For next time: Back in 121. We'll continue with the Normal
distribution, have quiz.
Math 251 Homework Day 6 (Wed., Sept. 7). (all with
SPSS) At least one in by Monday; all Due by Friday, Sept. 16, Day
10. Hand in each problem as you finish it! They don't
have to be done in order!! (But they're roughly in order from
Copy numbers from tables, don't print tables. Print graphs. Printing: bottom p. 1. Select from the Outline panel what you want, Do Print Preview, then Print. DON'T print everything! SOLUTIONS will be posted...
A. Do p.27, 1.44, Table 1.4 (IQ);
Use a histogram and a boxplot. And do p. 48 1.82 a,b. (mean,
s, median of IQ) You may
find it most efficient to do both together (Hand this one in along with
B) (Datasets SEVENTHGRADE and IQGPA are identical. Use one for
B. Use the data of the problem above (1.44,
You found the mean and standard deviation of IQ
there. Use these values to make a new variable of
standardized values of IQ, z = (x - mean)/sd. (Transform>Compute:
Handout p. 8) What is the standardized value of the first
case, IQ = 111? Make a histogram of the standardized values
to hand in. Save the file for yourself, for future use.
C. Still with SEVENTHGRADE and IQGPA, find for Males and
Females separately, the means, s.d.'s, and 5-number summaries
, and show the male and female
distributions compared in boxplots (following on 1.43, p.27, where you
did stemplots by
hand, Day 2). Discuss how the numerical summaries and the
boxplots "go together." Note in the SPSS dataset the genders are
numerically. Be sure to label on your output which is which (you
do this by hand, checking with the book, or optionally give the Values
column (in Variable view) Labels, Handout p.2 bottom .)
D. Investigate the issue with
data wrongly labeled Nominal or Ordinal. Use the file
from p.26 1.35a (studying
time. You did it for HW by hand):
Open the SPSS file and look at Variable View. Note that
"studytim" variable, Measure
is (incorrectly) Nominal,
but leave it that way. Using Graphs> Legacy Graphs:
Make parallel boxplots
and parallel dotplots, to compare M and F study hours. Examine
your plots carefully, and print them to hand in. Now change the
Measure for Studytim to Scale, and re-make the graphs. (The icon
in the graph dialog box for
Studytim should be a ruler now, not the 3 balls. If it isn't, hit
Reset, so the icon becomes the ruler. Then proceed.) Print the
graphs, compare to the first graphs, and write on your paper what was
wrong the first time.
Some procedures just look at the "Type" of the variable (Numeric/String) but the Dot procedure looks at the Measure! and treats Ordinal and Nominal data as if they are just Names, not numbers at all! No telling what a particular procedure will do!!
E. (Do after the previous one) Import the
text file STUDYTIME.txt (it's in \Data
Sets IPS7e\PC-Text\Chapter 1) (or Mac-Text)
Compare the imported file to the one already in SPSS form. What do you need to do to make sure it doesn't produce garbage?
Moral: Check the Measure column in Variable view--for every data set!
F. Timeplots (Read pp. 20-21)
a) Do pp. 27-8, 1.46 (MARATHON) using a Simple Scatterplot, adding an
Interpolation line.. Print your graph and discuss, as in the
b) Now get the data for EARTHDENSITY p.27, 1.42. Make a stemplot or histogram, I don't care which. Note a little skewness and that it's not exactly bellshaped--sort of flattish on top. Now make a Timeplot on the order the observations are in the dataset. (I believe these to be the order they were taken in.) What pattern do you see? Can this have an effect on the shape of the distribution?