Professional Documents
Culture Documents
#2
Group with: Cindy
Kluz, Shannon Meier
and me
Sara Kornoski
Assignment #2
Pearson r (Correlation)
The research question for the Pearson r (correlation) test is what is the relationship
between how often a person prays and their general happiness. The variable for general
happiness and how often a person prays is both ordinal variables. The independent variable is
how often a person prays, and the dependent variable is general happiness. The null hypothesis is
there is no relationship between how often a person prays and their general happiness. The
alternative hypothesis is that there is a relationship between how often a person prays and their
general happiness. For the variable of general happiness the sample size is three thousand five
hundred and thirteen people. It has positive mean of 1.85, this means that on average people
were pretty happy. The standard deviation is .647, which is low and indicates that the data points
are very close to the mean. It is within the normal range of kurtosis at .659, it has unremarkable
negative kurtosis. The skewness is also within the normal range at .150, it has unremarkable
positive skewness. There are zero outliers. The histogram is normally distributed. For the
variable of how often a person prays the sample size is also three thousand five hundred and
thirteen people which is the same amount for the variable general happiness. It has a mean of
positive mean of 2.74 which means that on average people prayed several times a week and more
than once a week. The standard deviation is 1.717 which is high and indicates that the data points
are spread and not close to the mean. It is within the normal range of kurtosis at -.871 and has an
unremarkable negative kurtosis. It is within normal range of skewness of .718. It has
unremarkable positive skewness. The histogram has positive skewness but the skewness is
unremarkable. There are zero outliers. Neither one of the variables were a candidate for
transformation. The transformation (extra credit) of square root transformation was done even
though a transformation was not needed. The transformation was on the variable how often a
Assignment #2
person prays because it has a less normal skewness and kurtosis result than the variable general
happiness. The square root transformation test changed the skewness to .511 and changed the
kurtosis to -1.088. The transformation made the kurtosis greater and the skewness lower, but s
expected the statistical significance stayed great.
Descriptive Statistics Table
Percent
Percent
Percent
GENERAL HAPPINESS
3513
98.7%
46
1.3%
3559
100.0%
3513
98.7%
46
1.3%
3559
100.0%
Mean
Standard
Skewness
Kurtosis
Deviation
General
3513
1.85
.647
1.50
-.659
3513
2.74
1.717
.718
-.871
Happiness
How Often
Person Prays
Square root transformations
Valid
Zscore:
Zscore: HOW
GENERAL
OFTEN DOES
HAPPINESS
R PRAY
3540
3529
19
30
N
Missing
Percent
Valid Percent
Cumulative
Percent
Valid
-1.31591
1043
29.3
29.5
29.5
.23016
1981
55.7
56.0
85.4
1.77623
516
14.5
14.6
100.0
Assignment #2
Total
Missing
3540
99.5
19
.5
3559
100.0
System
Total
100.0
Percent
Valid Percent
Cumulative
Percent
Valid
-1.01579
1039
29.2
29.4
29.4
-.43414
1056
29.7
29.9
59.4
.14751
404
11.4
11.4
70.8
.72917
214
6.0
6.1
76.9
1.31082
423
11.9
12.0
88.9
1.89247
393
11.0
11.1
100.0
3529
99.2
100.0
30
.8
3559
100.0
Total
Missing
System
Total
Correlations
Pearson Correlation
Zscore: GENERAL HAPPINESS
Zscore: HOW
GENERAL
OFTEN DOES
HAPPINESS
R PRAY
1
Sig. (2-tailed)
N
Pearson Correlation
Zscore:
.074
**
.000
3540
3513
**
.074
Sig. (2-tailed)
.000
3513
3529
Chi Square
The research question for the chi square test is, is there an association between religious
preference and trust in people. The variables religious preference and trust in people are both
Assignment #2
nominal. The independent variable is religious preference and the dependent variable is trust in
people. The null hypothesis is there is not an association between religious preference and trust
in people. The alternative hypothesis is there is an association between religious preference and
trust in people. This test was run to assess the relationship between religious preference and
whether or not it has an effect on people trust. The variable trust in people has the categories for
the variable trust in people is from zero to nine, o=IAP, 1=can trust, 2=cannot trust, 3= dont
know, 4=not assigned. It has a sample size of two thousand three hundred and eighty six people
out of a population size of three thousand five hundred and fifty nine. It has a percentage of sixty
seven who responded to this question. Out of the two thousand three hundred and eighty six
people who responded to this question eight hundred and fourteen people with the percentage of
34.1 indicated that that they can trust people, and one thousand four hundred and fifty nine with
the percentage of 61.1 people indicated that they cannot trust people, and one hundred and
thirteen with the percentage of forty seven of people indicated that their trust in people depends.
The rest of this tests data is missing. The variable religious preference has a sample size of three
thousand five hundred and forty eight out of the population size of three thousand five hundred
and fifty nine who answered this question which is 99.7 percent of people. The categories for this
variable are; 0=IAP, 1=Protestant, 2=Catholic, 3=Jewish, 4=None, 5=other (specify),
6=Buddhism, 7=Hinduism, 8=Other Eastern, 9=Muslim/Islam, 10=Orthodox-Christianity,
11=Christianity, 12=Native American, 13=Inter-Nondenominational, 98=Dont Know, 99=Not
Assigned. Out of the three thousand five hundred and forty eight of people a total number of one
thousand eight hundred and twenty with percentage of 51.3 identified themselves as protestant; a
total number of eight hundred and forty one with a percentage of 23.7 people identified
themselves as catholic, a number of sixty two with a percentage of 1.7 identified themselves as
Assignment #2
Jewish, a number of five hundred and eighty two with a percentage of 16.4 identified as having
no religious preference, a total number of twenty four with a percentage of 7 identified as having
a religious preference of other which means that their religious preference was no on the survey
answers, a total number of thirty two with a percentage of 9 identified themselves as Buddhism,
a total number of fifteen with a percentage of four identified themselves as Hinduism, a total
number of five with a percentage of one identified themselves as other eastern, a total number of
twenty seven with a percentage of eight identified themselves as Muslim/Islam, a total number of
sixteen with a percentage of five identified themselves as Orthodox-Christianity, a total number
of one hundred and nine with a percentage of 3.1 identified themselves as Christianity, a total
number of one with a percentage of .0 identified themselves as native American, and lastly a
total number of fourteen with a percentage of .4 identified themselves as being interdenominational. The rest of the data information is missing because out of the three thousand
five hundred and fifty nine a total number of people who filed out the General Social Survey of
2008 not every person completed this question, only three thousand five hundred and forty eight
people filled this question out, which means that eleven people didnt answer this question and
this is why some data is missing. There is a significant relationship between religious preference
and trust in people. The alternative hypothesis is supported since religious preference is not
independent of whether or not people can trust people. The statistical results for the chi square
test is 42.63, the degree of freedom is twenty four which are free to vary, and the p value is
<.011. The percentages for the religious preference of people that can trust are Hinduism with a
percentage of 77.8, Orthodox Christian with a percentage of 54.5, and Native American with a
percentage of 100, the percentage for religious preference of people that cannot trust are
Protestant with a percentage of 62.7, Catholic with a percentage of 60., Jewish with a percentage
Assignment #2
of 53.7, people with no religious preference has a percentage of 59, people with a religious
preference of other has a percentage of 80, Buddhism with a percentage of 55.6, Muslim/Islam
with a percentage of 62.5, Christian with a percentage of 62.5, and Interdenominational had a
percentage of 63.6.
For the extra credit for the chi square test the variables are religious preference, trust in
people, and religion raised in. The variables are all nominal. From the previous test that was
done we know that religious preference and trusts in people are related, but if we add religion
raised in will there be a correlation and statistical significance. The correlation between the three
variables is .010 which is very close to zero and the statistical significance is .620 which is much
bigger than .05.
CAN PEOPLE
PREFERENCE
BE TRUSTED
Valid
3548
2386
11
1173
N
Missing
RS RELIGIOUS PREFERENCE
Frequency
PROTESTANT
Valid Percent
Cumulative Percent
1820
51.1
51.3
51.3
841
23.6
23.7
75.0
JEWISH
62
1.7
1.7
76.7
NONE
582
16.4
16.4
93.2
OTHER (SPECIFY)
24
.7
.7
93.8
BUDDHISM
32
.9
.9
94.7
HINDUISM
15
.4
.4
95.2
OTHER EASTERN
.1
.1
95.3
MOSLEM/ISLAM
27
.8
.8
96.1
ORTHODOX-CHRISTIAN
16
.4
.5
96.5
109
3.1
3.1
99.6
CATHOLIC
Valid
Percent
CHRISTIAN
Assignment #2
NATIVE AMERICAN
.0
.0
99.6
14
.4
.4
100.0
3548
99.7
100.0
DK
.1
NA
.3
11
.3
3559
100.0
INTER-NONDENOMINATIONAL
Total
Missing
Total
Total
Percent
Valid Percent
Cumulative
Percent
CAN TRUST
814
22.9
34.1
34.1
1459
41.0
61.1
95.3
113
3.2
4.7
100.0
Total
2386
67.0
100.0
IAP
1169
32.8
DK
.1
NA
.0
1173
33.0
3559
100.0
CANNOT TRUST
Valid
DEPENDS
Missing
Total
Total
Chi-Square Tests
Value
df
24
.011
33.959
24
.085
Linear-by-Linear Association
.010
.919
N of Valid Cases
2378
Pearson Chi-Square
Likelihood Ratio
42.627
Frequency
Percent
Valid Percent
Cumulative
Percent
PROTESTANT
1820
51.1
51.3
51.3
841
23.6
23.7
75.0
62
1.7
1.7
76.7
582
16.4
16.4
93.2
OTHER (SPECIFY)
24
.7
.7
93.8
BUDDHISM
32
.9
.9
94.7
CATHOLIC
JEWISH
Valid
NONE
Assignment #2
HINDUISM
15
.4
.4
95.2
.1
.1
95.3
MOSLEM/ISLAM
27
.8
.8
96.1
ORTHODOX-CHRISTIAN
16
.4
.5
96.5
109
3.1
3.1
99.6
.0
.0
99.6
14
.4
.4
100.0
3548
99.7
100.0
DK
.1
NA
.3
11
.3
3559
100.0
OTHER EASTERN
CHRISTIAN
NATIVE AMERICAN
INTER-NONDENOMINATIONAL
Total
Missing
Total
Total
nt
Valid
Cumulativ
Percent
e Percent
814
22.9
34.1
34.1
1459
41.0
61.1
95.3
113
3.2
4.7
100.0
Total
2386
67.0
100.0
IAP
1169
32.8
Missi
DK
.1
ng
NA
.0
1173
33.0
3559
100.0
TRUST
CANNOT
Valid
TRUST
DEPENDS
Total
Total
Descriptive Statistics
Mean
Std. Deviation
RS RELIGIOUS PREFERENCE
2.31
2.238
2366
1.70
.550
2366
1.89
1.683
2366
Assignment #2
df1
df2
11
3518
Sig.
.000
T test/ANOVA
The research question for the T test/ANOVA tests is what is the relationship between
religious preference and general happiness. The variables are religious preference, and general
happiness. The independent variable is religious preference and the dependent variable is general
happiness. The variable religious preference is nominal and has 12 categories and the variable
general happiness is ordinal. The null hypothesis is there is not a relationship between religious
preference and general happiness. The alternative hypothesis is there is a relationship between
religious preference and general happiness. The homogeneity of variance test showed a high
statistical significance between religious preference and general happiness at .000. The ANOVA
test showed a statistical significance of .190 in general happiness between groups. The Post hoc
test could not be performed so another ANOVA test of the variables religious preference and
general happiness was done. The variable religious preference is nominal and the variable life
exciting or dull is ordinal. The categories for the variable life exciting or dull is 0=IAP,
1=Exciting, 2= Routine, 3=Dull, 8= Dont Know, 9= Not assigned. No tests and results showed
any statistical significance between the variables of religious preference ad whether people felt
their lives were exciting or dull. The Levenes Test for Homogeneity of Variance test statistics
are 1.509, and the degree of freedom is eleven, and a degree of freedom of 2,374 and
Assignment #2
significance of .121. This test had no significance. The ANOVA test showed a statistical
significance of .092, variance of the group means / mean of the within group variances is 1.60,
and the degree of freedom is 11. The Welch Test gave the following results; statistic was 1.21,
and the degree of freedom of 11 and 1.651, the p value of .304. The Post Hoc Test for each of
the categories showed statistical significance higher than .05 for every category.
df1
1.509
df2
11
Sig.
2374
.121
ANOVA
IS LIFE EXCITING OR DULL
Sum of
df
Squares
Between
Mean
Sig.
Square
6.632
11
.603
894.578
2374
.377
901.210
2385
1.600
.092
Groups
Within
Groups
Total
Implications
The results of my analyses would have implications for social work practice or social
welfare policy by knowing more information about the clients to be more effective and helpful to
them. For some people their religion plays a big part in their lives so it would be beneficial to
know their religious preference and how that could affect their trust in people as well as if people
pray how often and how it would affect their general happiness. In future research the
information from this study would be helpful to study different places with more cultural diverse
people to see if the findings would be different or the same as what I gathered.
10