I.

Statistical Inferences about Two Independent Populations 2

1. Gender and Game 2
2. Gender and Shopping 7
3. Gender and Emotion 11
4. Gender and Purchase: 16
5. Marital Status and Game: 21
6. Marital Status and Shopping 25
7. Marital Status and Emotion 30
8. Marital Status and Purchase 34
I. Statistical Inferences about Two Independent Populations
1. Gender and Game

● N (sample size): the number of people who participated in the survey

- n 0=16: There were 16 men participating in the survey
- n1=20: There were 20 women participating in the survey

● Mean: the average of average scores of “Game” factors related to gamification in livestream of each surveyed gender group
- x 0=3.425 : the average of average score of “Game” factors related to gamification in livestream of 16 surveyed men was 3.425,
which was a bit higher than women
x 0 ,1 + x 0 ,2 +...+ x 0 ,16 1+ 4.6+...+ 4.4
x 0= = =3.425
n0 16

- x 1=3.37 : the average of average score of “Game” factors related to gamification in livestream of 20 surveyed women was 3.37,
which was a bit lower than men
x1 , 1+ x 1 ,2 +...+ x 1 ,20 2.6+2+...+5
x 1= = =3.37
n1 20
 ● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Game” factors related to gamification
in livestream of each surveyed gender group from the mean
- s0 =0.9348: the average scores of “Game” factors related to gamification in livestream of 16 surveyed men were further from
the mean compared to the women
s0 =√ ❑

- s1=0.8467: the average scores of “Game” factors related to gamification in livestream of 20 surveyed women were more
clustered around the mean compared to the men’s
s1= √ ❑

Levene’s Test for Equality of Variances

2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0
2 2
H a :σ 0−σ 1 ≠ 0
Because the Sig. value of p-value of 0.847 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0
( x 0−x 1 )−( μ 0−μ1 )
t= 0.185
√❑

df = n 0+ n1−2=16+ 20−2=34

Because the Sig. (2-tailed) of the p-value of 0.854 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference = 0.55 ⇔ x 0−x 1= 3.425 - 3.37 =0.55

According to the 95% Confidence Interval of the Difference with the df = 34 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -0.5494 and 0.6594:

(x 0−x 1 )−t √ ❑

⇔ (3.425-3.37) - 2.04 √❑≤ μ 0−μ1 ≤ (3.425-3.37) +2.04 √❑

⇔ -0.5494 ≤ μ 0−μ1 ≤ 0.6594
 ● Equal variances not assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0

( x 0−x 1 )−( μ 0−μ1 )

t= = 0.1828
√❑

df = ¿ ¿ ≈ 30.708

Because the Sig. (2-tailed) of the p-value of 0.856 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null
hypothesis H 0 : μ0−μ1 =0, which meant that there was no difference between two population means.

Mean difference = 0.55⇔ x 0−x 1= 3.425 - 3.37 =0.55, which was really small as to be considered a difference

According to the 95% Confidence Interval of the Difference with the df = 30.708 ≈ 30 and α= 0.05, we are
95% confident that in the case of equal variance not assumed, the mean difference lies within -0.5494
and 0.6594:

(x 0−x 1 )−t √ ❑

⇔ (3.425-3.37) - 2.04 √❑≤ μ 0−μ1 ≤ (3.425-3.37) +2.04 √❑

⇔-0.5586 ≤ μ 0−μ1 ≤ 0.6686
 2. Gender and Shopping

● N (sample size): the number of people who participated in the survey

- n 0=16: There were 16 men participating in the survey
- n1=20: There were 20 women participating in the survey

● Mean: the average of average scores of “Shopping” factors related to gamification in livestream of each surveyed gender group
- x 0=3.2031: the average of average score of “Shopping” factors related to gamification in livestream of 16 surveyed men was
3.2031, which was a bit lower than women
x 0 ,1 + x 0 ,2 +...+ x 0 ,16 1+3.5+ ...+ 4.5
x 0= = =3.2031
n0 16

- x 1=3.3125: the average of average score of “Shopping” factors related to gamification in livestream of 20 surveyed women
was 3.3125, which was a bit higer than men
x1 , 1+ x 1 ,2 +...+ x 1 ,20 2.5+2+...+5
x 1= = =3.3125
n1 20
● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Shopping” factors related to
gamification in livestream of each surveyed gender group from the mean
- s0 =1.0417: the average scores of “Shopping” factors related to gamification in livestream of 16 surveyed men s0 =√❑

- s1=0.8954 : the average scores of “Shopping” factors related to gamification in livestream of 20 surveyed women s1= √ ❑
Levene’s Test for Equality of Variances

2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0

Because the Sig. value of p-value of 0.431 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0
( x 0−x 1 )−(μ 0−μ1 )
t= -0.3388
√❑
df = n 0+ n1−2=16+ 20−2=34
Because the Sig. (2-tailed) of the p-value of 0.737 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference = -0.1093 ⇔ μ0−μ 1= 3.2031 - 3.3125 = -0.1094 ≈ -0.1093

According to the 95% Confidence Interval of the Difference with the df = 34 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -0.7655 and 0.5468:

(x 0−x 1 )−t √ ❑

⇔ (3.2031-3.3125) - 2.04 √❑≤ μ 0−μ1 ≤ (3.2031-3.3125) +2.04 √❑

⇔ -0.7655 ≤ μ 0−μ1 ≤ 0.5468
● Equal variances not assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0

( x 0−x 1 )−(μ 0−μ1 )

t= = -0.333
√❑

df = ¿ ¿ ≈ 29.7634

Because the Sig. (2-tailed) of the p-value of 0.742 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null
hypothesis H 0 : μ0−μ1 =0, which meant that there was no difference between two population means.
 Mean difference = -0.10938⇔ μ0−μ 1= 3.2031 - 3.3125 = -0.1094 ≈ -0.1093

According to the 95% Confidence Interval of the Difference with the df = 29.7634 ≈ 30 and α= 0.05, we
are 95% confident that in the case of equal variance not assumed, the mean difference lies within -0.7804
and 0.5617:

(x 0−x 1 )−t √ ❑

⇔ (3.2031-3.3125) - 2.04 √❑≤ μ 0−μ1 ≤ (3.2031-3.3125) +2.04 √❑

⇔-0.7804 ≤ μ 0−μ1 ≤ 0.5617

3. Gender and Emotion

● N (sample size): the number of people who participated in the survey

- n 0=16: There were 16 men participating in the survey
- n1=20: There were 20 women participating in the survey

● Mean: the average of average scores of “Emotion” factors related to gamification in livestream of each surveyed gender group
 - x 0=3.1406 : the average of average score of “Emotion” factors related to gamification in livestream of 16 surveyed men was
3.1406, which was a bit higher than women
x 0 ,1 + x 0 ,2 +...+ x 0 ,16 1+3.5+ ...+ 4.25
x 0= = =3.1406
n0 16

- x 1=3.1375: the average of average score of “Emotion” factors related to gamification in livestream of 20 surveyed women was
3.1375, which was a bit lower than men
x1 , 1+ x 1 ,2 +...+ x 1 ,20 1.75+ 2+...+ 5
x 1= = =3.1375
n1 20
● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Emotion” factors related to
gamification in livestream of each surveyed gender group from the mean
- s0 =1.0838: the average scores of “Emotion” factors related to gamification in livestream of 16 surveyed men s0 =√❑

- s1=0.9982: the average scores of “Emotion” factors related to gamification in livestream of 20 surveyed women s1= √ ❑

Levene’s Test for Equality of Variances

 2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0

Because the Sig. value of p-value of 0.521 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0
( x 0−x 1 )−(μ 0−μ1 )
t= 0.0089
√❑
df = n 0+ n1−2=16+ 20−2=34

Because the Sig. (2-tailed) of the p-value of 0.993 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference = 0.00312 ⇔ μ0−μ 1= 3.1406- 3.1375= 0.00311 ≈ 0.00312

According to the 95% Confidence Interval of the Difference with the df = 34 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -0.7036 and 0.7099:

(x 0−x 1 )−t √ ❑

⇔ (3.1406-3.1375) - 2.04 √❑≤ μ 0−μ1 ≤ (3.1406-3.1375) + 2.04 √❑

⇔ -0.7036 ≤ μ 0−μ1 ≤ 0.7099
● Equal variances not assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0

( x 0−x 1 )−( μ 0−μ1 )

t= = 0.0088
√❑

df = ¿ ¿ ≈ 30.9968

Because the Sig. (2-tailed) of the p-value of 0.993 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null
hypothesis H 0 : μ0−μ1 =0, which meant that there was no difference between two population means.

Mean difference = 0.00312⇔ μ0−μ 1= 3.1406- 3.1375= 0.00311 ≈ 0.00312

According to the 95% Confidence Interval of the Difference with the df = 30.9968 ≈ 30 and α= 0.05, we
are 95% confident that in the case of equal variance not assumed, the mean difference lies within -0.7128
and 0.7191:

(x 0−x 1 )−t √ ❑

⇔ 3.1406-3.1375) - 2.04 √❑≤ μ 0−μ1 ≤ (3.1406-3.1375) + 2.04 √❑

⇔ –0.7128 ≤ μ 0−μ1 ≤ 0.7191

 4. Gender and Purchase

● N (sample size): the number of people who participated in the survey

- n 0=16: There were 16 men participating in the survey
- n1=20: There were 20 women participating in the survey

● Mean: the average of average scores of “Purchase” factors related to gamification in livestream of each surveyed gender group
- x 0=3.0208 : the average of average score of “Purchase” factors related to gamification in livestream of 16 surveyed men was
3.0208, which was lower than women
x 0 ,1 + x 0 ,2 +...+ x 0 ,16 1+1+...+ 4.67
x 0= = =3.0208
n0 16

- x 1=3.6166: the average of average score of “Purchase” factors related to gamification in livestream of 20 surveyed women was
3.6166, which was a bit lower than men
x1 , 1+ x 1 ,2 +...+ x 1 ,20 2+ 2+ ...+ 5
x 1= = =3.6166
n1 20
● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Purchase” factors related to
gamification in livestream of each surveyed gender group from the mean
 - s0 =1.2676: the average scores of “Purchase” factors related to gamification in livestream of 16 surveyed men s0 =√ ❑

- s1=0.9749: the average scores of “Purchase” factors related to gamification in livestream of 20 surveyed women s1= √ ❑

Levene’s Test for Equality of Variances

2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0

Because the Sig. value of p-value of 0.308 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0
 ( x 0−x 1 )−(μ 0−μ1 )
t= -4.8784
√❑
df = n 0+ n1−2=16+ 20−2=34

Because the Sig. (2-tailed) of the p-value of 0.120 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference =- 0.5958 ⇔ μ0−μ 1= 3.0208 - 3.6166= -0.5958

According to the 95% Confidence Interval of the Difference with the df = 34 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -1.3549 and 0.1632:

(x 0−x 1 )−t √ ❑

⇔ (3.0208 - 3.6166) - 2.04 √❑≤ μ 0−μ1 ≤ (3.0208 - 3.6166) + 2.04 √❑

⇔ -1.3549 ≤ μ 0−μ1 ≤ 0.1632
● Equal variances not assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0

( x 0−x 1 )−( μ 0−μ1 )

t= = -1.5489
√❑

df = ¿ ¿ ≈ 27.66
 Because the Sig. (2-tailed) of the p-value of 0.133 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null
hypothesis H 0 : μ0−μ1 =0, which meant that there was no difference between two population means.

Mean difference = -0.5958⇔ μ0−μ 1= 3.0208 - 3.6166= -0.5958

According to the 95% Confidence Interval of the Difference with the df = 30.9968 ≈ 30 and α= 0.05, we
are 95% confident that in the case of equal variance not assumed, the mean difference lies within -1.384
and 0.1925:

(x 0−x 1 )−t √ ❑

⇔ (3.0208 - 3.6166) - 2.04 √❑≤ μ 0−μ1 ≤ (3.0208 - 3.6166) + 2.04 √❑

⇔ -1.384 ≤ μ 0−μ1 ≤ 0.1925

5. Marital Status and Game

● N (sample size): the number of people who participated in the survey

- n 0=30: There were 30 people who are still single participating in the survey
- n1=4 : There were 4 people who are married participating in the survey

● Mean: the average of average scores of “Game” factors related to gamification in livestream of each surveyed gender group
- x 0=3.3333 : the average of average score of “Game” factors related to gamification in livestream of 30 single people was
3.3333, which was lower than the married’s
x 0 ,1 + x 0 ,2 +...+ x 0 ,30 1+ 4.6+...+5
x 0= = =3.3333
n0 30
- x 1=4.0 : the average of average score of “Game” factors related to gamification in livestream of 4 married people was 4.0,
which was higher than the single’s
x1 , 1+ x 1 ,2 +...+ x 1 ,4 3.8+ 4+3.2+5
x 1= = =4.0
n1 4
● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Game” factors related to gamification
in livestream of each surveyed gender group from the mean
- s0 =0.8840: the average scores of “Game” factors related to gamification in livestream of 30 single people s0 =√❑

- s1=0.7483: the average scores of “Game” factors related to gamification in livestream of 4 married people s1= √ ❑
Levene’s Test for Equality of Variances

2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0

Because the Sig. value of p-value of 0.308 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0
( x 0−x 1 )−(μ 0−μ1 )
t= -4.8784
√❑
df = n 0+ n1−2=16+ 20−2=34

Because the Sig. (2-tailed) of the p-value of 0.120 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference =- 0.5958 ⇔ μ0−μ 1= 3.0208 - 3.6166= -0.5958

According to the 95% Confidence Interval of the Difference with the df = 34 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -1.3549 and 0.1632:

(x 0−x 1 )−t √ ❑
⇔ (3.0208 - 3.6166) - 2.04 √❑≤ μ 0−μ1 ≤ (3.0208 - 3.6166) +2.04 √❑
⇔ -1.3549 ≤ μ 0−μ1 ≤ 0.1632
● Equal variances not assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0

( x 0−x 1 )−(μ 0−μ1 )

t= = -1.5489
√❑

df = ¿ ¿ ≈ 27.66

Because the Sig. (2-tailed) of the p-value of 0.133 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null
hypothesis H 0 : μ0−μ1 =0, which meant that there was no difference between two population means.

Mean difference = -0.5958⇔ μ0−μ 1= 3.0208 - 3.6166= -0.5958

According to the 95% Confidence Interval of the Difference with the df = 30.9968 ≈ 30 and α= 0.05, we
are 95% confident that in the case of equal variance not assumed, the mean difference lies within -1.384
and 0.1925:

(x 0−x 1 )−t √ ❑

⇔ (3.0208 - 3.6166) - 2.04 √❑≤ μ 0−μ1 ≤ (3.0208 - 3.6166) +2.04 √❑

⇔ -1.384 ≤ μ 0−μ1 ≤ 0.1925

6. Marital Status and Shopping

● N (sample size): the number of people who participated in the survey

- n 0=30: There were 30 people who are still single participating in the survey
- n1=4 : There were 4 people who are married participating in the survey

● Mean: the average of average scores of “Shopping” factors related to gamification in livestream of each surveyed gender group
- x 0=3.258 : the average of average score of “Shopping” factors related to gamification in livestream of 30 single people was
3.258, which was lower than the married’s
x 0 ,1 + x 0 ,2 +...+ x 0 ,16 1+ 4+...+5
x 0= = =3.258
n0 30
- x 1=3.75: the average of average score of “Shopping” factors related to gamification in livestream of 4 married people was
3.75, which was higher than the single’s
 x1 , 1+ x 1 ,2 +...+ x 1 ,4 3+ 4+ 4+3
x 1= = =3.75
n1 4
● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Shopping” factors related to
gamification in livestream of each surveyed gender group from the mean
- s0 =0.9342: the average scores of “Shopping” factors related to gamification in livestream of 30 single people s0 =√❑

- s1=0.9574 : the average scores of “Shopping” factors related to gamification in livestream of 4 married people s1= √ ❑

Levene’s Test for Equality of Variances

2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0

Because the Sig. value of p-value of 0.897 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed
 The hypotheses for t-test are: H 0 : μ0−μ1 =0
H a : μ0−μ1 ≠ 0
( x 0−x 1 )−( μ 0−μ1 )
t= -0.986
√❑
df = n 0+ n1−2=30+ 4−2=32

Because the Sig. (2-tailed) of the p-value of 0.331 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference =- 0.49167 ⇔ μ0−μ 1= 3.258 - 3.6166= -0.49167

According to the 95% Confidence Interval of the Difference with the df = 32 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -1.5069 and 0.5236:

(x 0−x 1 )−t √ ❑

⇔ ( 3.258 - 3.75) - 2.04 √❑≤ μ 0−μ1 ≤ ( 3.258 - 3.75) + 2.04 √❑

⇔ -1.5069 ≤ μ 0−μ1 ≤ 0.5236
● Equal variances not assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0

( x 0−x 1 )−(μ 0−μ1 )

t= = -0.967
√❑
 df = ¿ ¿ ≈ 27.66

Because the Sig. (2-tailed) of the p-value of 0.391 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null
hypothesis H 0 : μ0−μ1 =0, which meant that there was no difference between two population means.

Mean difference = -0.49167⇔ μ0−μ 1= 3.258 - 3.6166= -0.49167

According to the 95% Confidence Interval of the Difference with the df = 30.9968 ≈ 30 and α= 0.05, we
are 95% confident that in the case of equal variance not assumed, the mean difference lies within -1.384
and 0.1925:

(x 0−x 1 )−t √ ❑

⇔ ( 3.258 - 3.75) - 2.04 √❑≤ μ 0−μ1 ≤ ( 3.258 - 3.75) + 2.04 √❑

⇔ -1.19318 ≤ μ 0−μ1 ≤ 0.9484
 7. Marital Status and Emotion

● N (sample size): the number of people who participated in the survey

- n 0=30: There were 30 people who are still single participating in the survey
- n1=4 : There were 4 people who are married participating in the survey

● Mean: the average of average scores of “Emotion” factors related to gamification in livestream of each surveyed gender group
- x 0=3.066 : the average of average score of “Emotion” factors related to gamification in livestream of 30 single people was
3.066, which was lower than the married’s
x 0 ,1 + x 0 ,2 +...+ x 0 ,16 1+3.5+ ...+ 5
x 0= = =3.066
n0 30
- x 1=3.875: the average of average score of “Emotion” factors related to gamification in livestream of 4 married people was
3.875, which was higher than the single’s
x1 , 1+ x 1 ,2 +...+ x 1 ,4 3.5+ 4+3+ 5
x 1= = =3.875
n1 4
● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Emotion” factors related to
gamification in livestream of each surveyed gender group from the mean
- s0 =0.9933: the average scores of “Emotion” factors related to gamification in livestream of 30 single people s0 =√❑
 - s1=0.9574 : the average scores of “Emotion: factors related to gamification in livestream of 4 married people s1= √ ❑

Levene’s Test for Equality of Variances

2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0

Because the Sig. value of p-value of 0.572 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0
( x 0−x 1 )−( μ 0−μ1 )
t= -1. 548
√❑
 df = n 0+ n1−2=30+ 4−2=32

Because the Sig. (2-tailed) of the p-value of 0.132 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference =- 0.8083 ⇔ μ0−μ 1= 3.066 - 3.875= -0.809 ≈ -0.808

According to the 95% Confidence Interval of the Difference with the df = 32 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -1.8721 and 0.2554:

(x 0−x 1 )−t √ ❑

⇔ ( 3.066 - 3.875) - 2.04 √❑≤ μ 0−μ1 ≤ ( 3.258 - 3.75) + 2.04 √❑

⇔ -1.5069 ≤ μ 0−μ1 ≤ 0.5236
● Equal variances not assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0

( x 0−x 1 )−(μ 0−μ1 )

t= = -0.967
√❑

df = ¿ ¿ ≈ 27.66

Because the Sig. (2-tailed) of the p-value of 0.154 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null
hypothesis H 0 : μ0−μ1 =0, which meant that there was no difference between two population means.
 Mean difference = -0.8083⇔ μ0−μ 1= 3.066 - 3.875= -0.809 ≈ -0.808

According to the 95% Confidence Interval of the Difference with the df = 30.9968 ≈ 30 and α= 0.05, we
are 95% confident that in the case of equal variance not assumed, the mean difference lies within -2.076
and 0.4595:

(x 0−x 1 )−t √ ❑

⇔ ( 3.066-3.875) - 2.04 √❑≤ μ 0−μ1 ≤ ( 3.066-3.875) + 2.04 √❑

⇔ -2.076 ≤ μ 0−μ1 ≤ 0.4595

8. Marital Status and Purchase

● N (sample size): the number of people who participated in the survey

- n 0=30: There were 30 people who are still single participating in the survey
- n1=4 : There were 4 people who are married participating in the survey

● Mean: the average of average scores of “Purchase” factors related to gamification in livestream of each surveyed gender group
 - x 0=3.188 : the average of average score of “Purchase” factors related to gamification in livestream of 30 single people was
3.188, which was lower than the married’s
x 0 ,1 + x 0 ,2 +...+ x 0 ,16 1+1+...+5
x 0= = =3.188
n0 30
- x 1=4.416 : the average of average score of “Purchase” factors related to gamification in livestream of 4 married people was
4.416, which was higher than the single’s
x1 , 1+ x 1 ,2 +...+ x 1 ,4 4+ 4+ 4.67+5
x 1= = =4.416
n1 4
● Std. Deviation: the measure of variability (spread, dispersion, …) of average scores of “Purchase” factors related to
gamification in livestream of each surveyed gender group from the mean
- s0 =1.1536: the average scores of “Purchase” factors related to gamification in livestream of 30 single people s0 =√❑

- s1=0.5: the average scores of “Purchase” factors related to gamification in livestream of 4 married people s1= √ ❑

Levene’s Test for Equality of Variances

2 2
The hypotheses for Levene’s test are: H 0 :σ 0 −σ 1=0
Because the Sig. value of p-value of 0.164 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis (equal
variances assumed), which meant that we will rely on the first row of output when we look at the results for the actual Independent Samples t
Test (under the heading t-test for Equality of Means)

t-test for Equality of Means

● Equal variances assumed

The hypotheses for t-test are: H 0 : μ0−μ1 =0

H a : μ0−μ1 ≠ 0
( x 0−x 1 )−( μ 0−μ1 )
t= -2.080
√❑
df = n 0+ n1−2=30+ 4−2=32

Because the Sig. (2-tailed) of the p-value of 0.046 was larger than all of the values (0.05, 0.01, 0.001), we accepted the null hypothesis:
H 0 : μ0−μ1 =0 which meant that there was no difference between two population means.

Mean difference =-1.227 ⇔ μ0−μ 1= 3.188-4.416 = -1.228 ≈ -1.227

According to the 95% Confidence Interval of the Difference with the df = 32 and, we are 95% confident that in the case of equal variance
assumed, the mean difference lies within -2.43 and -0.255:

(x 0−x 1 )−t √ ❑

⇔ ( 3.188-4.416 ) - 2.04 √❑≤ μ 0−μ1 ≤ ( 3.188-4.416 ) + 2.04 √❑

⇔ -2.43 ≤ μ 0−μ1 ≤ -0.255
● Equal variances not assumed
 The hypotheses for t-test are: H 0 : μ0−μ1 =0
H a : μ0−μ1 ≠ 0

( x 0−x 1 )−(μ 0−μ1 )

t= = -3.756
√❑

df = ¿ ¿ ≈ 8.336

Because the Sig. (2-tailed) of the p-value of 0.005 which is smaller than all of the values (0.05, 0.01, 0.001), it is accepted to infer the
population’s regression line from the estimated regression line of the sample

Mean difference = -1.227 ⇔ μ0−μ 1= 3.188-4.416 = -1.228 ≈ -1.227

According to the 95% Confidence Interval of the Difference with the df = 8.336 and α= 0.05, we are 95% confident that in the case of
equal variance not assumed, the mean difference lies within -1.9763 and -479:

(x 0−x 1 )−t √ ❑

⇔ ( 3.188 - 4.416) - 2.04 √❑≤ μ 0−μ1 ≤ ( 3.188 - 4.416) + 2.04 √❑

⇔ -1.9763 ≤ μ 0−μ1 ≤−0.479