Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-l] Vary-One-Variable Protocols (was PsychologistsRespond etc...)

While the multivariable analysis is a powerful tool, it is often not usable
in educational settings. The problem is that if you take a number of
teachers, the relevant variables may be difficult or impossible to actually
quatify. So much research is done by picking and training specific teachers
to do specific things. The relevant variable is then varied between 2
discrete states. Notice that you often can not have a variable that is
continuously variable. There are some studies which use multivariable
analysis, but often the relevant variables are only surface features such as
the teacher's years or service, degrees... Deeper analysis is fairly
difficult to do.

Even large scale studies have a major difficulty, that teachers do not
follow the rules so strong effects turn into weak effects. This is why
Shayer, Adey, & Yates have kept tight control over Thinking Science and
discourage implementing it without the necessary training and monitoring.

Educational systems are very messy with unknown variables intruding into the
experiment. And measuring variables often perturbs the system to produce
unintended consequences.

There are few good examples of a multivariate educational experiment. The
following approaches being multivariate. One math experiment showed that
teaching for understanding resulted in near transfer, but teaching
procedures did not. Then they repeated the experiment and first taught for
understanding, followed by teaching a procedure, and finally reversed the
order. The result was that all of the students who were taught the
procedure failed to achieve near transfer, and the students who were only
taught conceptual understanding, and not a procedure could transfer. ETS
does do multivariate analysis, but the important variables are usually
either not varied, or were not measured.

Philip Sadler has done a multivariate analysis on easily determined surface
features. But the deeper features of instruction were not determinable.
Even then he came up with the fact that using a book is detrimental, and
finishing the book is very detrimental. Also the math courses taken were
more important than the physics course in HS to doing well in a college
physics course. But one does not know if the ability to do the advanced
math and college physics are really functions of another variable. In other
words students may not take advanced math because they are not capable of
it, and will likewise have trouble with physics. As far as I know he has no
measurement of the degree of interactive engagement factored into his
studies. IE is still rare enough, and difficult to quantify by student
questionnaires.

Some good multivariate analysis has been don on the standardized tests, and
they reveal that private HS is not better than public, and the Catholic
school system is lower in math than the public schools. But to find this
result you have to have the information about SES. Private schools appear
to do better because their students have higher SES. Shayer & Adey found
that all the measured schools in England achieved the same thing if you
looked at output scores compared to intake scores. These are facts that
people do not want to hear.

John M. Clement
Houston, TX


A powerful tool in the QC arsenal for analyzing the performance of
production systems is called the Taguchi Method. It is a multivariable
analysis technique that bypasses the single variable technique by reducing
the number of experiments whilst also showing variable interdependence.
This method is especially strong since it can show how variables interact.

E.C. Muehleisen

Too much of a good thing is wonderful!

Date: Mon, 19 Nov 2007 18:46:57 -0600
To: phys-l@carnot.physics.buffalo.edu
From: betwys1@sbcglobal.net
Subject: [Phys-l] Vary-One-Variable Protocols (was Psychologists Respond
etc...)

At 02:48 PM 11/19/2007, Richard H., you wrote:

... we not only reemphasize the importance of randomized,
controlled experimental tests of competing instructional procedures,
but also indicate that altering one variable at a time is an
essential feature of a properly controlled experiment. ...

The familiar "vary one variable" approach to experiment,
is now in particular situations, substituted by
multivariate experimentation.

If all confounding factors cannot be specified in advance,
methods of varying several factors at a time can extract
useful experimental responses due to these factors, from other,
perhaps unknown or unrecognized variables.

I sometimes think of this as the spatial or parallel approach to
randomizing the protocol, as opposed to the serial randomization
which is familiar.


Brian Whatcott Altus OK Eureka!

_______________________________________________
Forum for Physics Educators
Phys-l@carnot.physics.buffalo.edu
https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l

_________________________________________________________________
Share life as it happens with the new Windows Live.Download today it's
FREE!
http://www.windowslive.com/share.html?ocid=TXT_TAGLM_Wave2_sharelife_11200
7
_______________________________________________
Forum for Physics Educators
Phys-l@carnot.physics.buffalo.edu
https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l