|
I don’t think the only two choices
are 1) requiring students to calculate statistics by hand (a math colleague of
mine once told me proudly of how, back in the old days, he had to calculate factor
analysis by hand over a few days) or 2) teaching students to use SPSS blindly
to get a result they don’t understand. I believe (and different things
work for different professors) that the happy medium is teaching students to
use a spreadsheet to calculate definitional (not computational) formulas and
using interactive graphs in the spreadsheet and on the internet until they have
the “A ha!” experience about how the procedure works. This is much
different than typing in X and Y to get Z in a statistical package. That should
be done only after the concepts are mastered. Hand calculation is more likely
to lead to frustration than an epiphany. You can simulate hand calculation (minus
the frustration of calculation errors) by using spreadsheets to do the basic
calculations. For example, in my class, correlation is
taught immediately following z-score and we use the z-score formula for
correlation to show how two variables measured on two entirely different scales
can be compared in a correlation and how the summing of multiplied z-scores determines
the direction of the correlation (positive or negative). You can also allow
students to change numbers in the columns and see the effect it has on the
correlation and even on the linked scatterplot. Later, with t-tests and ANOVA,
you can show them how changing individual data points (to increase the
difference between the means or decrease the variability within the groups)
will lead to a higher or lower value of t or F. The increase in
conceptual understanding due to this kind of numerical and graphic interactivity
is just not possible with hand calculation (at least not before the limits of physical
exhaustion are reached J). Rick Dr. Rick Froman From: Michael
Scoles [mailto:[EMAIL PROTECTED] "Slogging
throught the arithmetic" might be a way of getting them to think about
what they are doing, regardless of how exhausting it might be. Three stories: 1) In graduate
school the group under my mentors had to "slog through" the most
tedious procedures for doing 4-way ANOVAs with repeated measures on one factor,
analysis of simple effects, multiple comparisons, etc. Students in other
areas could use the new-fangled procedures (involving large stacks of IBM cards
and BMDP) to avoid the tedium. On several occasions , they visited those
of us in the "backwards" part of the department, printouts in hand
saying, "I've just completed the analyses for my dissertation, can you
tell me what all of this means?" 2) My oldest boy
was a math-whiz at a very early age. One afternoon, he came home from
school with some homework. One question was "15 - 7 = ?",
when he got to this point, he reached for a calculator. I asked him why
he was doing that, since I was sure that he knew the answer. He said,
"Yeah, it's 8, but the teacher wants us to use a calculator."
(Later, I was on a committee to hire a director for a new charter school.
The "winner" was a person who said, rather spontaneously, "Some
people might think of it as rote learning, but children do need to learn
multiplication tables." The school has been remarkably successful.) 3) Earlier this
semester, I gave an undergraduate statistics test with problems that I thought
were simple enough to not require a calculator. One was to obtain the
mean of 4 numbers, which happened to add to 26. Probably about 1/3 of the
class gave answers like, "6 R2," "6.2," or simply "I
don't know how to do this without a calculator." Calculators and
statistical packages can be a high-tech way of "dumbing down." —---------------- From: Marc Carter [mailto:[EMAIL PROTECTED] (For most
students, just slogging through the Michael T. Scoles, Ph.D. --- You are currently subscribed to tips as: [email protected] To unsubscribe send a blank email to [EMAIL PROTECTED] |
