[Phys-L] Re: Charts or Graphs, and other Excel stuff

[oneill at SVGS.K12.VA.US (Thomas O'Neill)]

        We use Excel as our primary analysis to the point of exporting
data from Pasco's Data Studio.  My class (first year pre-med level
Physics) just completed an RC lab where we mandated not only that the
data be exported but that the submitted graphs actually plot the Ln(V)
versus time rather than using the exponential plot function in Excel.
Their analysis had to include a statement about the meaning of the slope
and intercept of the best-fit line in this case.
        We stay away from the other fits because once the students
discover exponential, power and polynomial fits, they then will fit a
polynomial to some data set and argue that since the R^2 is closer to
one that polynomial must be the proper one.  The fact that one can fit
any data set to some arbitrary polynomial is lost on them.  There is
nothing more dangerous than a six-year-old with a hammer, even more so
with a jackhammer
        Class (read my old fogy) rules mandate that data must be
presented with no more precision than was displayed on the meter.  This
means that my students have learned to format cells as a matter of
course.  They have also learned that not labeling data columns with
units is the path to destruction or at least a lower GPA.

THO

Thomas O'Neill
Physics
Shenandoah Valley Governor's School

-----Original Message-----
From: Edmiston, Mike [mailto:edmiston@BLUFFTON.EDU]
Sent: Monday, October 03, 2005 5:44 PM
To: PHYS-L@LISTS.NAU.EDU
Subject: Charts or Graphs, and other Excel stuff

I'm getting more and more students analyzing lab data with Excel, which
I encourage, and this is overall great... but there are some things that
irk me.  I am wondering if these things bother other people, or if I am
just an old fashioned crabby guy.

(1) Students aren't referring to their graphs as graphs anymore.
They're calling them charts.  Chart is a pretty broad word and could
include graphs (as well as tables and maps and other things), but I sure
would prefer they call a graph a graph.

(2) They are beginning to refer to the linear-regression line as the
trend line.  Again, trend line strikes me as too broad.  If the line is
a linear-regression line I would prefer they call it that.  If it is a
2nd-order polynomial fit, I prefer they call it that.

(3) Suppose the intercept of a linear regression is an important
parameter for the report.  Excel automatically runs the regression line
from the first data point to the last data point.  It doesn't extend the
regression line to visually show the intercept unless you manually
extend it.  I expect students to do this, but many don't, even after I
tell them.

(4) Students are not inclined to reformat the cells to show an
appropriate number of digits, or to make the table look nicer.  They use
the default "general format" that leaves some numbers showing a 5 while
others show 6.23 while others show 4.142856 (these are in the same
column).

Okay... That's enough examples.

When I harp about these things, students are inclined to say, "Watch
out.  With Edmiston it's his way or the highway."  I'm inclined to think
my way is a standard and/or traditional way of doing things, and I am
not really a quirky person.  Where do you stand on these types of
issues?

Michael D. Edmiston, Ph.D.
Professor of Chemistry and Physics
Bluffton University
Bluffton, OH 45817
(419)-358-3270
edmiston@bluffton.edu