You are on page 1of 7

HOMEWORK FOUR

Homework Four
Kristin M. Deel
1. T-test for Independent 1
First, I organized the data to be sorted by Group 1 and Group 2. Then, I began to run the ttest by navigating to Data > Data Analysis and picking the test called t-Test: Two-Sample
Assuming Equal Variances. I highlighted the column for Group 1 as the Variable 1 Range
and then selected the range for Group 2. I also checked the box to show Labels and selected
the cell for output. After clicking OK, the following was returned:
t-Test: Two-Sample Assuming Equal Variances
Group 1
Mean

5.433333333

Variance
Observations
Pooled Variance
Hypothesized Mean Difference
df

11.70229885
30
7.979885057
0
58
0.137103112
0.44571206
1.671552762
0.891424121
2.001717484

t Stat
P(T<=t) one-tail
t Critical one-tail
P(T<=t) two-tail
t Critical two-tail

Group 2
5.53333333
3
4.25747126
4
30

This was a two-tailed test and the significance level is .05. Since we had 30 in Group 1 and 30 in
Group 2, the degrees of freedom (df) is 58. This is because the degrees of freedom for this
particular test is n1+n2-2 (Salkind, 2013). The results would be written as t(58)=-0.14, p>.05
where 58 is degrees of freedom, -0.14 is the obtained value, and the probability is greater than
5% that on any test of the null hypothesis that the groups do not differ. 0.14 does not exceed the
critical value of 2 so we will retain the null hypothesis.
2. T-test for Independent 2
To test the hypothesis that boys raise their hands more than girls, I first organized the data to
be sorted by Gender 1 and Gender 2. Then, I began to run the t-test by navigating to Data >
Data Analysis and picking the test called t-Test: Two-Sample Assuming Equal Variances. I
highlighted the column for Gender 1 as the Variable 1 Range and then selected the range for
Gender 2. I also checked the box to show Labels and selected the cell for output. After
clicking OK, the following was returned:

HOMEWORK FOUR

t-Test: Two-Sample Assuming Equal Variances

Mean
Variance
Observations
Pooled Variance
Hypothesized Mean
Difference
df
t Stat
P(T<=t) one-tail
t Critical one-tail
P(T<=t) two-tail
t Critical two-tail

Gender 1
7.92857142
9
5.60989011
14
5.65593112
2

Gender 2
5.3125
5.69583333
3
16

0
28
3.00580583
7
0.00276863
1
1.70113093
4
0.00553726
2
2.04840714
2

In this sample of 14 males and 16 females, the df is 28. This is because 14+16-2=28, following
the formula for degrees of freedom which is n1+n2-2 for a t test for independent samples. The t
value (labeled in the chart as t Stat) is 3.006 and the critical value at .05 level is 1.701. This is a
one tailed test because the research hypothesis is directional, stating that boys raise their hands
more than girls. Since 3.006 (the obtained value) exceeds 1.701 (critical value), we can conclude
that boys raise their hands significantly more than boys.
3. T-test for dependent 2
To test the hypothesis that counseling made an impact on adolescents, I used the Excel
ToolPak for this one as well by going to Data > Data Analysis and selecting the t-test Paired
Two Sample for Means. Next I entered the cell ranges for variables 1 (A1:A21) and 2
(B1:B21), checked the box to show labels, selected the output range, and clicked OK. Here
is the result (interpreted below):
t-Test: Paired Two Sample for Means

Mean
Variance
Observations
Pearson Correlation
Hypothesized Mean Difference
df
t Stat

Before Intervention
5.48
5.140631579
20
0.193151428
0
19
-3.892546471

After Intervention
7.595
2.003657895
20

HOMEWORK FOUR

P(T<=t) one-tail
t Critical one-tail
P(T<=t) two-tail
t Critical two-tail

0.00048967
1.729132812
0.00097934
2.093024054

In a nutshell, the intervention worked well. How do I know that? The results state that t(19)=3.89, p<.05. In this case, we will accept the research hypothesis that counseling adolescents had
an impact on the tolerance of other adolescents who are ethnically different from them.
4. Single Factor ANOVA
The first step in determining whether amount of practice has to do with the speed of
swimmer is to rearrange the data so that it is separated by Practice times. Next, go to
Data>Data Analysis in Excel and select ANOVA: Single Factor. For the input range, I
entered the cell addressed for the three groups of data (including labels) and selected
Columns for grouping. As always, click the Labels box so labels are included. Lastly, select
the output range and voila:
Anova: Single
Factor
SUMMARY
Groups

Count

Sum

Average
58.05
57.9615
4
59.0307
7

Practice 1

10

580.5

Practice 2

13

753.5

Practice 3

13

767.4

ANOVA
Source of
Variation

Within Groups

SS
8.86876
1
914.063
5

Total

922.932
2

Between Groups

df

MS
2
33

4.43438
27.6988
9

Varianc
e
29.8783
3
22.6842
3
31.0789
7

F
0.16009
2

P-value
0.85272
3

F crit
3.28491
8

35

The means for each group are listed in the Average column so the means for Practice 1, 2, and 3
respectively are 58.05, 57.96, and 59.03. An important statistic in this test is the df between
groups and within groups. The probability of F(2, 33)=0.160 occurring by chance is .85 (Pvalue), indicating that number of hours does not impact swimming speed at all.
5. Factorial Design ANOVA

HOMEWORK FOUR

The first step that I took was to reorganize the dataset by pain levels with Drug 1, Drug 2 and
Placebo. This is a 3x2 factorial design since I am also examining pain severity of 1 and 2.
Next, I selected ANOVA: Two- Factor with replication and highlighted the entire area input
empty cells. I picked 10 rows per sample, input the output range, and then clicked OK.
Here are my results (conclusions below):
Anova: Two-Factor With Replication
SUMMARY

Drug #1

Drug #2

Placebo

Total

10
65

10
42

10
32

6.5
0.72222
2

4.2
1.06666
7

3.2
1.28888
9

10
70

10
61

10
32

7
1.33333
3

6.1
0.98888
9

3.2
1.73333
3

30
163
5.43333
3
3.97816
1

10
72
7.2
1.06666
7

10
42
4.2
1.51111
1

10
45
4.5
1.38888
9

30
159
5.3
3.11379
3

10
70

10
53

10
23

7
6.88888
9

5.3
1.34444
4

2.3

30
146
4.86666
7

0.9

6.74023

40
277
6.925
2.37884
6

40
198
4.95
1.79230
8

40
132
3.3

Severity 1

Count
Sum
Average
Variance

Count
Sum
Average
Variance

30
139
4.63333
3
2.92988
5

Severity 2

Count
Sum
Average
Variance

Count
Sum
Average
Variance
Total

Count
Sum
Average
Variance

1.85641

HOMEWORK FOUR
ANOVA
Source of
Variation
Sample
Columns
Interaction
Within
Total

SS
12.4916
7
263.516
7
40.4833
3

df
3
2
6

182.1

108

498.591
7

119

MS
4.16388
9
131.758
3
6.74722
2
1.68611
1

F
2.46952
2
78.1433
3
4.00164
7

P-value
0.06583
7
1.03E21
0.00117
5

F crit
2.68869
1
3.08038
7
2.18365
7

Conclusion: There is no main effect for severity (listed in the Sample row) but there is one
for treatment (Columns). There is no interaction between the 2 main factors.
6. Revisit Correlation and Linear Regression
First, I ran the correlation function in the ToolPak to obtain the correlation coefficient for each
tab. Then, it was necessary to determine the degrees of freedom and run a two-tailed test. I
chose to pursue income and education further because it had the highest correlation coefficient
out of all 4 tabs. The higher the absolute value of the correlation coefficient, the more accurate
we can predict one variable from the other based on that correlation. .63 was the highest
correlation so lets use linear regression to move forward.
Income and education: .63 correlation coefficient, df is 18, p=.05 - .44 is critical value so we will
reject the null because correlation is .63, which is greater than critical value of .44. OR,
r(18)=.63, p<.05.
Formulas > More Functions > Statistical > SLOPE
Data > Data Analysis > Regression, then fill out the form. Here are my results:
SUMMARY
OUTPUT
Regression Statistics
Multiple R
0.628926
R Square
0.395548
Adjusted R Square
0.361968
Standard Error
15574.01
Observations
20
ANOVA
df
Regression
Residual

1
18

SS
2857006147
4365893581

MS
2.86E+09
2.43E+08

F
11.77906

Significanc
eF
0.002973

HOMEWORK FOUR

Total

Intercept
Level of Education

19

7222899729

Coefficient
s
29638.53
16152.83

Standard Error
12045.19709
4706.447956

t Stat
2.460609
3.432063

P-value
0.024207
0.002973

Lower 95%
4332.505
6264.946

Upper
95%
54944.55
26040.71

This tells me the slope is 16152.83 and the intercept is 29638.53, so the formula for the regression line is
determined by using the data returned from the ToolPak, which is:
Y=16152.83x+29638.53
To create the linear regression line, I simply created a scatterplot and then selected for it to show the line,
as seen below:

Level of Education
3.5
3
2.5
Level of Education

Linear (Level of Education)

1.5
1
0.5
0
$0

$50,000

$100,000

$150,000

7. Chi Square
For this one, I started filling in the observed values in the chart, making the 1s in the sample
For, 2s are maybe, and 3s are against. Then, since there were 90 in the sample, I divided
that by 3 and put 30 in each field under the expected column. Calculating the difference
between the 2 values (observed and expected) allowed me to fill in the Difference column.
Completed chart:
Observed
23
17
50

Expecte
d
30
30
30

O-E
7
13
20

Square (OE)
49
169
400
ChiSquare=

Divide by
E
1.6
6
13
20.6

Lower
95.0%
4332.505
6264.946

HOMEWORK FOUR

X2=20.6, p<.05 tells us that there is a significant difference between the set of scores.

You might also like