Comment Elsewhere: Certainty

Context here:

Most comments on blogs – particularly well-read blogs on controversial topics – are not part of a conversation, nor do they add anything beyond a sort of “me too” vote to one side or the other. Hardly anyone actually reads previous comments before posting – I’ve skimmed through, reading more closely what caught my eye – because that’s not the point. The point is to express ourselves, vigorously and publicly: like doing cartwheels in a grocery store, to get noticed briefly and acknowledged, but not really to engage.

So I’m often loathe to comment on well-commented pieces, on the grounds that nobody cares, that what I’d say has probably been said, etc. But Aaron, if you’re still reading comments (Sorry, really), here’s something that hasn’t really been touched on directly here (though I do seem to remember it coming up on twitter).

Sources lie. But they’re all we have.

This is what I tell my history students (http://www.slideshare.net/jdresner/two-things-about-history-and-history-teaching) and it’s one of the most fundamental things about doing history: Sources lie deliberately, sometimes. Sometimes by omission. Sometimes accidentally. Sometimes because of their particular perspective, or because they made a mistake. Sometimes they lie because they are lost to us; a kind of 5th amendment right of history, to destroy materials.

In spite of that, we *have* to *use* our biased, incomplete, poorly written, fragmentary sources to come to some conclusions about the world. Absolute certainty about anything other than big events is rare; absolute certainty about causality or human experiences is damn near impossible. Sometimes even “preponderance of evidence” (The American civil court standard) is just not reasonable.

We have to be careful about filling in gaps in our knowledge with “likely,” or “logical,” or “reasonable,” suppositions. Human beings are many things, but they often aren’t logical or reasonable, and they often do unlikely things. That doesn’t mean that we can’t come to conclusions, but it means that we have to learn to be comfortable with uncertainty, epistemological humility. We acknowledge that we have built our arguments on possibly shifting ground. And we move on. We build other arguments on those arguments, and so on. Unless some new source or new perspective comes along and changes them. Then we go back and revise.

Nothing wrong with that. Nothing wrong with uncertainty, as long as we acknowledge it. And nothing wrong with living our lives by the fruit of our uncertain conclusions, as long as we’re satisified with the evidence and arguments that we have.

Certainty is nice, but rare. In its absence, resolve to work towards a better world keeps us moving in the right directions, more or less.

Test Three Results

As with the first test, each question was worth up to 4 points, for a possible total of 36. The highest score in the class before extra credit was 30, a little lower; the median was a C+, and the distribution of grades was much more towards the center. The grade scale works out like this:

Grade starts at distribution
A+ 30
A 28.5 7%
A- 27
B+ 25
B 22 30%
B- 20
C+ 18
C 14 45%
C- 12
D+ 10
D 7 17%
D- 5
F under 5 1%

If you answered 9 questions, but failed to answer one from each chapter, I took a 2 point penalty off your grade. (If you didn’t answer all 9 questions, I did not)

If you want to discuss your performance, and how you can improve it next time, feel free to come by my office hours. If you want to dispute your grade, feel free to do so in writing.

I will be in the office for a good portion of tomorrow, Thursday, at least from 10-3; if you want to pick up your test on Friday, let me know because I haven’t set a schedule yet.

Test Two Results

As with the first test, each question was worth up to 4 points, for a possible total of 36. The highest score in the class before extra credit was 32 again; the median was a C+, which is OK but lower than the first (and the distribution of grades shows this); again, nobody who took the test failed. The grade scale works out like this:

Grade starts at distribution
A+ 32
A 30 15%
A- 28.75
B+ 26.25
B 23.25 25%
B- 20.75
C+ 18.25
C 15.25 40%
C- 12.75
D+ 10.25
D 7.25 20%
D- 4.75
F under 4 0%

If you answered 9 questions, but failed to answer one from each chapter, I took a 4 point penalty off your grade. (If you didn’t answer all 9 questions, I did not)

If you want to discuss your performance, and how you can improve it next time, feel free to come by my office hours. If you want to dispute your grade, feel free to do so in writing.

Continue reading

Test 1 Results

One of the reasons it takes so long for me to grade tests is that I grade by questions or chapters, which means that I can be more consistent across tests. It also means that your grade on each question is independent of your grade on the other questions: it’s entirely possible to get 4 points on some, and zero on others, depending on how well you’ve understood and expressed everything.

Each question was worth up to 4 points, for a possible total of 36. The highest score in the class before extra credit was 32; the median was a B-, which is good, and nobody who took the test failed. The grade scale works out like this:

Grade starts at distribution
A+ 32
A 30 20%
A- 28.75
B+ 26.25
B 23.25 35%
B- 20.75
C+ 18.25
C 15.25 35%
C- 12.75
D+ 10.25
D 7.25 10%
D- 4.75
F under 4 0%

If you answered 9 questions, but failed to answer one from each chapter, I took a 2 point penalty off your grade. (If you didn’t answer all 9 questions, I did not)

If you want to discuss your performance, and how you can improve it next time, feel free to come by my office hours. If you want to dispute your grade, feel free to do so in writing.

Continue reading

Results from the First Test Post-Survey

The first test, like most of my survey tests, was a short-essay ID test, with terms taken from (and arranged by) the textbook chapters, supplemented with theory terms and a few names from the lectures. Twelve terms (out of 25 choices, out of about 50 terms on the study guide) over a 50-minute class.

The class after the first test, I did a short anonymous survey to see how people felt about the event. I feel reasonably good about the first question, though their views may change once they’ve actually seen grades: it suggests, though, that I prepped them appropriately for the kind of work. Question 2 is pretty typical results, it seems to me, though the “studied regularly” number should be higher given that I’ve assigned regular homework this semester which feeds directly into the test. The third question results are interesting: there doesn’t seem to be a correlation between amount of study time and expected grade (I’ve not actually performed any statistical tests on this data, though; N=28); the distribution of answers is a little more optimistic than my test results tend to be, but I won’t know until they’re graded. The comments are not particularly surprising, overall.

Quick Post-Test Survey

1. The test was

0%                  a. a cakewalk

15%                b. easier than I expected

60%                c. about as hard as I expected

20%                d. harder than I expected

5%                  e. impossible

2. I studied for the test

7.5%              a. not at all

15%                b. a little

25%                c. right before the test

20%                d. several hours in the week leading up to the test

7.5%              e. off and on for a few weeks

25%                f. every week, as part of regular homework

3. I think I got a grade of

10%                a. F

10%                b. D

27%                c. C

33%                d. B

20%                e. A

Other comments or thoughts?

  • Had problems with dates
  • I think 50 minutes for covering 12 definitions fully is not OK
  • Multiple choice?
  • Tough, but well rounded
  • Seeing my grade will help me study for the next test
  • Too many words on the study guide
  • I’m guessing that if I had studied more I would have had a better outcome.
  • Dates threw me off a lot.
  • So many things on first list but not on the test, that it caused me to not learn the subjects
  • I didn’t have the book long enough to study the material
  • Ran out of time. Need about 15-20 minutes more, or learn to write less or faster.
  • Not a good form/type of test.
  • Unrealistic.