To: sbaker@odu.edu
From: Thomas Meyer
tmeyer@ph.vccs.edu
276 656-0283
Patrick Henry Community College
Subject: Statistics - w/ Dr.Spencer Baker - Homework Assignment #4, Ch 13 - Problem 23, page 345.
Date: March 27, 2004
Use the following algorithm:
1. Get the data
2. Manipulate the data
To conduct the Independent Samples T-test at 95% confidence using two-tailed tests:
Having loaded the data
Click Analyze menu and choose Compare Means
Select Independent-Samples T-test
Move gender into the grouping variable
group 1: enter 1 for males;
group 2: enter 2 for females;
Continue
Move the variables " Independent Self-Construal," "Interdependent Self-Contrual,"
"Individualistic Self-Esteem," "Collectivistic Self-Esteem,"
"Individually-Oriented Achievement Motivation," and "Socially-Oriented
Achievement Motivation" one at a time into the Test Variable box.
Options: 95%
__________________
Note: Right-click the
Independent Samples Test Output and heed the advice on when to use "Equal
Variances Assumed and "Equal Variances Not Assumed."
Having found the proper row in the Independent Samples Table, report the
t-statistic, degrees of freedom, and the probability of these occurring within a
two-tailed test.
Here's how to apply the decision-making rule:
Take a look at question "a" below: If the value read from the significance
of the two-tailed test (.416) is greater than the value of alpha (.05) as it is
in this case, we would not conclude that their was a difference in
independent self-construal between Japanese and American respondents.
If the value read from the
significance of the two-tailed test (whatever that value may be) is less than
the value of alpha (.05) as it may come to be in other cases, we would conclude
that their is a difference in the variable being measured between
Japanese and American respondents.
Question:
The data in the data set named self.sav were used to test cultural differences
(citizen; 1 = Japanese, 2 = American). Using your SPSS data bank, perform
an independent samples t-test using a two-tailed test with an alpha of .05 to answer
the following questions. Use the computer printout to support your
answers.
a. Do Japanese and Americans differ on independent self-construal (indsc)?
Answer:
When Levene's Test for Equality of Variances is high >= .05, as it is in this case, use the data from the row in the Independent Samples Test for Equal Variances Assumed. In this case .144 >.05, so use the row values beside equal variances assumed.
A two-tailed test was conducted on Japanese and American respondents with
respect to Independent Self-Contrual.
Equal variances were assumed. The t-statistic .816 with 197 degrees of
freedom was calculated and compared with the probability of its occurrence at
the 95% confidence level, or alpha = .05. The probability of the
occurrence was .419.
If the value read from the significance of the two-tailed test (.419) is greater than the value of alpha (.05) as it is in this case, we would not conclude that there was a difference in independent self-construal between Japanese and American respondents.
Question:
b. Do Japanese and Americans differ on interdependent
self-construal (intsc)?
Answer:
When Levene's Test for Equality of Variances is high >= .05, as it is in this case, use the data from the row in the Independent Samples Test for Equal Variances Assumed. In this case .393 >.05, so use the row values beside equal variances assumed.
A two-tailed test was conducted on Japanese and American respondents with
respect to Interdependent Self-Contrual.
Equal variances were assumed. The t-statistic -1.277 with 197 degrees of
freedom was calculated and compared with the probability of its occurrence at
the 95% confidence level, or alpha = .05. The probability of the
occurrence was .203.
If the value read from the significance of the two-tailed test (.203) is greater than the value of alpha (.05) as it is in this case, we would not conclude that there was a difference in interdependent self-construal between Japanese and American respondents.
Question:
c. Do Japanese and Americans differ on individualistic
self-esteem (indse)?
Answer:
When Levene's Test for Equality of Variances is high >= .05, as it is in this case, use the data from the row in the Independent Samples Test for Equal Variances Assumed. In this case .164 >.05, so use the row values beside equal variances assumed.
A two-tailed test was conducted on Japanese and American respondents with
respect to Individualistic Self-Esteem.
Equal variances were assumed. The t-statistic -7.750 with 197 degrees of
freedom was calculated and compared with the probability of its occurrence at
the 95% confidence level, or alpha = .05. The probability of the
occurrence was .000 (probability less than .001, since zero probability is not
possible).
If the value read from the significance of the two-tailed test (.000) is less than the value of alpha (.05) as it is in this case, we would conclude that there is a difference in Individualistic Self-Esteem between Japanese and American respondents.
The Japanese mean is lower than the American mean and this is not explainable due to chance alone.
Question:
d. Do Japanese and Americans differ on collectivistic self-esteem (cses)?
Answer:
When Levene's Test for Equality of Variances is high >= .05, as it is in this case, use the data from the row in the Independent Samples Test for Equal Variances Assumed. In this case .388 >.05, so use the row values beside equal variances assumed.
A two-tailed test was conducted on Japanese and American respondents with
respect to Collectivistic Self-Esteem.
Equal variances were assumed. The t-statistic -4.791 with 197 degrees of
freedom was calculated and compared with the probability of its occurrence at
the 95% confidence level, or alpha = .05. The probability of the
occurrence was .000 (p<.001, since zero probability is not possible)..
If the value read from the significance of the two-tailed test (.000) is less than the value of alpha (.05) as it is in this case, we would conclude that there is a difference in Collectivistic Self-Esteem between Japanese and American respondents.
The Japanese mean is lower than the American mean and this is not explainable due to chance alone.
Question:
e. Do Japanese and Americans differ on individually-oriented achievement motivation (ioam)?
Answer:
When Levene's Test for Equality of Variances is high >= .05, as it is in this case, use the data from the row in the Independent Samples Test for Equal Variances Assumed. In this case .107 >.05, so use the row values beside equal variances assumed.
A two-tailed test was conducted on Japanese and American respondents with
respect to Individually-Oriented Achievement Motivation.
Equal variances were assumed. The t-statistic -6.720 with 197 degrees of
freedom was calculated and compared with the probability of its occurrence at
the 95% confidence level, or alpha = .05. The probability of the
occurrence was .000 (p<.001, since zero probability is not possible).
If the value read from the significance of the two-tailed test (.000) is less than the value of alpha (.05) as it is in this case, we would conclude that there is a difference in Individually-Oriented Achievement Motivation between Japanese and American respondents.
The Japanese mean is lower than the American mean and this is not explainable due to chance alone.
Question:
f. Do Japanese and Americans differ on socially-oriented achievement motivation (soam)?
Answer:
When Levene's Test for Equality of Variances is high >= .05, as it is in this case, use the data from the row in the Independent Samples Test for Equal Variances Assumed. In this case .469 >.05, so use the row values beside equal variances assumed.
A two-tailed test was conducted on Japanese and American respondents with
respect to Socially-Oriented Achievement Motivation.
Equal variances were assumed. The t-statistic -3.341 with 197 degrees of
freedom was calculated and compared with the probability of its occurrence at
the 95% confidence level, or alpha = .05. The probability of the
occurrence was .001.
If the value read from the significance of the two-tailed test (.001) is less than the value of alpha (.05) as it is in this case, we would conclude that there is a difference in Socially-Oriented Achievement Motivation between Japanese and American respondents.
The Japanese mean is lower than the American mean and this is not explainable due to chance alone.
filename: StatHW3Ch13Prob23page345TomMeyer.doc
Tom Meyer
Thomas Meyer