Tuesday, September 28, 2010
Thursday, September 23, 2010
Monday, September 20, 2010
Here are the four podcasts I listen to most often:
Focus 580. This is better than any of the national interview and talk show formats on NPR (I can't stand to listen to those, because the public call-in questions tend to be so annoying). The calls they take on Focus can also be annoying, but they sometimes aren't, and the hosts, David Inge and Celeste Quinn, are extremely good.
Le Show with Harry Shearer.
Planet Money, on NPR.
State Week in Review on WUIS with iTunes-U.
I also listen to these sometimes:
This American Life.
Brookings Audio Events.
Swarthmore College Featured Events.
The UIS ECCE Speakers Series.
Podcasts from the University of Oxford.
Since I have a 40-45 minute commute each way on days when I ride my bike to campus and back home, I can usually get in a couple episodes each day, and then I often go to bed at 11:00 or midnight and listen to a podcast as I fall asleep. When I'm doing work that doesn't require much language thought (like editing photographs or making a photo album) I'll also listen to podcasts when I'm not listening to music.
Thursday, September 16, 2010
Recently some faculty and administrators in my college were considering what sort of expectations we have for clinical instructors, in terms of their teaching, scholarship, and service. This inspired me to consider how much hours it takes to teach well, and what time this leaves us for committee meetings, service to the college and department and community, advising, research, keeping up-to-date in our fields of expertise, and so forth. I’d like to share my calculations here.
First of all, let’s model an ideal situation, where faculty work a reasonable number of hours per week. I think 44 hours per week is a reasonable workload, and in a 44 hour work week for university faculty, I would suppose about 8 hours of labor are actually spent at lunch, breaks, walking around campus to get to meetings, and various other activities that don’t really contribute to teaching, service, or scholarship. So, a typical faculty in this ideal state would about 36 hours of productive working time per week in a semester in which to do what they are paid to do.
And how long is a semester? Well, faculty probably work on their courses and administrative duties a couple weeks before classes start, and then classes go on for about 15 weeks, and then there is the week of final exams, and the week after final exams when faculty are grading those exams or final papers and so forth, so that’s 19 weeks of work. Usually there is a week of vacation in there somewhere, either spring break or the Thanksgiving break, but in this ideal scenario we’re just ignoring that in the hopes that in this ideal world the faculty get a week of vacation in there. That gives a typical faculty member about 684 hours to labor in a typical semester, at least in this model in which they are working a reasonable work week.
During the 15 weeks of courses, faculty must show up for classes, and that usually involves 2-4 hours of time per class each week (for simplicity’s sake, we’ll say 3 hours). Each week a faculty member teaching three courses must be present with the classes of students for about 9 hours. This time is the same whether a class has four students or 84 students.
Then there is a flexible amount of time a faculty member can devote to preparing for each class session or preparing for their class before the semester begins. They can design and prepare exercises, simulations, lab assignments, lecture notes, and so forth. They can update everything. They can make new tests, new discussion questions, and choose new readings. They must design a syllabus, and they must prepare assignment descriptions or tests. There is a minimum threshold to all this work, and there is probably a maximum number beyond which additional hours of labor get someone no discernible improvements in the quality of instruction.
Another significant variable shaping the need for this sort of labor is the experience of the faculty in teaching the course. If it is the faculty’s first time teaching the course, and the faculty is new, and the course is in an area the faculty knows about, but it’s not the faculty’s specialty, then preparation time can be immense, easily 10 hours per week just to do a reasonable job. If the faculty has taught the course a few times before, and the course is directly related to the faculty’s special interests and expertise, then class preparation times are drastically reduced, and one can probably do an excellent job with just 2-3 hours of preparation per week of classes. Typically there is more time devoted to this class preparation in the two weeks preceding the start of classes, and then a fairly constant amount of time throughout the weeks of classes, and then almost no time devoted to it in the weeks following classes, so all in all, the average over the 19 weeks of work in a semester is probably close to the average spent in a typical week. This is a number of labor hours that is also fairly constant and independent of class sizes. In fact, preparing lectures and group exercises and discussion questions for a class of 55 may require less time than the more detailed preparation one might need to do for a more intimate and personal class of 15.
Then there is the time devoted to evaluating student work and giving students individual attention, mainly through writing comments on their papers or lab work, but also in grading tests, answering questions or e-mails related to class during office hours, and so forth. On one extreme, a faculty member can give almost no feedback and use only multiple-choice tests. On the other extreme, a faculty can give copious feedback on writing style, reasoning, critical thinking, sources, grammar, and so forth. The minimum requirement that leads to almost no feedback aside from submission of grades at the end of the semester probably averages out over 19 weeks to about 5 minutes per student in class per week (slightly over 90 minutes per student per semester). If one gives five assignments and a few tests and gives copious helpful feedback on each assignment, assigns a great amount of writing and actually carefully reads all the writing and comments on a fair portion of it, then one could easily devote 12 hours per student per class per semester, averaging nearly an hour per week per student during the 15 weeks of classes (although not spread evenly across the 15 weeks of classes, but rather concentrated in weeks following submission of assignments).
So, here are my hour estimates for the time required to teach three courses with 18 students in each course over a semester, at a minimal level, a high level, and the highest level that still makes sense, using 1.5, 8, and 13 hours per student per semester as the times faculty would devote to student feedback; 2 hours, 5 hours, and 10 hours as the time faculty would devote to class preparation on average per week; and 3 hours per week for actual in-class time for each course during the 15 weeks of the semester for all three scenarios:
Time in a semester per course:
Average reasonable time for an experienced faculty member giving high quality feedback: 280.
Time for a newer faculty teaching some new courses and giving outstanding help to students in the class: 460 hours.
Time for an experienced faculty member doing a minimal effort with minimal feedback to students: 110 hours.
In actual practice, I think something like 200-250 hours per course is the actual labor time for teaching at a level of excellence with some degree of efficiency after you’ve had two or three years of experience as a professor. However, most faculty teach at a level of high quality (very good work) rather than a level of outstanding excellence, so most faculty get by with some sort of effort around 170-200 hours in a semester per class they teach, assuming the class sizes are around 18-20 students.
What is the amount of time this leaves faculty for their research, their scholarly work, their attending meetings of their department or college, their service to various governance committees and other sorts of clubs or committees on campus? What about their time meeting with their advisees, or the time they spend talking to prospective students or alumni? What if they must travel for a three-day trip to a conference where they present a paper and attend many hours of sessions? What if they are volunteering in the community or in a scholarly association, and helping to serve on an advisory board, or organize a conference, or do accreditation site visits? What about the memos they must write, the schedules and book requests they must submit? What about service on search committees, hosting potential colleagues to campus visits or reviewing scores of files of job applicants? What about training sessions? What about the 40-60 e-mails they receive each day? What about showing support for students by attending student events? Anyway, if a faculty member is spending 200 hours per course per semester, this leaves about 84 hours per semester for all the research and service. If a faculty member is spending 170 hours per course, they will have about the same number of hours per semester (174 hours) to do all their scholarly and service and administrative work.
Now, how many hours are faculty expected to devote to service, and how many hours are required to produce high quality scholarship? Assuming a professor puts 510 hours into teaching three courses with 54 students across the three courses over the semester, and then divides their time with 90 hours in scholarship and 84 hours in service, that gives them about 4 hours and 45 minutes each week for scholarship and 4 hours and 25 minutes each week for service and administration.
1). If a faculty is expected to teach 4 courses per semester rather than 3 (as clinical instructors may be asked to do), there will be no significant time left for scholarship or service. Faculty teaching 4 courses per semester should have no expectations for service or scholarship.
2). If class sizes increase from 18 to 25, average hours devoted to feedback to each student will decrease (assuming hours allocated to scholarship and service remain constant). If a faculty member devotes 3-4 hours to individualized attention and feedback for each student (in reading and critiquing assignments and so forth) over the course of a whole semester, can this give the student an equally high quality education as when faculty devote 6-8 hours per semester to each student? We are assuming here that class preparation hours from the faculty and actual class time are held constant, and we’re just reducing individualized instruction and feedback. Is that individualized feedback and evaluation so important that a reduction of two or three hours per student in the time faculty give to individualized attention (over the course of a whole semester) going to harm students? I think it will.
3) Taxpayers who help fund universities should insist that public universities organize their spending and administrative structures in such a way that average class sizes remain at or below 20 students per class. A change in average class sizes from the 15-20 range up to the 20-25 range probably diminishes the quality of education in colleges that are not highly selective (students at highly selective schools probably need less individualized time from their instructors). Certainly a rise from the 15-20 class size range to the 25-30 students per class will greatly diminish individualized feedback from faculty. Faculty will need to compensate by training students to give peer feedback and using peer feedback to supplement the greatly reduced faculty feedback.
But still, faculty and the taxpaying public should fight to keep class sizes smaller. Universities ought to spend more than 50% of all revenue gained from tuition payments and allocations from unrestricted state funding to the costs of direct student advising and teaching (and library services). Efficiency gains in the structure of universities and allocation of money toward direct provision of services should be a high priority, and take precedence over attempts to squeeze greater “teaching efficiency” from faculty. Teaching 25 students rather that 15 students per class is not necessarily more efficient in terms of actual learning gained by the students in a classroom. That is, economic efficiency is not the same as pedagogical efficiency. Given the constraints of reasonable workloads and working hours and the need for quality scholarship and service, and given the actual numbers of hours needed for class preparation and individualized feedback, maximum efficiencies in teaching are probably obtained in class sizes of 15-20. Average class sizes of 20 are sustainable if public subsidy of teaching costs are set at close to 50% and average tuition costs actually paid for a year of public university education are set at about 20% of median year-round full-time income and faculty earn incomes at close to the median year-round full-time income for college-educated males, provided that over half of tuition and unrestricted state subsidies go to direct instructional and advising and library costs, and administrative and support costs are kept at less than 50%.
3). If quality scholarship requires more than 4-5 hours of work per week over a semester, and faculty are expected to have a reasonable work week of about 44 hours, then faculty will need to reduce time devoted to courses or to service and administration. Universities should be careful not to lay administrative burdens on faculty that would likely exceed an average of 4-5 hours per week. If university governance requires that some faculty give significantly more than 5 hours per week to departmental service or university service, then a system should be created to allow those faculty to teach fewer courses or smaller courses.
Monday, September 06, 2010
When a certain type of scientist makes this claim, their position always runs along these lines: science can explain whatever phenomena a religious person uses God to explain, and since science can explain it without requiring God or supernatural powers, it follows that God and supernatural powers don’t exist. Boiled down to its essence, the argument is that if we use parsimony (Occam’s Razor) we won’t have any reason to believe in God, at least not any reason based on our desire to explain or describe what we perceive in the universe.
The Hawking and Mlodinow argument follows this line. They point out that when people didn’t know much about science, they made up stories about gods or supernatural forces to explain their experiences. Even now, some people suppose there must be a God because there is something rather than nothing. We exist in a universe that seems to have just the right settings to allow life to evolve, and our planet seems remarkably well-suited to human evolution. This seems like a remarkable coincidence, so perhaps God designed the universe to be just right for us. Yet, science (as presented in the Hawking and Mlodinow book) has shown that there can be many universes, and universes can emerge from nothing, and if there are a nearly infinite number of universes it follows that there must be some universes such as ours where life as we know it can evolve. And since life has evolved here, this just happens to be a universe where it’s possible for life to emerge and evolve. So, we don’t need God to explain this universe, its creation, or its remarkable suitability for life.
The sort of God this argument dismisses is one who would fashion a creation in such a way that there would be gaps, cracks, holes, and incongruities within the creation so that intelligences within the creation would be able to logically discern the existence of the Creator, and understand that the Creator had intervened within the creation. It also dismisses a God who would set up a universe with two conflicting systems, one a system of physical laws and forces working according to a particular set of rules, and another a system of supernatural forces working according to an entirely different set of rules. In other words, the "God of the gaps" idea that we believe in God because only God can explain certain observed phenomena (like the existence of the universe) is refuted by science which has eliminated all gaps. This is an argument against a God who inspires belief because people need a "god hypothesis" to fill in explanations for things that aren't understood.
But really, there is no point in entering into a debate with such assertions. Occam’s Razor (the rule of parsimony) will always rule out God. Any evidence of God (defined as an all-powerful, all-knowing, Source of creation and Sustainer of reality with attributes analogous to what we experience as will and consciousness and emotion) can be explained parsimoniously as evidence for more advanced civilizations, rather than God. That is, in a universe such as ours it is quite plausible that there could be entities who belong to races or civilizations that evolved and advanced past our current technological state hundreds of thousands of years ago, or even millions or tens of millions of years ago. There could be beings in this universe that passed our present state of evolution and knowledge over a billion years ago. Such entities and their civilizations might be able to manifest all sorts of powers and technologies which would look to us like the work of an all-powerful, all-knowing Creator God. But, if we apply the rule of parsimony, we must stick with the hypothesis that all indications of this “God” are in fact the manifestations of natural and scientific technologies far beyond our current understanding. Advanced alien races and technologies, rather than supernatural gods or magical powers of a deity, can always be suggested as an alternative scientific explanation for what appears to be evidence of God. Such explanations, which conform to Occam’s razor, give us questions about any god that science eventually finds. For example, from what race of beings did this god evolve, and how did the god develop powers and attributes, and by what technologies or physical laws does this god manifest seemingly miraculous powers and attributes? This is quite different from the typical religious contention that God was uncreated, and God is above questions related to causes, technologies, and so forth.
Some people come to belief in God through the teachings of Messengers of God. Unlike many other branches of human religion that follow Messengers of God whose lives and revelations are considerably obscured by historical distance and lack of solid evidence, Baha’is follow the Revelation of a Manifestation who lived from 1817-1892. There are plenty of obscuring legends and exaggerations around the life of Baha’u’llah, but we have enough physical and historic evidence, as well as original source materials, to make some factual observations about His life and message.
Yet, even these, which I take as convincing proof and a basis for faith, could never overcome a strong attachment to the rule of parsimony. Baha’u’llah revealed scriptures of high excellence and wisdom, sometimes spontaneously and rapidly. This seems like a possession by the Holy Spirit or a Revelation from God to those who have faith. But, rather than resort of supernatural explanations, one could use the rule of parsimony to point out that many people have revealed poems, stories, and other miraculous material while in trances, or under hypnosis. There have been cases of automatic writing (spirit writing) and so forth, all related to dissociative mental activity, and this isn’t so different from the way Baha’u’llah revealed Baha’i scriptures. Baha’u’llah evidently saw and heard things that others couldn’t hear, and this sort of phenomena is fairly common in people who suffer from schizophrenia, or persons who aren’t mentally ill, but who experience objectively non-real experiences such as alien or fairy abduction or conversations with dead ancestors (at least non-real to mortal observers, but I have no idea about what is going on in the brains of such people). Delusions are also widely known, and Baha’u’llah’s beliefs about His station could be explained as delusions without invoking a supernatural God to explain them. The fact that Baha’u’llah seems to have fit in with prophetic predictions about a Manifestation of God could be dismissed, if we use parsimony, as coincidence, or as stretching the facts to fit vague prophetic traditions. A person could just argue that Baha’u’llah didn’t in fact fulfill prophesies. Likewise the remarkable or miraculous events surrounding Bah’u’llah’s life or his remarkable abilities could be dismissed as untrue stories, or merely remarkable coincidences or luck. It is always more scientific to invoke Occam’s razor and explain away aspects of Baha’u’llah’s life as coincidence, false stories, or something else that doesn’t require a divine supernatural God.
Even if one did accept that something seemingly supernatural was going on in Baha’u’llah’s life and Revelation, it would be more parsimonious to suggest Baha’u’llah was merely using natural forces that science hasn’t yet detected and explored, rather than to accept that God was involved. In fact, in some of Baha’u’llah’s writings (such as the Tablet of Wisdom) it appears that Baha’u’llah was picking up some ideas current in his time and culture about history, rather than picking up direct factual information about history. This fits with an idea that natural technologies or powers were involved, rather than an all-powerful, all-knowing Divine Force.
Baha’i writings about miracles also seem to take this view, that miracles and miraculous events should not be relied upon as evidence, unless one has directly experienced the miracle or event for oneself. The supernatural stories surrounding Baha’u’llah are evidence to those who were present when the miraculous events took place, but for the rest of us, such stories are merely stories, and we should not take them as evidence for Baha’u’llah‘s status as a Messenger of God. After all, there are many people who have sometimes exhibited miraculous powers or encountered supernatural forces, and this sort of experience isn’t really all that unusual.
What frustrates me is that atheistic scientists who present these arguments that Occam’s razor and the rule of parsimony remove God from the question because science can explain everything without God are claiming that the domains of logic and empirical evidence (the sort of evidence that can be reproduced and manipulated through experiment) are sufficient domains to explain the world. In essence, they are claiming that science and logic and rational thought are sufficient for our beliefs and models of the world. This is, I believe, a mistake. There are other domains of life and other models of understanding that work quite well, and are needed, and some of these domains are outside of science. Religion is one of these domains. I think when religion makes propositions about the natural world, it is the duty of the religious person to subject their assertions to scientific investigation. But the scientific method is not an especially useful way to investigate the realms of meaning, value, ethics, theology, and so forth. There are other realms such as arts, economics, human relationships, dreams, and various other aspects of the world that may also lie mostly outside of the scientific method. That is, we use models of understanding and meaning within these realms in a way that makes science and rational thought not really very useful. Scientific methods can be applied in order to try to unify all knowledge under the umbrella of reason and rational thought and the scientific method, but I think the scientific method will not eventually capture a satisfactory understanding of all aspects of human experience and life.
The continuing attempt by certain scientists and atheists to remove superstitious belief in a God who must intervene to make things happen in this universe is usually quite unsatisfactory. They are arguing against a level of theological understanding that seems to only exist in certain fundamentalists or very conservative and traditional theologians. It seems a waste of their intellect. I suppose there are enough literally-minded religious traditionalists and fundamentalists running around in our world to keep the atheists and scientists busy, but rarely do I see anything written in a sensitive way that would really persuade a literalistic believer. Nor do I see a sophistication in the dialogue that would engage with someone who had a more sophisticated critique of scientism or a more post-modern theological belief system. Mostly, these books are intended for an audience that already accepts science and tends to have a simplistic view of religious believers, or a condescending view toward religious faith. It is of course interesting news that cosmologists have answers they find logical and scientific to questions such as why is there something instead of nothing, and how did the universe emerge out of the pre-universe, and why is it that our universe has scientific laws and constants that seem to make it such a nice place for life. But to then take answers to such questions and make a claim that this tells us something new and important about God is not very convincing.
I continue to have faith, but I see my faith as being outside of my scientific approach to the world. My faith is unscientific. That’s not a problem that needs to be resolved.