know. I’ve constructed a chance to think in depth about a topic, but I’ve also forced overworked students to spend time on material they probably shouldn’t use in the exam.

Description of my workload

My workload for this kind of teaching was profoundly rough at first. In my first year here, I was working intensively for >70h/week in term time. I kept a workload diary, and am confident of the number. I think it would be fair to describe the over-work as mostly relating to the course’s depth rather than its breadth: while some of the time was the high teaching contact hours of my role (perhaps around 20h/week), most of it was trying to engage with very specific questions outside my immediate expertise. If you set students a question on the magnetic moment of Eu(III) salts, you had better know what the answer is. This expressed itself as a lot of time spent on marking and reading, as well as a lot of thinking (often after a tutorial: “yeah, why is the bending of RSnSnR favourable if the bond order becomes lower?“). The reflective task of improving my own teaching was very narrowly tied to the content at first, and it’s only recently that my reflection has broadened into thinking about non-content issues like student affect.

The combination of hours and intensity was very grim, and honestly I don’t think I could ever do it again. The tiredness under such workloads is crushing, and my first year of teaching here was not my finest work (not even close). It’s hard to keep engaged with work you aren’t proud of, and this under-performance made the overall experience very negative in a professional sense as well as a personal one; I seriously considered quitting before the pandemic hit at the end of my first year. I was also very disagreeable! I hated how grumpy I was when I felt bad at my job. This undoubtedly had an impact on the way my relationships with colleagues developed, and I suspect I am still perceived as quite difficult or prickly. I was overwhelmed despite being relatively well-prepared for the coverage by my academic trajectory: this is the degree I read as a student, and I taught every inorganic supervision topic as a PhD student at Cambridge before engaging deeply with the educational literature about how to teach chemistry effectively while I was a T&S lecturer at Hull. 2019-20 was still awful, despite these advantages.

I’m in my fourth year now, and the content-related workload has become much more manageable. Some of that is familiarity with the topics (I really know what a Mott-Hubbard insulator is now), some of it is familiarity with the specific tutorial questions I set (I know how to discuss the NiO and EuI2 Mott-Hubbard insulator systems in Question 4 now), and some of it is familiarity with the particular questions students often ask (I have developed ways to discuss the work function in the context of the band structure now).

In the teaching context of workloads, being alert and engaged is the scarce resource (both for me and for students). In general, tactics which minimise in-term workload are probably the scholarly thing I am thinking about the most in terms of my personal teaching practice. Removing redundant tutorial questions from my problem sets, or changing “legwork” questions into interpretation questions about data I provide. Running a class on Group Theory in week 1 of second year is possible if I make my students read Vincent in the summer between first and second year. Making a physical booklet of the problems saves me lots of time on in-term emails. Workload-neutral decisions like sequencing easier topics near the end of term (when everyone is tired) lets students engage more deeply with the hard topics, but this isn’t always possible to reconcile with the sequencing of lectures. Students will often have a few tutorials every year before starting the lecture course on those topics - they must learn completely independently, and tutorials often involve more teaching in these cases. Similarly, weak delivery of some lecture courses means that the tutorial has to be used to teach fundamental material rather than discuss it. This alters the types of question and discussion which advance student learning in certain topics.

Two mistakes I made

In my first year, I was (generously!) given copies of the tutorial problems which several colleagues used. My tutorials are their buy-out teaching, and I decided to run the tutorials that students would have had if their tutor wasn’t doing funded work (i.e. a different set of problems for each college).

I really regretted doing this. The specific questions in each tutorial sheet had their own quirks and invited different student questions. For example, one set of diffraction work had questions about the fundmanental physics of the structure factor, and the discussion was very mathematical. Another had a tricky sequence of chemical questions about observed vs true symmetry in the PXRD data for isoelectronic ion pairs in substances like CsI, MgO, CaTiO3.

On the one hand, this meant more work for me in an already-heavy year. On the other hand, it forced me to engage with the tutorial decisions of several experienced tutors; in the long term, seeing these perspectives has enriched my understanding of what kinds of tutorial questions are successful. I’m a better tutor for having survived, but it seems likely that I could have reached the same outcome through less-brutal ways like mentorship. Mentorship would also have helped me understand how many contact hours were expected of me, something which would have been good to know much earlier.

The second big mistake I made was in my second year, when I tried to align my coverage with the published course syllabus. I thought the syllabus was the authoritative document about the course coverage, but I now believe I was wrong about this. There are exam questions which don’t match any knowledge or skills on the syllabus, and there are objects on the syllabus which don’t ever make it to the exam papers. I now align my teaching with past papers as I have no say over the assessment, and I re-wrote my teaching materials accordingly the following year.

In broad terms, the examiners (drawn from the permanent academic staff in the central University’s Chemistry Department) seem free to set questions on whatever they like. I once heard someone defend this approach to assessment as “constructive unalignment”, but I confess that it seems more like destructive alignment to me. If there is no agreement on what material we’re teaching, how can students be assessed fairly? What kind of student behaviours are encouraged when the examinable material is not demarcated? Perhaps these questions are beyond the remit of a mere tutor, but they have answers which I need to engage with. The more poorly-defined the examinable content, the more over-prepared I must aim to make my students.

My development as a scientist and an educator

I’ve become much better at the dynamics of a tutorial, but I struggle to explain exactly what this means. A strong personal grasp of the material is necessary but not sufficient to run a tutorial, yet most of the generic tutorial skills are more social than disciplinary. The generic skills are always used in disciplinary contexts, though. To take a small example: I have learned to write upside-down so students opposite me can read as I write. This is a generic skill in some (incredibly-)narrow sense, but it feels very disciplinary to me when I write an eta or a mu.

The social dimensions are mostly about holding a scientific conversation, and the responsiveness of the pedagogy is its central feature. There is something about establishing rapport, something about encouraging quiet students and directing confident ones. Perhaps there is something about responding positively to wrong answers, and about explaining the same idea in meaningfully-different ways if a student keeps asking for clarification (this is hardest for the fundamental concepts - first-year students always ask the hardest questions!). Finding ways of asking productive technical questions without “giving away” the answers is important. Getting a feel for the amount of time available and the balance of who is speaking (and who they're speaking to) is also part of it. Being approachable and open is somewhere in the mix as well, especially as students asking questions is the central mechanism of the tutorial’s responsiveness. The pandemic made the social dynamics very different; I found online tutorials just awful to teach. I wonder if some of this was disciplinary - I’ve heard people in essay-driven contexts say they rather liked online tutorials.

There is also something more holistic about tutorials which feel successful. The times I feel like the tutorial has worked involve developing students’ piecemeal knowledge and skills into a coherent understanding. Sometimes this demands an alertness to the difference between getting a question right and understanding it (the disproportionation condition on a Frost diagram [“sits above the line joining two other species”] is easy to apply but hard to understand, for example). In general, seeing specific chemical systems as examples of general chemical themes is the perspective I’ve found most successful for satisfying tutorial teaching in Inorganic Chemistry. It’s hard to see a general way of applying that principle, but four years of reflective practice is something which helps to provide context-driven opportunities to make it work.

Scientifically, I’m not sure I’m a better chemist for this teaching but I have become an extraordinarily good chemistry graduate. I know a bit more content within all of the topics, but my main development has been to develop better ways of explaining things I already understood. This often speaks to the synoptic perspective afforded to tutors; seeing links between ideas is easier if you’re teaching first years on Monday and third years on Tuesday. To ask students bluntly what they thought of a tutorial or a topic also helps me to reflect more deeply on what’s working and what isn’t.

Reflective tutorial teaching is something which I have found some conceptions of Schulman’s “Pedagogical Content Knowledge” to describe fairly well. Responding to both the structures of knowledge within the discipline and also the pedagogic practices I judge to be effective creates a very particular kind of professional knowledge. This knowledge might be expressed as things like the specific picture I draw to explain the difference between Fischer and Schrock carbenes, the language I use to avoid talking about plusses and minusses in electrochemistry, or the series of compounds I use as examples to explain the most stable VSEPR isomers in trigonal bipyramidal molecules. Part of this personal growth has been a much deeper engagement with what kinds of in-tutorial discussion are provoked by different questions (both in the problem sheets and in the session).

My students’ learning

I’m not sure any tutor will ever get a fully candid view of how students find their tutorials, but I think the tutorial system has a few general merits for students when it works well.

1. Someone is making students do work, forcing them to grapple with concepts they do not (yet) understand.

2. Staff and students develop a professional academic relationship which persists for longer than a lecture course.

3. Staff are deeply aware of the course’s trajectory, and can become hyper-aware of things like prior learning and the later uses of knowledge (“organometallics is worth getting good at in second year because there are several courses on it in third year”). 

4. There is also something quite hard-to-define about what it means to hold a professional scientific conversation which I think tutorials can develop in students (though as term progresses I can feel myself becoming more didactic as I become more tired).

5. Tutors have a relationship to the course which parallels students’ experience of it: students need to study the whole course, and tutors need to teach the whole (inorganic) course. I think this helps get everyone on the same page somehow, and the way that tutors aren’t examiners means that beating the exam is a unifying goal (perhaps reminiscent of school teaching when exams are set nationally): tutors and students are “on the same side”.

On the other hand, the autonomy of tutors has significant risks for students. Your relationship with your tutor is absolutely critical for your success at the University, but is subject to all the personal dynamics of any professional relationship. If you have a bad tutor (whatever that means), you may have them for your whole course.

My Students’ Workload

A specific consequence of tutor autonomy is the consideration of student workload. I can set my students as much or as little work as I wish. If the departmental teaching load is high (and in Chemistry it is), then Oxford’s “signature pedagogy” arguably comes at the expense of students’ free time (though you could argue instead that the culprit is lectures or labs or maths classes). Student workload isn’t a hypothetical worry, having been discussed both in recent QAA analysis of the University and also by several participants in a recent MChem project on student mental health.

To see my students succeed in the University’s exams I believe they probably need to do nearer 60h/week of work than the University’s stated expectation of 40h/week. However radically I winnow my tutorial questions, however cunningly I assign vacation work, the outcomes expectations embodied in the exams are hard to reconcile with reasonable student workloads. If these numbers are right (and I’m sure some would say that they aren’t), then tutors are forced to make a decision: do I let my students have a life, or do I let them succeed in exams?

Surprises

The single thing which has surprised me most about tutorial teaching is how little scrutiny I experience as a tutor. I try my best, but having taught now in 5-or-so different colleges I don’t think anyone is checking up on me. I could be a truly terrible tutor and I don’t think any non-student would know. As the tutorial pedagogy is the distinctive feature of the University, this seems significant.

Relatedly, it is interesting to see how the University’s signature pedagogy is so poorly-incorporated into career structures. It’s not clear to me how the hypothetical best tutor in the world could make a career - or even a permanent job - out of being an Oxford tutor. When the work of a tutor is both so specialised and so hard to master, this seems odd.

Which I think brings me to the most disappointing surprise of my time here so far: I am professionally lonely. Perhaps it was the timing of the pandemic, or the workload-driven isolation of that first year, or the way my role is based in a department rather than a college, but I have very little interaction with other academics and I don’t regularly talk about my tutorial teaching with anyone. It’s very hard for me to understand exactly why this is. The proportion of teaching-only contracts is very low in the HESA data, and perhaps my odds of encountering someone interested in scholarly tutorial teaching are correspondingly poor on a simple statistical basis. I sort of wish I didn’t already have my FHEA, as getting it here might have been a good way to meet other educators.

Conclusions

So lots of my experience has been very ambiguous! I’ve had an immersion in the discipline which has dramatically improved my fluency in teaching inorganic chemistry, but that immersion felt a lot like drowning. I’ve been granted tremendous autonomy, but I feel like I need to use that autonomy to violate the University’s student workload policies. I’ve been given extraordinary opportunities to reflect on my teaching, but no communities of practice to reflect within.

For all that, I take real joy in tutorials. It is fun to talk about problems with students, and to find fresh perspectives on fundamental ideas. It is stimulating to be asked questions you’d never considered. It’s completely thrilling to see students grow into the potential you saw in them at interview.

So I’m not sure I have a neat conclusion about how teaching so broadly and to such depth has affected me. It has certainly changed me. There are technical ways I’m better at tutorials than I used to be, and I am increasingly aware of the political ways that my decisions will necessarily centre some types of student and marginalise others. I now have perspectives on the curriculum which have challenged elements of my old thinking, and a deeper view of the curriculum about which some of my old thinking has substantial criticisms. In disciplinary terms, I think my knowledge of inorganic chemistry is much more richly interconnected and I think my appreciation of what concepts are most significant has become much more focused.

The most urgent tensions which remain unresolved for me are the student workload and the poor definition of examinable material. These issues are substantially intertwined, but in my view both of them seem to lie within the authority of the central University (rather than the colleges). The ongoing internal review of the Chemistry curriculum has great scope to address workload and assessment, and I hope addressing these problems is a central part of renewing the department’s ambitions for its teaching.