Jo,
I have spent a lot of time on trying to understand how we
use ordinary English to express complex conditional
statements about rates, probabilities, etc. I've studied the
Harper Collins Cobuild Corpus at the Univ. of Birmingham.
I'm still learning!
I have produced a short booklet on using English to
describe ratios such as rates and percentages .
E.g., "the percentage of males who are runners"
versus "the percentage of males among runners"
or "the male death rate" vs. "the death rate of males."
See "Describing Rates and Percentages in Tables" at
www.augsburg.edu/ppages/~schield/MiloPapers/2001BusOfComm.pdf
I have taken to teaching this ratio material to better prepare my students
for the subtlety of conditional probability. I have three other papers on
such matters.
See my web site at: www.augsburg.edu/ppages/~schield/
1. "Difficulties in Describing and Comparing Rates and Percentages"
The section on chance and probability might be useful in helping
students discern the part and whole (the random variable from a
fixed variable/value/condition).
2. "Using Bayesian Inference in Classical Hypothesis Testing"
regarding the use of "if" and "when" in defining Type 1 error:
E.g., "Type I error: occurs if Ho is rejected when it is true."
"When" trumps "if" in identifying the whole.
3. "Interpreting Statistical Confidence"
I wouldn't give these last three papers to students, but they identify
some elements that result in student misunderstandings.
Let me know if I can be of further help in this critical area.
Milo
======================
At 10:27 AM 7/21/2001 +0100, you wrote:
>Dear list,
>
>Does anyone know of good language support materials helpful for students
>whose first language is not English who are learning statistics in English?
>(Particularly those whose first language is Chinese.)
>
>I teach A-level statistics in England and we have a lot of Chinese students
>whose English is quite weak - teaching conditional probability and
>hypothesis testing is a nightmare, as they find it difficult to follow the
>discussions that develop the relevant concepts. The exam questions tend to
>use 'everyday' English, and they find it very hard to convert this into a
>symbolic form that they can deal with. The English support that they do
>have is not very relevant to statistics.
>
>Any suggestions would be very welcome!
>
>Regards,
>
>Jo Tomalin
Milo Schield, Professor. Department of Business Administration
Project Coordinator, Keck Statistical Literacy Project
Statistical Literacy leader, Quantitative Literacy Team, Project Kaleidoscope
Augsburg College Minneapolis, MN 55454 612:330-1153
Home Page: www.augsburg.edu/ppages/~schield
|