Date:         Sat, 25 Jul 2009 15:11:23 -0400
Reply-To:     Amy Green <jtjt456@GMAIL.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Amy Green <jtjt456@GMAIL.COM>
Subject:      What statistical method should I choose from?

This may be a simple question for you, but I still confuse, please help!

There was a survey reported in a news that classified a random sample of
PP members according to employment (business or government) and membership
(<2 years = new member, 2+ years = established member). Next, I asked the
PP to provide me with the actual years of membership and corresponding
salary of each member in the sample.
I am interested in seeing whether years of membership has an effect on
salaries.
a. What should be my first approach to answer this?
b. What is the more detailed approach to determine this?
c. How would I determine the correlation between these two random
variables?
d. How would I determine the fraction of the variance in the salaries
which is explainable by its relationship with the years of membership?
e. I was asked to determine if the mean salary in government is
40,000 per year but I have an outlier in the data. Which method should be
used to determine this?

Choose from the possible methods:
1 boxplot
2 normal probability plot
3 descriptive statistics
4 sign test
5 confidence interval for a single mean
6 independent samples t-test
7 paired samples t-test
8 mann whitney U test
9 ANOVA
10 Levene's test
11 Kruskal-Wallis test
12 Chi square test
13 Pearson's correlation
14 Spearman's correlation
15 Rsquare
16 Simple Linear Regression
17 Fisher's exact test or Likelihood ratio test
18 Scatter Plot

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD