genbar.gif - 8.9 K

Homework for Introduction to Statistics - Fall 2000


Suggestions:

6.18
a) Either Consumers in Indiana or U.S. Consumers
b) The first of the three will be 18.67+-1.96*(24.95/sqrt(201))
c) If the confidence intervals in (b) overlap, then there probably is no difference otherwise there is a real difference.

6.22
The margin of error was higher because they asked far fewer men. Sample size determines the size of the margin of error.
6.10
(a) Compute x-bar +- z*(sigma/sqrt(n)) as 10.0023 +- 2.326(0.0002/sqrt(5))
(b) Set m=0.0001, z*=2.326, sigma=0.0002, and solve for n in the margin of error formula.

6.46
The hypothesis of "no difference" tells you that this is a two sided test.
Look up the area to the left of z=-1.37 and then double it to get your P value.
Compare this to the values in (a) and (b).
7.6
(a) df=25-1=24
(b) t=1.059, t=1.318
(c) P=1-0.70=0.30 and P=1-0.80=0.20
(d) You can do this.
7.12
(a) H0:mu=0
Ha:mu>0 (it says "increases")
t=(332-0)/(108/sqrt(200))
(b) 332+-2.626*(108/sqrt(200))
(c) How many cases? See page 380's yellow box

7.32
a) Think about how a subject who knows they are being watched could increase the number of species of trees in logged areas.
b) Put the data into Excel or SPSS and analyze it using a two-sample t-test.
c) Use the output from the software to build a 90% confidence interval.

7.36
a) Plot each by hand on a stem and leaf plot.
b) Put the data into Excel or SPSS and run a two-sample t-test.
c) Use the output from the software to build a 90% confidence interval.

7.42
For the information in Analysis 1, plug these numbers into the statistic on the preceding page. Repeat for Analysis 2. Would you draw the same conclusion in each case?
7.44
This is a good problem to use in test preparation. It covers lots of good ideas.


Back to Greg Crow's Homepage
To Dept of Math & C.S.
To Point Loma Nazarene University


Last modified on 12-Aug-2000
Send comments and suggestions to
E-mail: gcrow@ptloma.edu