Job Discussion Forums

[adventious]

[Non Sequitur]

Points taken.
NS

[Non Sequitur]

PS I've developed an on-the-fly mark slip for Oral English.
Happy for anyone interested to PM me with an email address and I'll send it on over.
NS

[wangdaning]

[adventious]

Thanks for acknowledging a criticism and a degree of maturity too often absent with any public forum.

I'll address what's related to the OP: How teachers derive a percentile. Wangdaning and I haven't related much more than the use of rubrics while NS and DD have cited a few useful activities that lend themselves to assessment, but I'll say this: It is difficult to provide a formula, or detail an algorithm, without shared sets of reference.

I recommended two solutions. The first involved assessing rubrics in terms of axes to effect greater discrimination/differentiation. The second solution, a ranking, merely exploits what the university has required. Large classes are particularly amenable to a ranking to achieve a distribution with quantiles approaching percents.

[donkeydonkey]

Deep breath, and hear we go.
I'm home from the bar and drunk on the freedom of 5 days off for Mid-Autumn Festival. Oh, and a couple of Laoshans, some baijiu, yaojiu, and some other jiu that had a bulls *beep* soaking in the jar. I'm now ready to tackle a few questions raised.
"How have you derived that ranking? What do you rely on?
Absent criteria, its utility to the validity of a rubric isn't evident."
Yep, i was very sloppy is my explanation or lack there of. I feel like I need to discuss how my raw scores are derived in order to talk about the ranking. Maybe a complete over view of what I do will be helpful.

Here is my grading in its' entirety.
My required split is 60% Final exam and 40% everything else. I am required by the department to include percentages for participation, attendance, homework/quizzes. I use my Rubric for my midterm and final exams and those scores stand alone as 60% for the final and 15% for the midterm. Attendance: I allow 2 absences after which I mark an absence and use that as an percentage. Freshman don't start until week 6 of the 18 week semester so there are 12 weeks of two classes per week. So If one of my freshman misses 4 classes then 2 are excused. Leaving an attendance score of 22/24 or 91.66%. This is rounded up to a 92% in attendance. Attendance is 10% of the final grade. Homework and quizzes. Homework is graded as credit/nocredit. If it is not handed in or is of such low quality that it shows zero effort then no credit is given. A credit for homework counts as 1 point. If 10 homework assignments are given in a semester and the student receives credit for 7 of them then 7/10 or 70% for Homework. Quizzes are always 10 question of 1 point each. I add the points gained and divide against points possible for a percentage. The quizzes and homework are added together to account for 10% of the final grade. Participation: 5% of the Final grade. Here I admit to woefully submitting to subjectivity. This essentially devolves into a ":do I remember you sleeping often?" Are you one of the ones that I must tell to leave your cell phone alone all the time? Is my Impression that you are generally attentive and interested while in class[b] It pains me to write it, but there it is, my shameful and subjective participation grade there for all the world to see. I need to think about the method of using seating as an indication of participation. My initial reaction is that it may be very good for the "individualistic" west where single students make these decisions, but the group behavior of my Chinese students may effect the outcomes disproportionately. I think the main consideration of my students when they chose a seat is "can myself and 5 roommates fit in this row." Disinterested roommates wind up in front, while attentive and participatory roommates are dragged to the back. It's a bit more of a human tetris game than an indication of individual interest. Then again I'm under the influence of an alcohol that had a bull's *beep* soaking in it, so perhaps this is something to think about another day.

[donkeydonkey]

The raw scores derived above can be fixed into a bell curve. I do this in different ways. I will look at the entire semester's grades. I will exclude the final in order to minimize the effect of the rubric on the curve but leave in the midterm so that it is not completely absent. The curve allows for 10% A, 20% B , 40 % C, 20% D, 10% F. This also creates a ranking from highest score to lowest. A few tensions are created by differing scores. One is that the students grade is heavily influenced by the midterm and final (75% of the final grade and dependent exclusively on the rubric). Another tension is that I have my own observations of student abilities. This often points to what language is in my rubric. How do the articulated expressions of achievement match with the grades that I have assigned throughout the semester? I can watch the movement of individual students go up or down and pontificate, analyze, and guestimate as to why it is that one student gets an A in one scheme while receiving a C in the other. At its' most basic level it is just generating data points and then interpreting them. For example, This led me to realize that in my first year of teaching, I used a lot of meaningless, undefinable, nonsense in my rubric's descriptive sections. Things such as "Displays a high level of fluency." This seemed pretty good at the time, but now I am absolutely embarrassed by it. What does "fluency" mean? What is the differentiation between a high level and another level of "fluency?" Is this just a rubber stamp in academic clothing? It's nonsense. Fluff and jargon that sounds meaningful, but is really just a cover for a subjective, gut feeling of "I thank she speaks real purrtay." I prefer something that is a little more meaty. "Pronunciation is mostly correct and does not hinder the transfer of meaning." That is sexy. I know what that means. My students know what that means. I can defend that statement and, therefore, can defend the grades that are derived from that statement. "Um, Mr. Donkey Donkey, what do you mean by fluency? I want to know so that I can achieve it next semester. And how do I know if I'm at a high, intermediate, or low level of fluency?" "Um, Um, Um hold on. I'm going to drink my bull *beep* alcohol and think about it for a few minutes, Liu Zhengxia." Complete nonsense. I can't defend an almost fluent, but not quite there type statement.

[adventious]

[donkeydonkey]

Thank you for the links. They provide an interesting read. The flipped classes are particularly interesting to me as potentially serving to better utilize resources such as teachers and technology. The links discuss younger students and some high school students in relation to learning outcomes and homework. i am curious what the literature has to say about University students who meet once or twice per week in regards to homework. My initial, and far from exhaustive, search discovers U.S based university debates, but without much discussion of class frequency, size, and subject. I will continue to search, thank you for setting me on the path. Certainly the links provide good food for thought and are appreciated.

Ranking based Grading in ESL or Are my Chinese University students normal?
It's the norm, upon which relative grading is based, that I think is disrupted in University classrooms. This is why I, obviously, fall on the criterion-based side of these things. Gaussian models, to my knowledge, in no way predict that small, manipulated sample sizes should adhere to normative distributions. Only that the larger populations from which these samples are pulled should be normative. A sample size or 30, 40, 50, or 60 students who have been selected because of their superior academic performance should not be expected to adhere to normal distributions. It is, in my humble opinion, like using normative height data to predict the heights of the L.A Lakers. While normative distribution properly describes the population at large many, if not all, university students are intentionally selected because they are outliers. Here is a discussion that highlights these concerns. For the relevant sections, Scroll down to the bottom of the page to Grade Distributions followed by Criterion-referenced grades.
http://www.sonoma.edu/users/w/warmotha/grading.html

To be clear, I don't doubt that it is possible to create a normative distribution based around a central point that adheres to 1 SD, 2 SD, 3SD. Instead, I question the validity of the assumed normativity, the basis of the deviations, and the general shape of the graph. The placement seems arbitrary to me. In other words, a graph can be created that shows a normative distribution, but if you zoom in on the graph at a small sample size the distribution will almost always be non-normal. I think University students, because of the selection process and their academic achievement, are a non-normal subsection of an otherwise normative graph. So I don't think my Chinese university students are normal. Thoughts?

[adventious]

[OhBudPowellWhereArtThou]

This is kind of funny. In the States, the only teachers who pursue conversations such as this at such length are new grads who majored in education and those who teach in underperforming schools and need to baffle parents and administrators with, um, brilliance.

[adventious]

[OhBudPowellWhereArtThou]

Adventious,

Why would anyone consider my response an ad hominem attack ? I merely expressed amusement that this dialog is occurring on an ESL board.

LOL

[adventious]

[donkeydonkey]