Re: experimental analysis

["David T. Marx" <dtmarx@SIU.EDU>]

<snip>
> For many years I taught an advanced laboratory for senior physic
> majors and I would start out with experimental statistics.  I finally
> gave up!  The subject put me and all of the students to sleep so fast
> it was difficult to recover and do any experiments.       WBN
> Barlow Newbolt
> Department of Physics and Engineering
> Washington and Lee University
> Lexington, VA 24450

        I took such a course at U. of Wisconsin-Madison.  I had exactly that
experience - sleep.  However, I felt that such discussions would have been
more useful if they were introduced as they were encountered in the various
experiments.  Physics first - statistical analysis second - as a rule.  The order
should be:
        1.  Understand the background material of the experiment.
        2.  Understand the equipment to be used in the experiment.  How does it
work?  What are its limitations?
        3.  Gather experimental data.
        4.  Analyze the data within the context of the limitations of the
experimentor and the equipment.  Perform statistical analysis.
        5.  Formulate conclusions.

        In my experience, lab instructors place most emphasis on steps 1 and 3
above and too little on steps 2, 4, and 5.


David T. Marx, PhD
Associate Scientist
Center for Advanced Friction Studies
Southern Illinois University
Carbondale, IL  62901-4343
phone: (618) 453-1168 or -7932
fax: (618) 453-5260
web: www.siu.edu/~cafs