Doubts of my students: Expert teaching is no better than good-enough teaching
February 24, 2012 at 10:27 am 23 comments
Teaching is a great job. I particularly appreciate how teaching keeps me thinking and questioning, which is particularly important for an education researcher. I’m teaching two classes this semester. I’ve mentioned recently how my data structures class has me thinking about new kinds of practice activities.
I am also teaching a course on educational technology, where we’re reading How People Learn. Chapter 7 is a fascinating read with three detailed accounts of high-quality learning environments with expert teachers, one each in history, mathematics, and science. The chapter includes some strong claims about teaching:
The interplay between content knowledge and pedagogical content knowledge illustrated in this chapter contradicts a commonly held misconception about teaching–that effective teaching consists of a set of general teaching strategies that apply to all content areas. This notion is erroneous….These examples provide glimpses of outstanding teaching in the disci- pline of history. The examples do not come from “gifted teachers” who know how to teach anything: they demonstrate, instead, that expert teachers have a deep understanding of the structure and epistemologies of their disciplines, combined with knowledge of the kinds of teaching activities that will help students come to understand the discipline for themselves. As we previously noted, this point sharply contradicts one of the popular—and dangerous—myths about teaching: teaching is a generic skill and a good teacher can teach any subject.
We had a great discussion in class about this last night. HPL is claiming that an expert teacher has (1) discipline knowledge, (2) understanding about teaching and learning, (3) understanding of conceptual barriers that students face in the discipline, and (4) a set of effective strategies for addressing those conceptual barriers. (3) and (4) on that list is what we call pedagogical content knowledge, discipline-specific knowledge for how to teach that discipline. My students don’t argue that CS PCK (pedagogical content knowledge about teaching CS) doesn’t exist. They just argue that it’s not necessary to be “effective.”
It may be a “dangerous myth” but my students cling to it pretty stubbornly. “If you know the content, and you know about how people learn, then you can teach that content. You may not be as good as a teacher with years of experience, but you’re good enough.” That’s almost an exact quote from one of the students in my class last night. I tried to argue that, not only is it better to have CS PCK, but we can also teach CS PCK, so that a first year teacher can be much more effective than a brand new teacher who doesn’t know anything about student problems or teaching strategies. They pushed back. “How much more does PCK contribute to being a good teacher, beyond just knowledge of the discipline and knowledge of learning sciences?” Since I don’t know how to measure knowledge of CS well, nor how to measure CS PCK, I have two unknowns, so I can’t really answer the question.
One way of interpreting my students’ comments is sheer hubris. These are young, smart Georgia Tech undergrads (and a smattering of grad students). In their minds, they are intellectually invulnerable, able to tackle any academic challenge, and certainly better than any teacher from a school of education. Several of them mentioned Teach for America in their comments, an organization whose existence encourages them to think that teaching is not so hard. Maybe their comments also are the thoughts of expert learners — these students have had to teach themselves often, so they don’t see expert teaching as a necessity.
Another way of interpreting my students’ comments which is much more intellectually challenging is that the difference between an effective and expert teacher is hard to see. A recent NYTimes article speaks to the enormous value of expert teachers — over a student’s lifetime. Barbara has pointed out that, in her experience, the first year that a teacher teaches AP CS, none of his or her students will pass the AP CS (with a score of 3 or better). Even some veteran teachers have few test-passers, but all the teachers who get many test-passers are veterans with real teaching expertise. But how do you make those successes visible? As we’ve talked about here before: How do we measure good teaching?
As a teacher of education research, I wasn’t so successful yesterday. I failed at convincing my class (at least, a vocal group of students in my class) that there is some value in expert teaching, that it’s something to be developed and valued. What I worry is that these are not just the thoughts of a few undergraduates. How many more people think that it’s easy to learn to be a teacher? How many other adults, voting citizens, even members of school boards agree with my students — that expert teaching is not that much better than effective teaching, so hiring a bunch of young, smart kids to teach is good enough?
Entry filed under: Uncategorized. Tags: cognitive science, computing education, learning, learning science, teachers, teaching.
23 Comments Add your own
Leave a comment
Trackback this post | Subscribe to the comments via RSS Feed
Computing Ed Research - Guzdial's Take 
1.
Darrin Thompson | February 24, 2012 at 10:41 am
Have your students done much actual teaching? If you had asked me 15 years ago if I could have taught basic CS concepts, I would have had no trouble telling you that I would figure it out. A couple rounds of motivated students and “The Camel has Two Humps” results and my hubris vanished.
2.
Mark Guzdial | February 24, 2012 at 10:44 am
Some have, Darrin, and I agree that actual teaching will shake that hubris. However, that can’t be our general answer. We will not get every member of every School Board to teach (and especially not every voting citizen), in order to convince them that expert teaching is valuable.
3.
Darrin Thompson | February 24, 2012 at 10:50 am
I’m definitely not suggesting that one must always fail first or that failing is the answer.
One of my students, in the Camel’s lesser hump, if you will, told me he was grateful for my class. It taught him that computer programming was not for him. He went into AV production. I’ve felt awful about that poor kid ever since.
So no, I’m not suggesting that every teacher must ruin 10-12 students as part of their own education.
I am surprised that students with actual field experience would be so convinced. But hey, surprise is my second favorite part of learning.
4.
Alan Kay | February 24, 2012 at 11:00 am
Hi Mark
If a person hasn’t learned much — and many different surveys now show that this is not just true of most Americans, but most college students — then it is easy for them to believe that “teaching is easy — it’s just the process that got me where I am and it was easy for both the teacher and me — there’s not that much you need to learn and it was easy for me to learn it”.
This is why I keep harping about “high minimum thresholds” and being really tough about them with the entire system. Desires for “success” and retention of various kinds in most parts of the system seem to be dominating content.
I think I’ve mentioned this before, but: the highest level of assessment in national reading is “proficient”. This is pretty low on their 500 point scale, and seems to have a rather low threshold. (This is in accord with the current practice of people trying to make many kinds of assessments look better than they are.)
With this kind of gaming of the system, one would expect that a very high percentage of 4 year college graduates would be found to be “proficient” in reading.
So it’s a big surprise to find that the 1992 national assessment could only identify 40% of 4 year college graduates as “proficient”, and 11 years later could only find 31% to be “proficient” (this large decrease is marked “significant” by the assessors).
This would be alarming with a high threshold. With the low threshold they use, it is at least a national disaster.
It is also an interesting view of what 4 year colleges might be actually about these days.
We can take this:
(a) as an exemplar and analogy for thinking about how well other subjects are being learned, and
(b) as a deep barrier to a very important set of processes that are likely more important than whether particular teachers are expert or not, namely, the meta-skills of being able to learn from books and other media with human teaching as a side strategy.
“(b)” is out of the scope of this reply, but given the metes and bounds of actual knowledge these days it is likely the more important topic of conversation.
Cheers,
Alan
5.
David Klappholz | February 24, 2012 at 11:35 am
Re: “One of my students, in the Camel’s lesser hump, if you will, told me he was grateful for my class. It taught him that computer programming was not for him. He went into AV production. I’ve felt awful about that poor kid ever since.”
About ten years ago a CS major with a near straight-A average told us, at the end of senior design, that working on teams to develop software wasn’t for him, and that he’d be looking for something else to make a living. Neither he nor we felt awful; we had helped him make an important career decision, and were glad we had.
6.
Jason Orendorff (@jorendorff) | February 24, 2012 at 12:43 pm
I think the title of this blog post is unfair to your skeptical students. If you can’t measure it, they are right to be skeptical.
Mark Lepper and some colleagues did a bunch of work where they video-recorded tutoring sessions in elementary math. One of their findings was that expert tutors have pedagogical content knowledge—i.e., successful tutoring correlates with PCK. How they defined and measured those two variables, I don’t know.
7.
gasstationwithoutpumps | February 24, 2012 at 1:01 pm
I believe that pedagogical content knowledge is important, but I’m not so sure that anyone is teaching it in a useful way.
I see a little in physics with the the Modeling Instruction workshops, and a little in math, but I’ve not seen any coherent presentation of the PCK for teaching programming.
Please give a few good references (books or meaty articles) that would help me learn the PCK for teaching programming to non-programmers (biologists), as I am tasked to do that next year, and I want to do a good job. I don’t think that my usual teaching strategies will work well with this group, and I’ve already been surprised by some anecdotes about what college students learning programming for the first time have difficulty with (since I did not see the same problems teaching elementary-school kids informally).
8.
Mark Guzdial | February 24, 2012 at 1:13 pm
I know of only one book on CS PCK: http://www.amazon.com/Guide-Teaching-Computer-Science-Activity-Based/dp/0857294423/ref=ntt_at_ep_dpt_2 I know that this is used at Columbus State for teaching CS methods. Purdue uses a collection of papers. So, in short, there are VERY few courses for teaching CS PCK. That doesn’t mean it doesn’t exist — the fact that there are expert CS teachers is evidence. However, there are few efforts to distill that knowledge, to make it accessible to new teachers, and to teach it to new teachers.
9.
gasstationwithoutpumps | February 24, 2012 at 7:46 pm
Thanks. Our library had electronic subscriptions that allowed me to download both “Guide to Teaching Computer Science” and “Reflections on the Teaching of Programming”. Reading them on-screen will be a lot slower and more awkward than reading them on paper, but at least I have access.
10.
Laura | February 24, 2012 at 1:51 pm
I’m participating in a P2PU course on teaching web programming and it’s primarily about the PCK knowledge. How do you anticipate where people will get stuck and be prepared to address that? I have a lot of knowledge about teaching and learning, more than I do about my subject matter, I’d say, but even I’ve been surprised that students don’t grasp loops immediately. And if they don’t grasp those, then they’re going to struggle later. 3 & 4 are what separate the pretty good from the great, imo.
11.
Franklin | February 24, 2012 at 1:57 pm
I will offer two perspectives here briefly:
(1) Thinking of my formal student days, I usually encountered so much teaching that I considered poor (and even malicious and detrimental) that I did not even have a mindset of evaluating what might be good or great. And so I simply blocked out “teaching” and fended for myself, using the classroom as a place to nap and daydream, and doing my real work at home with my books, etc. That said, I do remember a few teachers that lit up a love of learning or thinking or taught some other life lesson that I remember to this day. But none of that excellence had anything whatsoever to do with content. It had to do with recognizing when I was bored, or lonely, or discouraged, or such things, and offering in a very unique way some kind of encouragement, alternate assignments, etc.
(2) As a teaching assistant in grad school, I had my first real experience teaching (undergrad CS). I was shocked by how ineffective I was, and knew it, and did not have the tools to fix the problems. So I know that teaching is hard. Meanwhile, I’m sure that the ten percent of the students who “really” got the material would have done just fine without my existence, and these are probably the kinds of students who go around thinking that there isn’t much of a different between good and great.
12.
Bri Morrison | February 25, 2012 at 11:54 am
I think I might have asked your students if they ever had the experience of a “not so good” teacher and a “great” teacher in the same subject and what they think made the difference. This is especially true if the discipline was one that was not their strongest. I do believe that in any given class there is a subset of students that will learn regardless of the effectiveness of the teaching and there is a subset that probably won’t grasp the material to the level of what we would define as “proficient” no matter the effectiveness of the instructor. Where the real difference occurs is with the students between those two groups. An effective teacher with discipline knowledge, learning science knowledge, and PCK will allow those students to master the concepts while another teacher without that arsenal of knowledge won’t.
That’s where I believe the “measurement” of teacher effectiveness lies. Not that I’m arguing for measuring this will a single class. But over time given enough data (students), measuring the performance of the “mid 50%” in that class and follow-on classes define the performance of that instructor in that class.
13.
Alan Kay | February 25, 2012 at 12:48 pm
Hi Bri
How would you characterize the subset that will learn regardless?
Adele and I used to say “10%” because that’s what we thought we were seeing. We also used “10%” for those who needed more help than could be given with the time and resources allotted.
We used to say “Good pedagogy is all about the 80% in between”.
That led to criteria we’ve used since the late 70s: that real assessment of a “curriculum and process” needs about 3 years of trials, and the threshold for success is 90% fluency (the 10% that are independent of the pedagogy plus the 80% that aren’t).
A weaker (but perhaps more realistic) threshold could be gotten from normal distribution ideas — perhaps expand to 17% on the top and bottom categories and aim the pedagogy at the middle 66%.
This never seemed to be in accord with our experience. So we always went for the 10-80-10 categories instead.
14.
Bri Morrison | February 28, 2012 at 10:37 am
Alan,
My experience is similar to yours. Generally I find about 10% of a class will learn regardless of what I do. These students are usually very self-motivated and have mastered the “education game.” They have the meta-cognitive skills to be able to know what they don’t know. Interestingly I think the lack of meta-cognitive skills applies to the bottom 10%, they think they know a lot when in reality they have mastered very few skills.
I completely agree with your 3 year trial edict. I find this is true when I teach a new class…it’s usually the third time before I am remotely satisfied with the structure and content. However I think if your 90% fluency standard was ever required for new curricula then most current educational improvements would be doomed. Perhaps they should…
15.
Alan Kay | February 28, 2012 at 2:02 pm
Hi Bri
We applied the “90% fluent” threshold to our research with children and wound up failing for more than 20 years!
But we paid for this research ourselves from our other research funds, so we didn’t have to answer to anyone but ourselves.
Eventually we did enough work and made enough systems to get to the 90% fluency levels (early 2000s) for the parts of math, science and computing we were trying to make curriculum for.
As researchers, the “90 fluency threshold” was very helpful in helping us find out that we knew very little about most aspects of education, children, even science and math.
One way to think about your last sentence is to note that as more and more accurate measurements were done by Brahe and others, proponents of the geocentric theory added more and more epicycles to save their worldview.
The problem is that adding epicycles does provide a version of “improvement” but one never gets to escape from the weak paradigm into a more workable one.
I think they need to set thresholds that can only be achieved by really powerful rather accurate methods that actually help children.
Cheers,
Alan
16.
AnnMaria | February 26, 2012 at 3:29 pm
I’m surprised at their attitudes. I have a Ph.D., 30 years experience and teach statistics very well. I would never think I could teach Calculus or (shudder) English just as well without years of study and practice .
17.
guy | February 29, 2012 at 11:58 am
Very interesting dialogue… The book that turned me on to this area was Schools for Thought by John T. Bruer. John covers concepts and techniques that have been used by cognitive scientists, then goes on to detail what we’ve learned about teaching math, science, reading and writing. It would be nice to read about someone or some group that has done the kind of task analysis, protocol analysis, encoding, … (all the cogsci stuff that’s been done for other subjects) for learning introductory programming. What does a student that just gets iteration, conditionals, procedure definitions, etc… know that others that struggle don’t know? Do metacognition skills, some students have developed, make a difference?
I guess that this kind of research is not as much fun as inventing a new programming language or programming environment which will solve all the issues student have with learning to program. It seems to me that if we do not know why some students run into a wall when introduced to iteration (or pick your favorite stumbling block), new tools for teaching that are developed are just shooting in the dark.
18. reestheskin – Expertise is important. But what sort of expertise | March 12, 2012 at 5:28 am
[…] is a really nice post over at Computing Ed on the issue of expertise in teaching. The issue is about (i) subject knowledge (ii) teaching […]
19.
We used to know how to teach CS in Logo « Computing Education Blog | April 7, 2012 at 1:37 pm
[…] I get into conversations about CS pedagogical content knowledge (PCK), I realize that Gary Stager is right (below). We used to have CS PCK. I remember “Ask […]
20.
Blog Post #999: Research Questions in Computing Education « Computing Education Blog | May 3, 2012 at 6:16 am
[…] call that knowledge pedagogical content knowledge. How do we best teach computer science? How do we help future educators develop the unique skills to teach computer science? Share this:EmailDiggRedditFacebookPrintStumbleUponTwitterLike this:LikeBe the first to like this […]
21.
Learning to Teach Computer Science: The Need for a Methods Course : CACM « Computing Education Blog | November 2, 2012 at 7:12 am
[…] Pedagogical content knowledge (PCK) is the knowledge that teachers have about teaching specific content. “How People Learn” suggests that it’s much more important for student learning than general teaching knowledge. To create credentialing for CS, we need to offer CS methods courses that teach CS PCK. Aman Yadav and Tim Korb teach one of these courses at Purdue, and have an article in this month’s CACM on how it works. […]
22. On Pedagogical Content Knowledge | Teaching Software Carpentry | November 3, 2012 at 10:09 pm
[…] Mark Guzdial discusses in this post, “an expert teacher has (1) discipline knowledge, (2) understanding about teaching and […]
23.
MOOCs are a fundamental misperception of how teaching works « Computing Education Blog | January 4, 2013 at 8:29 am
[…] on the domain. Teaching is not a generalized skill. The most effective teachers have a lot of pedagogical content knowledge — they know how to teach the domain. The same general course structure is not as effective as […]