| Hand in
Wednesday (From
D&V unless otherwise noted) Scatterplots: (Copied from Day 12) SPSS Handout: Repeat the work of page 1, and do problems 1-5 on p. 3. Keep this work and hand it all in when all problems have been assigned. HW in Activstats: Go to Chapter 7, use the menu
button
with the House icon. Scroll thru the problem list to find the
ones
given. In each problem involving data, a button will allow you to
launch SPSS and open the correct file. Then save the file for
yourself,
do the analysis. MRA-95-13 (SPSS, and pencil) Corn plants. This is a
first introduction
to the idea of predicting or estimating a "typical" y for a given x
value.
Ch. 8 will do an important special case of that. |
Read, to
discuss ----- +++++ Correl. p.130, 21Politics, 24Sample survey |
Optional
- - - - - -
|
Relationships:(D&V Ch 7 thru p.117,
AS7-1&2 ) Day
12
Handout on SPSS Scatterplots etc.
(D&V Ch. 7-10, AS 7,8,9)
govsal_vs_pay.sav
is the file used for most of the handout. (In SPSS for Class 05 folder)
Correlation experiments:
ActivStats 7-3, 2nd activity: Slider to see shapes ~~ r's.
3rd activity: non-linear data and r's. 4th: center and scale change.
Website, http://www.whfreeman.com/scc,"Statistical
Applets", Correlation/Regression. Play with data points,
observing the Correlation Coefficient.
Check in the "Show
Mean X & Mean Y lines" box. See how much is in each quadrant.
Compare with correlation coefficient.
The
correlation
coefficient r is a numerical measure for how strongly
linear
(and in what direction) the relationship is. Doesn't
substitute
for a scatterplot.
Use if data is: 2 quantitative variables,
& "nice":
&& One cluster/cloud/band.
"Straight enough."
Outlier(s)? Do with/without & be cautious.
Using SPSS (p.4, Scatterplot handout) Analyze>Correlate>Bivariate
Properties:

--You won't have to calculate a correlation coefficient by hand. This
formula is a bad one for hand computation (roundoff error); if you must
do one by hand, find the computational formula in an old textbook.
--Eyeballing: sketch xbar and ybar lines, see how much data is
in + quadrants, how much in - quadrants.
Strength of correlation says NOTHING about causality!
Strong
correlation could be:
A causes B/ B causes A/ C causes both
A and B (lurking C)/ just Chance that they go together in this data
set.
| Sievers home | Math151-Fall05/Dayf14.htm | 2pm | 9/25/05 |