Professional Documents
Culture Documents
Example 1
A group of researchers conducted a study to determine if there is a relationship between
playing hours and study hours. Twenty high school learners involved in the study. The researcher
assumes that playing hours affect the study hours of learners.
Table 1
Participants Playing Hours Study Hours
1 11 10
2 2 4
3 24 23
4 18 3
5 19 21
6 25 17
7 12 24
8 15 20
9 5 2
10 6 25
11 14 22
12 21 15
13 7 2
14 20 9
15 8 7
16 17 19
17 16 8
18 13 11
19 1 1
20 23 26
Research Hypothesis
The null hypothesis states that there is no correlation between playing hours and study
hours. The research hypothesis states that there is a correlation between playing hours and study
hours .
H0: ρs = 0
H1: ρs ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the variables using an ordinal, or rank, procedure. Therefore, we
will convert the values in each variable to ordinal data. Additionally, we will be comparing the
two variables, playing hours and study hours. Since we are comparing two variables in which
one or both are measured on an ordinal scale, we will use Spearman rank-order correlation.
Compute the Test Statistics
First, rank the scores for each variable separately. Rank them from the lowest score to the
highest score to form an ordinal distribution for each variable.
Table 2.
Original Scores Ranked Scores
Participants Playing Hours Study Hours Playing Hours Study Hours
1 11 10 7 8
2 2 4 2 4
3 24 23 19 17
4 18 3 14 3
5 19 21 15 15
6 25 17 20 12
7 12 24 8 18
8 15 20 11 14
9 5 2 3 2
10 6 25 4 19
11 14 22 10 16
12 21 15 17 11
13 7 13 5 10
14 20 9 16 7
15 8 7 6 5
16 17 19 13 13
17 16 8 12 6
18 13 11 9 9
19 1 1 1 1
20 23 26 18 20
6 ∑ Di
2
R s=1−
n ( n2−1 )
6(748)
R s=1−
20 ( 202−1 )
4488
R s=1−
7980
R s=1−0.56
R s=0.44
Research Hypothesis
The null hypothesis states that there is no correlation between alcohol consumptions and
tardiness. The research hypothesis states that there is a correlation between alcohol consumptions
and tardiness..
H0: ρs = 0
H1: ρs ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the variables using an ordinal, or rank, procedure. Therefore, we
will convert the values in each variable to ordinal data. Additionally, we will be comparing the
two variables, alcohol consumptions and tardiness.. Since we are comparing two variables in
which one or both are measured on an ordinal scale, we will use Spearman rank-order
correlation.
Compute the Test Statistics
First, rank the scores for each variable separately. Rank them from the lowest score to the
highest score to form an ordinal distribution for each variable.
Table 5.
Original Scores Ranked Scores
Participants Alcohol Number of Alcohol Number of
Consumptions Absences Consumptions Absences
(month) (month) (month) (month)
1 22 20 16 15
2 19 12 14 10
3 12 6 9 6
4 1 11 2 9
5 30 1 20 1
6 17 28 12 18
7 18 23 13 16
8 14 30 11 20
9 4 17 5 14
10 8 15 8 13
11 23 2 17 2
12 2 9 3 7
13 13 10 10 8
14 3 4 4 4
15 6 29 7 19
16 26 5 19 5
17 24 3 18 3
18 5 13 6 11
19 0 24 1 17
20 21 14 15 12
6 ∑ Di
2
R s=1−
n ( n −1 )
2
6(1768)
R s=1−
20 ( 202−1 )
10608
R s=1−
7980
R s=1−1.33
R s=−0.33
Research Hypothesis
The null hypothesis states that there is no correlation between learner’s performance and
general average. The research hypothesis states that there is a correlation between learner’s
performance and general average.
H0: ρs = 0
H1: ρs ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the variables using an ordinal, or rank, procedure. Therefore, we
will convert the values in each variable to ordinal data. Additionally, we will be comparing the
two variables, learner’s performance and general average. Since we are comparing two variables
in which one or both are measured on an ordinal scale, we will use Spearman rank-order
correlation.
Compute the Test Statistics
First, rank the scores for each variable separately. Rank them from the lowest score to the
highest score to form an ordinal distribution for each variable.
Table 8.
Original Scores Ranked Scores
Participants Number of General Number of General
Days Present Average Days Present Average
(academic year) (academic year)
1 125 78 7 4
2 112 83 5 8
3 100 85 1 10
4 189 97 16 20
5 147 93 11 16
6 192 96 17 19
7 113 75 6 2
8 194 81 19 7
9 104 74 3 1
10 131 95 9 18
11 142 90 10 15
12 175 94 15 17
13 197 87 20 12
14 159 84 13 9
15 126 76 8 3
16 166 86 14 11
17 111 80 4 5
18 101 88 2 13
19 193 82 18 7
20 157 89 12 14
6 ∑ Di
2
R s=1−
n ( n −1 )
2
6(729)
R s=1−
20 ( 202−1 )
4374
R s=1−
7980
R s=1−0.55
R s=0.45
Research Hypothesis
The null hypothesis states that there is no correlation between age of pregnancy and
number of children. The research hypothesis states that there is a correlation between age of
pregnancy and number of children.
H0: ρs = 0
H1: ρs ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the variables using an ordinal, or rank, procedure. Therefore, we
will convert the values in each variable to ordinal data. Additionally, we will be comparing the
two variables, age of pregnancy and number of children. Since we are comparing two variables
in which one or both are measured on an ordinal scale, we will use Spearman rank-order
correlation.
Compute the Test Statistics
First, rank the scores for each variable separately. Rank them from the lowest score to the
highest score to form an ordinal distribution for each variable.
Table 11.
Original Scores Ranked Scores
Participants Age of Number of Age of Number of
Pregnancy Children Pregnancy Children
1 18 20 2 20
2 29 1 10 1
3 31 19 12 19
4 42 4 19 4
5 20 13 4 13
6 38 14 16 14
7 16 12 1 12
8 43 5 20 5
9 22 9 7 9
10 40 7 17 7
11 21 16 5 16
12 28 17 9 17
13 36 3 15 3
14 41 11 18 11
15 34 15 13 15
16 23 10 7 10
17 35 18 14 18
18 19 6 3 6
19 27 2 8 2
20 30 8 11 8
6 ∑ Di
2
R s=1−
n ( n2−1 )
6(1675)
R s=1−
20 ( 20 −1 )
2
10050
R s=1−
7980
R s=1−1.26
R s=−0.26
Research Hypothesis
The null hypothesis states that there is no correlation between age and weekly allowance.
The research hypothesis states that there is a correlation between age and weekly allowance
H0: ρs = 0
H1: ρs ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the variables using an ordinal, or rank, procedure. Therefore, we
will convert the values in each variable to ordinal data. Additionally, we will be comparing the
two variables, age and weekly allowance. Since we are comparing two variables in which one or
both are measured on an ordinal scale, we will use Spearman rank-order correlation.
Compute the Test Statistics
First, rank the scores for each variable separately. Rank them from the lowest score to the
highest score to form an ordinal distribution for each variable.
Table 14.
Original Scores Ranked Scores
Participants Age Weekly Age Weekly
Allowance Allowance
1 23 280 10 1
2 26 401 13 5
3 12 770 4 11
4 30 777 16 12
5 18 340 8 3
6 27 536 14 7
7 11 1379 3 18
8 25 693 12 9
9 31 1394 17 19
10 16 1142 6 15
11 24 1267 11 16
12 35 443 20 6
13 13 824 5 13
14 9 1057 1 14
15 34 291 19 2
16 17 368 7 4
17 33 1291 18 17
18 28 755 15 10
19 10 557 2 8
20 22 1445 9 20
6 ∑ Di
2
R s=1−
n ( n2−1 )
6(1538)
R s=1−
20 ( 20 −1 )
2
9228
R s=1−
7980
R s=1−1.16
R s=−0.16
Research Hypothesis
The null hypothesis states that there is no correlation between gender and number of
exes. The research hypothesis states that there is a correlation between gender and number of
exes.
H0: ρpb = 0
H1: ρpb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Gender is a discrete
dichotomous variable and number of exes is an interval scale variable.
Compute the Test Statistics
Compute the standard deviation of all values from the interval scale data. It is helpful to
organize the data if we put in tables.
Table 2.
Participant Gender Number of Exes x i−x (x−x )2
1 M 14 5.1 26.01
2 M 7 -1.9 3.61
3 M 9 0.1 0.01
4 M 3 -5.9 34.81
5 M 5 -3.9 15.21
6 M 11 2.1 4.41
7 M 11 2.1 4.41
8 M 15 6.1 37.21
9 M 2 -6.9 47.61
10 M 15 6.1 37.21
11 F 14 5.1 26.01
12 F 15 6.1 37.21
13 F 15 6.1 37.21
14 F 4 -4.9 24.01
15 F 6 -2.9 8.41
16 F 8 -0.9 0.81
17 F 2 -6.9 47.61
18 F 9 0.1 0.01
19 F 12 3.1 9.61
20 F 1 -7.9 62.41
∑ x i=¿178 ∑ ( x i−x)2=463.8
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=178 ÷ 20
x=8.9
S x=
√ ∑ ( x i−x)2 =
n−1 √ 18.9
20−1
S x =0.9974
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean male’s numbers of exes were
x p = ∑ x p ÷ nM
= 9.2
The mean females’ numbers of exes were
x q = ∑ xq ÷ n F
= 8.6
The males’ proportion was
P p=nM ÷ n
P p=10 ÷ 20
P p=0.5
Pq =10÷ 20
Pq =0.5
Compute the point-biserial correlation coefficient using the values computed earlier:
x p−x q
r pb=
sx
√ P p Pq
9.2−8.6
r pb=
0.9974
√( 0.5 ) (0.5)
r pb =0.30
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, table requires that the degrees of freedom be known. Since df =
n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since, we are conducting a two-tailed test
and α = 0.05, the critical value is 0.444.
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtain value is |R pb|=
0.30. If the critical value is less than or equal to the obtain value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater than the absolute value of the obtained value, we do
not reject the null hypothesis.
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no significant correlation
between gender and number of exes.
Reporting the Results
A group of researchers conduct a study to determine the relationship of gender and
number exes. Ten males and ten females participated in the investigation. A point-biserial
correlation produced less significant results (rpb(18), 0.30, p > 0.05. These data suggest that there
is no relationship between gender and number of exes.
Example 2
A group of researchers conducted a study to determine the relationship between gender
and speed in driving motorcycle. Thirty high school students were involved in the study, 20
males and 10 female. Speed was measured though kilometer per hour basis. Gender is the
discrete dichotomous variable and speed is an interval/scale variable. The researchers assume
that there is a strong relationship between gender and speed in driving motorcycle. Therefore, the
researchers will be using point-biserial correlation.
Table 3.
Participant Gender Speed (km/h)
1 M 197
2 M 76
3 M 151
4 M 148
5 M 125
6 M 66
7 M 141
8 M 74
9 M 134
10 M 198
11 M 63
12 M 121
13 M 72
14 M 142
15 M 184
16 M 105
17 M 183
18 M 62
19 M 168
20 M 68
21 F 121
22 F 146
23 F 144
24 F 109
25 F 67
26 F 66
27 F 66
28 F 197
29 F 184
30 F 70
Research Hypothesis
The null hypothesis states that there is no correlation between gender and speed in
driving. The research hypothesis states that there is a correlation between gender and monthly
salary gender and speed in driving.
H0: ρpb = 0
H1: ρpb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Gender is a discrete
dichotomous variable and speed is an interval scale variable.
Compute the Test Statistics
Compute the standard deviation of all values from the interval scale data. It is helpful to
organize the data if we put in tables.
Table 4.
Participant Gender Speed (km/h) x i−x (x−x )2
1 M 197 75.4 5685.16
2 M 76 -45.6 2079.36
3 M 151 29.4 864.36
4 M 148 26.4 696.96
5 M 125 3.4 11.56
6 M 66 -55.6 3091.36
7 M 141 19.4 376.36
8 M 74 -47.6 2265.76
9 M 134 12.4 153.76
10 M 198 76.4 5836.96
11 M 63 -58.6 3433.96
12 M 121 -0.6 0.36
13 M 72 -49.6 2460.16
14 M 142 20.4 416.16
15 M 184 62.4 3893.76
16 M 105 -16.6 275.56
17 M 183 61.4 3769.96
18 M 62 -59.6 3552.16
19 M 168 46.4 2152.96
20 M 68 -53.6 2872.96
21 F 121 -0.6 0.36
22 F 146 24.4 595.36
23 F 144 22.4 501.76
24 F 109 -12.6 158.76
25 F 67 -54.6 2981.16
26 F 66 -55.6 3091.36
27 F 66 -55.6 3091.36
28 F 197 75.4 5685.16
29 F 184 62.4 3893.76
30 F 70 -51.6 2662.56
∑ x i=3648 ∑ ( xi −x ) 2=66551.2
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=3648 ÷ 30
x=121.6
S x=
√ ∑ ( x i−x)2 =
n−1 √ 66551.2
30−1
S x =151.5
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean males’ driving speed was
x p = ∑ x p ÷ nM
= 123.9
The mean females’ driving speed was
x q = ∑ xq ÷ n F
= 117
The males’ proportion was
P p=nM ÷ n
P p=20÷ 30
P p=0.67
Pq =10÷ 20
Pq =0.33
Compute the point-biserial correlation coefficient using the values computed earlier:
x p−x q
r pb=
sx
√ P p Pq
123.9−117
r pb=
151.5
√ ( 0.67 ) (0.33)
r pb =0.02
z =r pb [ √ n−1 ]
¿
z ¿ =0.02 [ √ 30−1 ]
z ¿ =0.11
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Since the data is large, we can determine the correlation significant through critical
region of z-scores. For two-tailed testing with α = 0.05, we must reject the null hypothesis if
¿
−1.96 ≤ z ≤ 1.96 .
Comparing the Obtain Value with Critical Value
Notice that z* is in the positive tail of the distribution (0.11 < 1.96). Therefore, we should
reject the null hypothesis. This suggests that the correlation between gender and speed in driving
motorcyle is highly significant.
Interpret the Results
We do reject the null hypothesis, suggesting that there is a significant correlation between
gender and speed in driving
Reporting the Results
A group of researchers conduct a study to determine the relationship between gender and
speed in driving motorcyle. Twenty males and 10 females participated in the investigation. A
point-biserial correlation produced significant results (rpb(28), 0.02, p < 0.05. These data suggest
that there is a significant relationship between gender and speed in driving. Moreover, the mean
on male’s driving speed is ( x M =¿ 123.9), while the means on females driving speed ( x F = 117).
These suggest that males have a higher driving speed than females.
Example 3
A group of researchers conducted a study to determine the relationship between the
perceptions of random people and age. Perception was measured through yes/no basis. Twenty-
five random people were involved in the study, fifteen random people answered yes, and 10
random people answered no. They were asked if they believed in the existence of Deity.
Perception is the discrete dichotomous variable and age is an interval/scale variable. Therefore,
the researchers will be using point-biserial correlation.
Table 5.
Participant Perceptions in Deity Age
(Yes/No)
1 YES 62
2 YES 31
3 YES 41
4 YES 51
5 YES 44
6 YES 67
7 YES 44
8 YES 60
9 YES 69
10 YES 23
11 YES 61
12 YES 69
13 YES 30
14 YES 42
15 YES 27
16 NO 33
17 NO 60
18 NO 53
19 NO 48
20 NO 49
21 NO 20
22 NO 41
23 NO 37
24 NO 26
25 NO 56
Research Hypothesis
The null hypothesis states that there is no correlation between age and their perceptions in
the existence of Deity. The research hypothesis states that there is a age and their perceptions in
the existence of Deity.
H0: ρpb = 0
H1: ρpb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Perception is a
discrete dichotomous variable and age is an interval scale variable.
Compute the Test Statistics
Compute the standard deviation of all values from the interval scale data. It is helpful to
organize the data if we put in tables.
Table 6.
Participa Perceptions in Deity Age x i−x (x−x )2
nt (Yes/No)
1 YES 62 16.24 263.7376
2 YES 31 -14.76 217.8576
3 YES 41 -4.76 22.6576
4 YES 51 5.24 27.4576
5 YES 44 -1.76 3.0976
6 YES 67 21.24 451.1376
7 YES 44 -1.76 3.0976
8 YES 60 14.24 202.7776
9 YES 69 23.24 540.0976
10 YES 23 -22.76 518.0176
11 YES 61 15.24 232.2576
12 YES 69 23.24 540.0976
13 YES 30 -15.76 248.3776
14 YES 42 -3.76 14.1376
15 YES 27 -18.76 351.9376
16 NO 33 -12.76 162.8176
17 NO 60 14.24 202.7776
18 NO 53 7.24 52.4176
19 NO 48 2.24 5.0176
20 NO 49 3.24 10.4976
21 NO 20 -25.76 663.5776
22 NO 41 -4.76 22.6576
23 NO 37 -8.76 76.7376
24 NO 26 -19.76 390.4576
25 NO 56 10.24 104.8576
∑ x i=11 44 ∑ ( xi −x ) 2=5328.56
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=1144 ÷ 25
x=45.76
S x=
√ ∑ ( x i−x)2 =
n−1 √ 5328.56
25−1
S x =14.9
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean age of random people who answered yes was
x p = ∑ x p ÷ nY
= 48.07
The mean age of random people who answered no was
x q=∑ xq ÷ n N
= 42.3
The yes proportion was
P p=nY ÷ n
P p=15 ÷ 25
P p=0.6
Pq =10÷ 20 5
Pq =0.4
Compute the point-biserial correlation coefficient using the values computed earlier:
x p−x q
r pb=
sx
√ P p Pq
48.07−42.3
r pb=
14.9
√ ( 0.6 ) (0.4 )
r pb=0.189
z ¿ =r p b [ √n−1 ]
z =0.19 [ √ 25−1 ]
¿
¿
z =0.93
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Since the data is large, we can determine the correlation significant through critical
region of z-scores. For two-tailed testing with α = 0.05, we must reject the null hypothesis if
−1.96 ≤ z¿ ≤ 1.96 .
Comparing the Obtain Value with Critical Value
Notice that z* is in the positive tail of the distribution (0.93 < 1.96). Therefore, we should
reject the null hypothesis. This suggests that the correlation between age and perceptions of
random people to the existence in Deity is highly significant.
Interpret the Results
We do reject the null hypothesis, suggesting that there is a significant correlation between
age and random people’s perceptions in the existence of Deity.
Reporting the Results
A group of researchers conduct a study to determine the relationship between age and
random people’s perceptions to the existence in Deity. Fifteen males and ten females participated
in the investigation. A point-biserial correlation produced significant results (r pb(23), 0.189, p <
0.05. These data suggest that there is a significant relationship between age and perceptions of
random people. Moreover, the mean age of people who answered yes ( x Y =¿ 48.07), while the
means age of people who answered no is ( x N = 42.3).
Example 4
A group of researchers conducted a study to determine the relationship of gender and
hours in listening to music. There are 20 college students involved in the study, 11 males and 9
females. Gender is discrete dichotomous variable and hours in listening to music is an
interval/scale variable. Therefore, the researchers will be using point-biserial correlation.
Table 7.
Participants Gender Hours in Listening to Music
1 M 3
2 M 2
3 M 1
4 M 5
5 M 5
6 M 4
7 M 3
8 M 4
9 M 1
10 M 4
11 M 5
12 F 5
13 F 3
14 F 4
15 F 3
16 F 4
17 F 2
18 F 4
19 F 3
20 F 1
Research Hypothesis
The null hypothesis states that there is no correlation between gender and hours in
listening to music. The research hypothesis states that there is a correlation between gender and
hours in listening to music
H0: ρpb = 0
H1: ρpb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Gender is a discrete
dichotomous variable and hours in listening to music is an interval scale variable.
Compute the Test Statistics
Compute the standard deviation of all values from the interval scale data. It is helpful to
organize the data if we put in tables.
Table 8.
Participant Gender Hours in x i−x (x−x )2
Listening to
Music
1 M 3 -0.3 0.09
2 M 2 -1.3 1.69
3 M 1 -2.3 5.29
4 M 5 1.7 2.89
5 M 5 1.7 2.89
6 M 4 0.7 0.49
7 M 3 -0.3 0.09
8 M 4 0.7 0.49
9 M 1 -2.3 5.29
10 M 4 0.7 0.49
11 M 5 1.7 2.89
12 F 5 1.7 2.89
13 F 3 -0.3 0.09
14 F 4 0.7 0.49
15 F 3 -0.3 0.09
16 F 4 0.7 0.49
17 F 2 -1.3 1.69
18 F 4 0.7 0.49
19 F 3 -0.3 0.09
20 F 1 -2.3 5.29
∑ i
x =¿66 ∑ (x i−x)2=34.2
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=66 ÷ 20
x=3.3
S x=
√ ∑ ( x i−x)2 =
n−1 √ 34.2
20−1
S x =1.34
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean male’s hours in listening to music were
x p = ∑ x p ÷ nM
= 3.4
The mean females’ hours in listening to music were
x q = ∑ xq ÷ n F
= 1.5
The males’ proportion was
P p=nM ÷ n
P p=11 ÷ 20
P p=0.55
Pq =9 ÷20
Pq =0. 4 5
Compute the point-biserial correlation coefficient using the values computed earlier:
x p−x q
r pb=
sx
√ P p Pq
3.4−1.5
r pb=
1.34
√( 0.5 5 ) (0. 4 5)
r pb =0.74
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, table requires that the degrees of freedom be known. Since df =
n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since, we are conducting a two-tailed test
and α = 0.05, the critical value is 0.444.
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtain value is |R pb|=
0.74. If the critical value is less than or equal to the obtain value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is less than the absolute value of the obtained value, we do
reject the null hypothesis.
Interpret the Results
We do reject the null hypothesis, suggesting that there is a significant correlation between
gender and hours in listening to music.
Reporting the Results
A group of researchers conduct a study to determine the relationship of gender and hours
in listening to music. Eleven males and nine females participated in the investigation. A point-
biserial correlation produced high significant results (rpb(18), 0.74, p < 0.05. These data suggest
that there is a relationship between gender and hours in listening to music. Moreover, the mean
males hours in listening to music was (xM=3.4) and the mean females hours in listening to music
was (xF=1.5. These suggest that males have a higher hours in listening to music than females.
Example 5
Table 9.
Participants Gender Number of Years in College
1 M 4
2 M 5
3 M 5
4 M 5
5 M 8
6 M 6
7 M 6
8 M 5
9 M 8
10 M 7
11 M 5
12 M 7
13 M 4
14 M 7
15 M 4
16 F 6
17 F 6
18 F 4
19 F 7
20 F 6
Research Hypothesis
The null hypothesis states that there is no correlation between gender and number of
years studied in college. The research hypothesis states that there is a correlation gender and
number of years studied in college.
H0: ρpb = 0
H1: ρpb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Gender is a discrete
dichotomous variable and number of years studied in college is an interval scale variable.
Compute the Test Statistics
Compute the standard deviation of all values from the interval scale data. It is helpful to
organize the data if we put in tables.
Table 10.
Participant Gender Number of Years x i−x (x−x )2
in College
1 M 4 0.0625 0.0625
2 M 5 0.0625 0.0625
3 M 5 0.0625 0.0625
4 M 5 0.0625 0.0625
5 M 8 0.0625 0.0625
6 M 6 0.0625 0.0625
7 M 6 0.0625 0.0625
8 M 5 0.0625 0.0625
9 M 8 0.0625 0.0625
10 M 7 0.0625 0.0625
11 M 5 0.0625 0.0625
12 M 7 0.0625 0.0625
13 M 4 0.0625 0.0625
14 M 7 0.0625 0.0625
15 M 4 0.0625 0.0625
16 F 6 0.0625 0.0625
17 F 6 0.0625 0.0625
18 F 4 0.0625 0.0625
19 F 7 0.0625 0.0625
20 F 6 0.0625 0.0625
∑ x i=¿115 ∑ (x i−x)2=31.75
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=115 ÷ 20
x=5.75
S x=
√ ∑ ( x i−x)2 =
n−1 √ 31.75
20−1
S x =1.29
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean male’s numbers of years studied in college were
x p = ∑ x p ÷ nM
= 5.73
The mean females’ number of years studied in college was
x q = ∑ xq ÷ n F
x p=6+ 6+ 4+7+ 6÷ 5
= 5.8
The males’ proportion was
P p=nM ÷ n
P p=15 ÷ 20
P p=0.75
Pq =5÷ 20
Pq =0.25
Compute the point-biserial correlation coefficient using the values computed earlier:
x p−x q
r pb=
sx
√ P p Pq
5.73−5.8
r pb=
1.29
√ ( 0.7 5 ) (0.2 5)
r pb=−0.023
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, table requires that the degrees of freedom be known. Since df =
n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since, we are conducting a two-tailed test
and α = 0.05, the critical value is 0.444.
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtain value is |R pb|=
0.023. If the critical value is less than or equal to the obtain value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater than the absolute value of the obtained value, we do
not reject the null hypothesis.
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no significant correlation
between gender and number of years studied in college.
Reporting the Results
Research Hypothesis
The null hypothesis states that there is no correlation between achievements and
performance ratings. The research hypothesis states that there is a correlation between
performance ratings and achievements.
H0: ρb = 0
H1: ρb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Achievement is a
continuous dichotomous variable and performance rating is an interval scale variable.
Table 2.
Participant Achievements Performance x i−x (x−x )2
Ratings (60-
100%)
1 PROMOTED 62 -18.45 340.4025
2 PROMOTED 67 -13.45 180.9025
3 PROMOTED 87 6.55 42.9025
4 PROMOTED 76 -4.45 19.8025
5 PROMOTED 89 8.55 73.1025
6 PROMOTED 91 10.55 111.3025
7 PROMOTED 69 -11.45 131.1025
8 PROMOTED 95 14.55 211.7025
9 PROMOTED 64 -16.45 270.6025
10 PROMOTED 68 -12.45 155.0025
11 NOT PROMOTED 99 18.55 344.1025
12 NOT PROMOTED 96 15.55 241.8025
13 NOT PROMOTED 84 3.55 12.6025
14 NOT PROMOTED 63 -17.45 304.5025
15 NOT PROMOTED 88 7.55 57.0025
16 NOT PROMOTED 91 10.55 111.3025
17 NOT PROMOTED 65 -15.45 238.7025
18 NOT PROMOTED 94 13.55 183.6025
19 NOT PROMOTED 62 -18.45 340.4025
20 NOT PROMOTED 99 18.55 344.1025
∑ x i=¿1609 ∑ ( xi −x ) 2=3714.95
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=1609 ÷ 20
x=80.45
S x=
√ ∑ ( x i−x)2 =
n−1 √ 3714 .95
20−1
S x =13.98
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean performance ratings of promoted employee was
x p = ∑ x p ÷ nP
= 76.8
The mean performance ratings of not promoted employee was
x q=∑ xq ÷ n NP
= 84.1
The proportion of employee who were promoted was
P p=nP ÷ n
P p=10 ÷ 20
P p=0.5
Pq =10÷ 20
Pq =0.5
We will compute the value using critical value tables for normal distribution. It provide the z-
score at the point Pp and Pq, z = 0.00
2
1 −z 2
y= e
√2 π
2
1 0.00 2
y= e
√2 π
y=0.4
Compute, the biserial correlation coefficient using the values computed earlier:
[
rb =
sx ]
x p −x q P p P q
y
rb =
[ 13.98 ]
76.8−84.1 (0.5)(0.5)
0.4
r b =−0.33
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since
df = n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since we are conducting a two-tailed
test and α = 0.05, the critical value is 0.444
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtained value is |R b|=
0.33. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater the absolute value of the obtained value, we do not
reject the null hypothesis
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no significant correlation
between achievements and performance ratings..
Reporting the Results
A group of researchers conduct a study to compare employee’s achievement and
performance ratings. Ten employees were promoted (nP = 10) and 10 employees weren’t
promoted (nNP = 10). A biserial correlation produced no significant results (rb(18) = -0.33, p >
0.05). The data suggest that there is no significant correlation between achievements and
performance ratings
Example 2
A group of researchers wished to know the relationship between SCASTS results and
general average of learners. SCASTS result will be measured through pass/fail basis. Twenty
freshmen students were involved in the study, where twelve of them pass the SCAST
examination, while those eight failed. SCASTS result is a continuous dichotomous variable and
general average is an interval/scale variable. Therefore, the researchers will be using biserial
correlation.
Table 3.
Participant SCASTS Result General Average
1 PASS 89
2 PASS 83
3 PASS 81
4 PASS 93
5 PASS 89
6 PASS 76
7 PASS 98
8 PASS 86
9 PASS 75
10 PASS 98
11 PASS 95
12 PASS 99
13 FAIL 77
14 FAIL 93
15 FAIL 91
16 FAIL 93
17 FAIL 92
18 FAIL 95
19 FAIL 86
20 FAIL 91
Research Hypothesis
The null hypothesis states that there is no correlation between SCASTS results and
general average. The research hypothesis states that there is a correlation between SCASTS
results and general average.
H0: ρb = 0
H1: ρb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. SCASTS result is a
continuous dichotomous variable and general average is an interval scale variable.
Table 4.
Participant SCASTS Result General x i−x (x−x )
2
Average
1 PASS 89 0 0
2 PASS 83 -6 36
3 PASS 81 -8 64
4 PASS 93 4 16
5 PASS 89 0 0
6 PASS 76 -13 169
7 PASS 98 9 81
8 PASS 86 -3 9
9 PASS 75 -14 196
10 PASS 98 9 81
11 PASS 95 6 36
12 PASS 99 10 100
13 FAIL 77 -12 144
14 FAIL 93 4 16
15 FAIL 91 2 4
16 FAIL 93 4 16
17 FAIL 92 3 9
18 FAIL 95 6 36
19 FAIL 86 -3 9
20 FAIL 91 2 4
∑ x i=¿1780 ∑ ( xi −x ) 2=1026
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=1780 ÷ 20
x=89
S x=
√ ∑ ( x i−x)2 =
n−1 √ 1026
20−1
S x =7.35
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean general average of students who passed the exam was
x p = ∑ x p ÷ nP
= 88.5
The mean general average of students who failed the exam was
x q = ∑ xq ÷ n F
= 89.75
The proportion of students who passed the exam was
P p=nP ÷ n
P p=12 ÷ 20
P p=0.6
Pq =8 ÷20
Pq =0.4
We will compute the value using critical value tables for normal distribution. It provide the z-
score at the point Pp and Pq, z = 0.13
2
1 −z 2
y= e
√2 π
2
1 −0.13 2
y= e
√2 π
y=0.4
Compute, the biserial correlation coefficient using the values computed earlier:
[
rb =
sx ]
x p −x q P p P q
y
rb =
[ 7.35 ]
88.5−89.75 (0.6)(0.4 )
0.4
r b =−0.10
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since
df = n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since we are conducting a two-tailed
test and α = 0.05, the critical value is 0.444
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtained value is |R b|=
0.10. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater the absolute value of the obtained value, we do not
reject the null hypothesis
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no significant correlation
between SCASTS results and general average.
Reporting the Results
A group of researchers conduct a study to compare SCASTS results and general average.
Twelve students failed the exam (nP = 12) and 8 students failed the exam (n F = 8). A biserial
correlation produced no significant results (rb(18) = -0.10p > 0.05). The data suggest that there is
no significant correlation between SCASTS results and general average.
Example 3
A group of researchers wished to know the relationship between job application and
working years’ experience. Job application will be measured through accepted/not accepted
basis. Twenty applicants were involved in the study, where 5 of them were accepted, while those
15 wasn’t. Job application is a continuous dichotomous variable and working years’ experience
is an interval/scale variable. Therefore, the researchers will be using biserial correlation.
Table 5.
Participant Job Application Working Years’ Experience
1 ACCEPTED 12
2 ACCEPTED 13
3 ACCEPTED 13
4 ACCEPTED 9
5 ACCEPTED 3
6 NOT ACCEPTED 5
7 NOT ACCEPTED 10
8 NOT ACCEPTED 12
9 NOT ACCEPTED 4
10 NOT ACCEPTED 2
11 NOT ACCEPTED 6
12 NOT ACCEPTED 10
13 NOT ACCEPTED 9
14 NOT ACCEPTED 4
15 NOT ACCEPTED 12
16 NOT ACCEPTED 4
17 NOT ACCEPTED 8
18 NOT ACCEPTED 15
19 NOT ACCEPTED 11
20 NOT ACCEPTED 3
Research Hypothesis
The null hypothesis states that there is no correlation between job application and
working years’ experience. The research hypothesis states that there is a correlation between job
application and working years’ experience.
H0: ρb = 0
H1: ρb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Job application is a
continuous dichotomous variable and working years’ experiences is an interval scale variable.
Table 6.
Participant Job Application Working x i−x (x−x )2
Years’
Experience
1 ACCEPTED 12 3.75 14.0625
2 ACCEPTED 13 4.75 22.5625
3 ACCEPTED 13 4.75 22.5625
4 ACCEPTED 9 0.75 0.5625
5 ACCEPTED 3 -5.25 27.5625
6 NOT ACCEPTED 5 -3.25 10.5625
7 NOT ACCEPTED 10 1.75 3.0625
8 NOT ACCEPTED 12 3.75 14.0625
9 NOT ACCEPTED 4 -4.25 18.0625
10 NOT ACCEPTED 2 -6.25 39.0625
11 NOT ACCEPTED 6 -2.25 5.0625
12 NOT ACCEPTED 10 1.75 3.0625
13 NOT ACCEPTED 9 0.75 0.5625
14 NOT ACCEPTED 4 -4.25 18.0625
15 NOT ACCEPTED 12 3.75 14.0625
16 NOT ACCEPTED 4 -4.25 18.0625
17 NOT ACCEPTED 8 -0.25 0.0625
18 NOT ACCEPTED 15 6.75 45.5625
19 NOT ACCEPTED 11 2.75 7.5625
20 NOT ACCEPTED 3 -5.25 27.5625
∑ x i=¿165 ∑ ( xi −x ) 2=311.75
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=165 ÷ 20
x=8.25
S x=
√ ∑ ( x i−x)2 =
n−1 √ 311.75
20−1
S x =4.05
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean working years’ experience of accepted applicants was
x p=∑ x p ÷ n A
x p=12+13+13+ 9+3 ÷5
= 10
The mean working years’ experience of not accepted applicants was
x q=∑ xq ÷ n NA
= 7.67
The proportion of accepted applicants was
P p=n A ÷ n
P p=5 ÷ 20
P p=0.25
Pq =15÷ 20
Pq =0.75
We will compute the value using critical value tables for normal distribution. It provide the z-
score at the point Pp and Pq, z = 0.66
2
1 −z 2
y= e
√2 π
2
1 −0.66 2
y= e
√2π
y=0.3
Compute, the biserial correlation coefficient using the values computed earlier:
[
rb =
sx ]
x p −x q P p P q
y
rb =
[ 4.05 ]
10−7.67 (0.25)(0.75)
0.3
r b =0.36
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since
df = n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since we are conducting a two-tailed
test and α = 0.05, the critical value is 0.444
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtained value is |R b|=
0.36. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater the absolute value of the obtained value, we do not
reject the null hypothesis
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no significant correlation
between job application and working years’ experience
Reporting the Results
A group of researchers conduct a study to compare job application and working years’
experience. Five applicants were accepted (nA = 5) and 15 applicants weren’t accepted (n NA =
15). A biserial correlation produced no significant results (rb(18) = 0.36, p > 0.05). The data
suggest that there is no significant correlation between job application and working years’
experience.
Example 4
A group of researchers wished to know the relationship between thesis proposal and
number of errors. Thesis proposal will be measured through accepted/denied basis. Twenty
college students were involved in the study, where 5 of them were accepted, while those 15 was
denied. Thesis proposal is a continuous dichotomous variable and number of error is an
interval/scale variable. Therefore, the researchers will be using biserial correlation.
Table 7.
Participant Thesis Proposal Number of Errors
1 ACCEPTED 2
2 ACCEPTED 9
3 ACCEPTED 4
4 ACCEPTED 5
5 ACCEPTED 6
6 DENIED 5
7 DENIED 7
8 DENIED 7
9 DENIED 3
10 DENIED 10
11 DENIED 5
12 DENIED 1
13 DENIED 2
14 DENIED 5
15 DENIED 8
16 DENIED 3
17 DENIED 5
18 DENIED 4
19 DENIED 6
20 DENIED 4
Research Hypothesis
The null hypothesis states that there is no correlation between thesis proposal and number
of errors. The research hypothesis states that there is a correlation thesis proposal and number of
errors.
H0: ρb = 0
H1: ρb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Thesis proposal is a
continuous dichotomous variable and number of error is an interval scale variable.
Table 8.
Participant Thesis Proposal Number of x i−x (x−x )2
Errors
1 ACCEPTED 2 -3.05 9.3025
2 ACCEPTED 9 3.95 15.6025
3 ACCEPTED 4 -1.05 1.1025
4 ACCEPTED 5 -0.05 0.0025
5 ACCEPTED 6 0.95 0.9025
6 DENIED 5 -0.05 0.0025
7 DENIED 7 1.95 3.8025
8 DENIED 7 1.95 3.8025
9 DENIED 3 -2.05 4.2025
10 DENIED 10 4.95 24.5025
11 DENIED 5 -0.05 0.0025
12 DENIED 1 -4.05 16.4025
13 DENIED 2 -3.05 9.3025
14 DENIED 5 -0.05 0.0025
15 DENIED 8 2.95 8.7025
16 DENIED 3 -2.05 4.2025
17 DENIED 5 -0.05 0.0025
18 DENIED 4 -1.05 1.1025
19 DENIED 6 0.95 0.9025
20 DENIED 4 -1.05 1.1025
∑ x i=¿101 ∑ ( xi −x ) 2=104.95
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=101 ÷ 20
x=5.05
S x=
√ ∑ ( x i−x)2 =
n−1 √ 104.95
20−1
S x =2.35
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean number of errors of accepted thesis was
x p=∑ x p ÷ n A
x p=2+9+ 4+ 5+6 ÷ 5
= 5.2
The mean number of errors of denied thesis was
x q=∑ xq ÷ n D
=5
The proportion of accepted thesis was
P p=n A ÷ n
P p=5 ÷ 20
P p=0.25
Pq =15÷ 20
Pq =0.75
We will compute the value using critical value tables for normal distribution. It provide the z-
score at the point Pp and Pq, z = 0.66
2
1 −z 2
y= e
√2 π
2
1 −0.66 2
y= e
√2π
y=0.3
Compute, the biserial correlation coefficient using the values computed earlier:
rb =
[ sx ]
x p −x q P p P q
y
rb =
[ 2.35 ]
5.2−5 (0.25)( 0.75)
0.3
r b =0.05
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since
df = n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since we are conducting a two-tailed
test and α = 0.05, the critical value is 0.444
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtained value is |R b|=
0.05. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater the absolute value of the obtained value, we do not
reject the null hypothesis
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no significant correlation
between thesis proposal and number of errors.
Reporting the Results
A group of researchers conduct a study to compare thesis proposal and number of errors.
Five thesis proposal was accepted (nA = 5) and 15 thesis proposal was denied (nD = 15). A
biserial correlation produced no significant results (rb(18) = 0.05, p > 0.05). The data suggest that
there is no significant correlation between thesis proposal and number of errors.
Example 5
A group of researchers wished to know the relationship between thesis proposal
and number of errors. Thesis proposal will be measured through accepted/denied basis. Twenty
college students were involved in the study, where 5 of them were accepted, while those 15 was
denied. Thesis proposal is a continuous dichotomous variable and number of error is an
interval/scale variable. Therefore, the researchers will be using biserial correlation.
Table 9.
Participant Thesis Proposal Number of Errors
1 ACCEPTED 2
2 ACCEPTED 9
3 ACCEPTED 4
4 ACCEPTED 5
5 ACCEPTED 6
6 DENIED 5
7 DENIED 7
8 DENIED 7
9 DENIED 3
10 DENIED 10
11 DENIED 5
12 DENIED 1
13 DENIED 2
14 DENIED 5
15 DENIED 8
16 DENIED 3
17 DENIED 5
18 DENIED 4
19 DENIED 6
20 DENIED 4
Research Hypothesis
The null hypothesis states that there is no correlation between thesis proposal and number
of errors. The research hypothesis states that there is a correlation thesis proposal and number of
errors.
H0: ρb = 0
H1: ρb ≠ 0
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance.
Test Statistics
It is decided to analyze the relationship between the two variables. A correlation will
provide the relative strength of the relationship between the two variables. Thesis proposal is a
continuous dichotomous variable and number of error is an interval scale variable.
Table 8.
Participant Thesis Proposal Number of x i−x (x−x )2
Errors
1 ACCEPTED 2 -3.05 9.3025
2 ACCEPTED 9 3.95 15.6025
3 ACCEPTED 4 -1.05 1.1025
4 ACCEPTED 5 -0.05 0.0025
5 ACCEPTED 6 0.95 0.9025
6 DENIED 5 -0.05 0.0025
7 DENIED 7 1.95 3.8025
8 DENIED 7 1.95 3.8025
9 DENIED 3 -2.05 4.2025
10 DENIED 10 4.95 24.5025
11 DENIED 5 -0.05 0.0025
12 DENIED 1 -4.05 16.4025
13 DENIED 2 -3.05 9.3025
14 DENIED 5 -0.05 0.0025
15 DENIED 8 2.95 8.7025
16 DENIED 3 -2.05 4.2025
17 DENIED 5 -0.05 0.0025
18 DENIED 4 -1.05 1.1025
19 DENIED 6 0.95 0.9025
20 DENIED 4 -1.05 1.1025
∑ x i=¿101 ∑ ( xi −x ) 2=104.95
Calculate the mean and the standard deviation for the interval data:
x=∑ x i ÷ n
x=101 ÷ 20
x=5.05
S x=
√ ∑ ( x i−x)2 =
n−1 √ 104.95
20−1
S x =2.35
Compute the means and proportions of the values associated with each item from the
dichotomous variable. The mean number of errors of accepted thesis was
x p=∑ x p ÷ n A
x p=2+9+ 4+ 5+6 ÷ 5
= 5.2
The mean number of errors of denied thesis was
x q=∑ xq ÷ n D
=5
The proportion of accepted thesis was
P p=n A ÷ n
P p=5 ÷ 20
P p=0.25
Pq =15÷ 20
Pq =0.75
We will compute the value using critical value tables for normal distribution. It provide the z-
score at the point Pp and Pq, z = 0.66
2
1 −z 2
y= e
√2 π
2
1 −0.66 2
y= e
√2π
y=0.3
Compute, the biserial correlation coefficient using the values computed earlier:
rb =
[ sx ]
x p −x q P p P q
y
rb =
[ 2.35 ]
5.2−5 (0.25)( 0.75)
0.3
r b =0.05
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since
df = n – 2 and n = 20, then df = 20 – 2. Therefore, df = 18. Since we are conducting a two-tailed
test and α = 0.05, the critical value is 0.444
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 0.444 and the obtained value is |R b|=
0.05. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater the absolute value of the obtained value, we do not
reject the null hypothesis
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no significant correlation
between thesis proposal and number of errors.
Reporting the Results
A group of researchers conduct a study to compare thesis proposal and number of errors.
Five thesis proposal was accepted (nA = 5) and 15 thesis proposal was denied (nD = 15). A
biserial correlation produced no significant results (rb(18) = 0.05, p > 0.05). The data suggest that
there is no significant correlation between thesis proposal and number of errors.
CHI-SQUARE
Example 1
A group of researchers conduct an activity to determine if there is a significant
preferences for a type of smartphone brands that were sold in shopping malls. The target group is
the high school students. It was assumed that there is no preference when the study was started.
The different smartphone brands are Vivo, Huawei, iPhone, Samsung, and Realme. The sample
size for this study was n=30.
Table 1.
Smartphone Brands Observed Frequency
Vivo 8
Huawei 7
iPhone 6
Samsung 5
Realme 9
The researchers want to determine if there is any preference for one of the five
smartphone brands that were purchased by high school students in shopping malls. Since the data
only need to be classified into categories, and no sample mean nor sum of squares needs to be
calculated, the x2 statistic goodness-of-fit test can be used to test the nonparametric data.
Research Hypothesis
The null hypothesis states that there is no preference among the different categories.
There is an equal proportion of frequency of participations selecting each smartphone brands.
The research hypothesis states that one or more of smartphone brand is preferred over the others
by high school.
H0: In the population of high school students, there is no preference of one smartphone
brands over any other. Thus, the five different brands are selected equally often and the
population distribution has the proportions shown in Table 2.
Table 2.
Smartphone Brands Proportion
Vivo 20%
Huawei 20%
iPhone 20%
Samsung 20%
Realme 20%
HA: In the population of high school students, there is at least one smartphone brands
preferred over others.
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance
Test Statistics
First, tally the observed frequencies, f0, for the 30 high school students who were in the
study. Refer to Table 1 for observed frequency distribution.
Smartphone Brands Observed Frequency
Vivo 8
Huawei 7
iPhone 6
Samsung 5
Realme 9
Calculate the expected frequency for each category. In this case, the expected frequency,
fe, will be same for all five categories since our research problem assumes that all categories are
equal:
f e =Pi n
1
f e = (30)
5
f e =6
Table 3
Smartphone Brands Expected Frequencies
Vivo 6
Huawei 6
iPhone 6
Samsung 6
Realme 6
Using the values for the observed and expected frequencies, the x2 statistics may be
calculated:
2
( f o−f e )
x =∑
2
fe
¿¿¿
2
x =2.5
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since,
there are five categories, C = 5. To find degrees of freedom, use df = C-1. Therefore, df = 5-1 is
4. The critical value is found in the x2 table for three four degrees of freedom, df = 4. Since we
set the α = 0.05, the critical value is 9.49. A calculated value that is greater than or equal to 9.49
will lead us to reject the null hypothesis.
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 9.49 and the obtained value is x2=
2.5. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater than the obtained value of the obtained value, we do
not reject the null hypothesis.
Note that the critical value for α = 0.01 is 13.28. Since the obtained value is 2.5, a value
less than 13.28, the data indicate that the results are not highly significant.
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no real difference among
smartphone brands preferred by high school students in shopping malls.
Reporting the Results
The reporting of results for the x2 goodness of fit should include such information as the
total number of participants in the sample and the good that were classified in each category. In
some cases, x2 statistics, degrees of freedom, and p-values relation to α is necessary. For this
study, p < 0.01, should also be indicated with the data to show the degree of significance of the
x2.
For this example, 30 high school students were surveyed to determine which smartphone
brands preferred to buy. The five choices are Vivo, Huawei, iPhone, Samsung, and Realme. The
choices were 8, 7, 6, 5, and 9 respectively. The x2 goodness-of-fit test was not significant (x2(4) =
2.5, p > 0.01).
Example 2
A researchers conduct an activity to determine if there is a significant preferences for a
most played sports that were commonly play by college students . The target group is the
college students. It was assumed that there is no preference when the study was started. The
different sports are Basketball, Volleyball, Badminton, and Football. The sample size for this
study was n=25.
Table 4.
Sports Observed Frequency
Basketball 8
Volleyball 7
Badminton 6
Football 4
The researchers want to determine if there is any preference for one of the four most
played sports by the college students in DORSU. Since the data only need to be classified into
categories, and no sample mean nor sum of squares needs to be calculated, the x2 statistic
goodness-of-fit test can be used to test the nonparametric data.
Research Hypothesis
The null hypothesis states that there is no preference among the different categories.
There is an equal proportion of frequency of participating each sports. The research hypothesis
states that one or more sports are preferred by college students.
H0: In the population of college students, there is no preference of one sports over any
other. Thus, the five different brands are selected equally often and the population distribution
has the proportions shown in Table 2.
Table 5.
Sports Proportion
Basketball 25%
Volleyball 25%
Badminton 25%
Football 25%
HA: In the population of college students, there is at least one sports preferred over others.
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance
Test Statistics
First, tally the observed frequencies, f0, for the 30 high school students who were in the
study. Refer to Table 4 for observed frequency distribution.
Sports Observed Frequency
Basketball 8
Volleyball 7
Badminton 6
Football 4
Calculate the expected frequency for each category. In this case, the expected frequency,
fe, will be same for all four categories since our research problem assumes that all categories are
equal:
f e =Pi n
1
f e = (30)
4
f e =7.5
Basketball 7.5
Volleyball 7.5
Badminton 7.5
Football 7.5
Using the values for the observed and expected frequencies, the x2 statistics may be
calculated:
( f o−f e )2
x =∑
2
fe
¿¿¿
2
x =2
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since,
there are five categories, C = 4. To find degrees of freedom, use df = C-1. Therefore, df = 4-1 is
3. The critical value is found in the x2 table for three four degrees of freedom, df = 3. Since we
set the α = 0.05, the critical value is 7.81. A calculated value that is greater than or equal to 7.81
will lead us to reject the null hypothesis.
The researchers want to determine if there is any preference for one of the four favorite
sports cars by random males. Since the data only need to be classified into categories, and no
sample mean nor sum of squares needs to be calculated, the x 2 statistic goodness-of-fit test can
be used to test the nonparametric data.
Research Hypothesis
The null hypothesis states that there is no preference among the different categories.
There is an equal proportion of frequency of participating each sports car. The research
hypothesis states that one or more sports cars are preferred by random males.
H0: In the population random males, there is no preference of one sports car over any
other. Thus, the four different sports cars are selected equally often and the population
distribution has the proportions shown in Table 8.
Table 8.
Sports Car Proportion
Lamborghini 25%
Ducati 25%
Ford Mustang 25%
Honda Civic 25%
HA: In the population of random males, there is at least one sports car preferred over
others.
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance
Test Statistics
First, tally the observed frequencies, f0, for the 30 high school students who were in the
study. Refer to Table 7 for observed frequency distribution.
Sports Car Observed Frequency
Lamborghini 7
Ducati 10
Ford Mustang 9
Honda Civic 4
Calculate the expected frequency for each category. In this case, the expected frequency,
fe, will be same for all four categories since our research problem assumes that all categories are
equal:
f e =Pi n
1
f e = (30)
4
f e =7.5
Lamborghini 7.5
Ducati 7.5
Ford Mustang 7.5
Honda Civic 7.5
Using the values for the observed and expected frequencies, the x2 statistics may be
calculated:
( f o−f e )2
x =∑
2
fe
¿¿¿
2
x =2 .8
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since,
there are five categories, C = 4. To find degrees of freedom, use df = C-1. Therefore, df = 4-1 is
3. The critical value is found in the x2 table for three four degrees of freedom, df = 3. Since we
set the α = 0.05, the critical value is 7.81. A calculated value that is greater than or equal to 7.81
will lead us to reject the null hypothesis.
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 7.81 and the obtained value is x2=
2.8. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater than the obtained value of the obtained value, we do
not reject the null hypothesis.
Note that the critical value for α = 0.01 is 11.34. Since the obtained value is 2.8, a value
less than 11.34, the data indicate that the results are not highly significant.
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no real difference among
sports car preferred by random males.
Reporting the Results
The reporting of results for the x2 goodness of fit should include such information as the
total number of participants in the sample and the good that were classified in each category. In
some cases, x2 statistics, degrees of freedom, and p-values relation to α is necessary. For this
study, p < 0.01, should also be indicated with the data to show the degree of significance of the
x2.
For this example, 30 random males were surveyed to determine which sports card
preferred to buy. The four choices are Lamborghini, Ducati. Ford Mustang, and Honda Civic.
The choices were 7, 10, 9, 4. The x2 goodness-of-fit test was not significant (x2(3) = 2.8, p > 0.01).
Example 4
A group of researchers conduct an activity to determine if there is a significant
preferences for a shoes brand that are sold in shopping malls. The target group is the college
students. It was assumed that there is no preference when the study was started. The different
shoes brands are Nike, Adidas, Word Balance, and Robertson.
Table 10.
Shoes Brand Observed Frequency
Nike 7
Adidas 10
Word Balance 9
Robertson 4
The researchers want to determine if there is any preference for one of the four shoes
brand preferred by college students in shopping malls. Since the data only need to be classified
into categories, and no sample mean nor sum of squares needs to be calculated, the x 2 statistic
goodness-of-fit test can be used to test the nonparametric data.
Research Hypothesis
The null hypothesis states that there is no preference among the different categories.
There is an equal proportion of frequency of participating and selecting each shoes brand. The
research hypothesis states that one or more shoes brand preferred by college students
H0: In the population college students, there is no preference of one shoes brand over any
other. Thus, the four different shoes brand are selected equally often and the population
distribution has the proportions shown in Table 11.
Table 11.
Shoes Brand Proportion
Nike 25%
Adidas 25%
Word Balance 25%
Robertson 25%
HA: In the population of college students, there is at least one shoes brands preferred over
others.
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance
Test Statistics
First, tally the observed frequencies, f0, for the 30 high school students who were in the
study. Refer to Table 11 for observed frequency distribution.
Shoes Brand Observed Frequency
Nike 7
Adidas 10
Word Balance 9
Robertson 4
Calculate the expected frequency for each category. In this case, the expected frequency,
fe, will be same for all four categories since our research problem assumes that all categories are
equal:
f e =Pi n
1
f e = (30)
4
f e =7.5
Nike 7.5
Adidas 7.5
Word Balance 7.5
Robertson 7.5
Using the values for the observed and expected frequencies, the x2 statistics may be
calculated:
( f o−f e )2
x =∑
2
fe
¿¿¿
2
x =2.8
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since,
there are five categories, C = 4. To find degrees of freedom, use df = C-1. Therefore, df = 4-1 is
3. The critical value is found in the x2 table for three four degrees of freedom, df = 3. Since we
set the α = 0.05, the critical value is 7.81. A calculated value that is greater than or equal to 7.81
will lead us to reject the null hypothesis.
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 7.81 and the obtained value is x2=
2.8. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater than the obtained value of the obtained value, we do
not reject the null hypothesis.
Note that the critical value for α = 0.01 is 11.34. Since the obtained value is 2.8, a value
less than 11.34, the data indicate that the results are not highly significant.
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no real difference among
shoes brands preferred by college students.
Reporting the Results
The reporting of results for the x2 goodness of fit should include such information as the
total number of participants in the sample and the good that were classified in each category. In
some cases, x2 statistics, degrees of freedom, and p-values relation to α is necessary. For this
study, p < 0.01, should also be indicated with the data to show the degree of significance of the
x2.
For this example, 30 college students were surveyed to determine which shoes brand they
preferred to buy. The four choices are Nike, Adidas, Word Balance, and Robertson. The choices
were 7, 10, 9, 4,. The x2 goodness-of-fit test was not significant (x2(3) = 2.8, p > 0.01).
Example 5
A group of researchers conduct an activity to determine if there is a significant
preferences for a social media platforms that were used in shopping. The target group is the
college students. It was assumed that there is no preference when the study was started. The
different social media platforms for shopping is Shopee, Lazada, TikTok, and Facebook
Table 13.
Social Media Platforms Observed Frequency
Shopee 7
Lazada 9
Tiktok 9
Facebook 5
The researchers want to determine if there is any preference for one of the four social
media platforms for shopping preferred by college students. Since the data only need to be
classified into categories, and no sample mean nor sum of squares needs to be calculated, the x 2
statistic goodness-of-fit test can be used to test the nonparametric data.
Research Hypothesis
The null hypothesis states that there is no preference among the different categories.
There is an equal proportion of frequency of participating and selecting each social media
platforms. The research hypothesis states that one or more shoes brand preferred by college
students
H0: In the population college students, there is no preference of one social media
platforms over any other. Thus, the four different shoes brand are selected equally often and the
population distribution has the proportions shown in Table 13.
Table 13.
Social Media Platforms Proportion
Shopee 25%
Lazada 25%
Tiktok 25%
Facebook 25%
HA: In the population of college students, there is at least one social media platforms
preferred over others.
Level of Significance
The level of risk also called an alpha (α), is frequently set at 0.05. We will use α = 0.05 in
our example. In other words, there is a 95% chance that any observed statistical difference will
be really and not due to chance
Test Statistics
First, tally the observed frequencies, f0, for the 30 high school students who were in the
study. Refer to Table 13 for observed frequency distribution.
Social Media Platforms Observed Frequency
Shopee 7
Lazada 9
Tiktok 9
Facebook 5
Calculate the expected frequency for each category. In this case, the expected frequency,
fe, will be same for all four categories since our research problem assumes that all categories are
equal:
f e =Pi n
1
f e = (30)
4
f e =7.5
Shopee 7.5
Lazada 7.5
Tiktok 7.5
Facebook 7.5
Using the values for the observed and expected frequencies, the x2 statistics may be
calculated:
( f o−f e )2
x =∑
2
fe
¿¿¿
2
x =1.47
Determine the Value of Need for Rejection of the Null Hypothesis Using the Appropriate
Table of Critical values for the Particular Statistic.
Using the critical values, the table requires that the degrees of freedom be known. Since,
there are five categories, C = 4. To find degrees of freedom, use df = C-1. Therefore, df = 4-1 is
3. The critical value is found in the x2 table for three four degrees of freedom, df = 3. Since we
set the α = 0.05, the critical value is 7.81. A calculated value that is greater than or equal to 7.81
will lead us to reject the null hypothesis.
Comparing the Obtain Value with Critical Value
The critical value for rejecting the null hypothesis is 7.81 and the obtained value is x2=
1.47. If the critical value is less than or equal to the obtained value, we must reject the null
hypothesis. If the critical value is greater than the obtained value, we must not reject the null
hypothesis. Since the critical value is greater than the obtained value of the obtained value, we do
not reject the null hypothesis.
Note that the critical value for α = 0.01 is 11.34. Since the obtained value is 1.47, a value
less than 11.34, the data indicate that the results are not highly significant.
Interpret the Results
We do not reject the null hypothesis, suggesting that there is no real difference among
social media platforms for shopping prefer by college students.
Reporting the Results
The reporting of results for the x2 goodness of fit should include such information as the
total number of participants in the sample and the good that were classified in each category. In
some cases, x2 statistics, degrees of freedom, and p-values relation to α is necessary. For this
study, p < 0.01, should also be indicated with the data to show the degree of significance of the
x2.
For this example, 30 college students were surveyed to determine which social media
platforms for shopping they preferred to use. The four choices are Shopee, Lazada, TikTok, and
Facebook. The choices were 7, 9, 9, and 5. The x2 goodness-of-fit test was not significant (x2(3) =
1.47, p > 0.01).
STAT 122 – NON-
PARAMETRIC STATISTICS
SPEARMAN-RANK ORDER CORRELATION
POINT-BISERIAL CORRELATION
BISERIAL CORRELATION
CHI – SQUARE
EXERCISES
2.1. List the elements of each of the following sample spaces:
2.2. Use the rule method to describe the sample space S consisting of all points in the first
quadrant inside a circle of radius 3 with center at the origin.
S = {(x, y) | x 2 + y 2 < 9; x ≥ 0, y ≥ 0}
2.3. Which of the following events are equal?
(a) A = {1, 3};
A = {1, 3}
(b) B = {x | x is a number on a die};
B = {1, 2, 3, 4, 5, 6}
(c) C = {x | x2 − 4x+3 = 0};
C = {x | x 2 − 4x + 3 = 0} = {x | (x − 1)(x − 3) = 0} = {1, 3}
(a) The probabilities that an automobile salesperson will sell 0, 1, 2, or 3 cars on any given day in
February are, respectively, 0.19, 0.38, 0.29, and 0.15.
Sum of the probabilities exceeds 1
(b) The probability that it will rain tomorrow is 0.40, and the probability that it will not rain
tomorrow is 0.52.
Sum of the probabilities is less than 1
(c) The probabilities that a printer will make 0, 1, 2, 3, or 4 or more mistakes in setting a
document are, respectively, 0.19, 0.34,−0.25, 0.43, and 0.29.
A negative probability
(d) On a single draw from a deck of playing cards, the probability of selecting a heart is 1/4, the
probability of selecting a black card is 1/2, and the probability of selecting both a heart and a
black card is 1/8.
Probability of both a heart and a black card is zero
2.50 Assuming that all elements of S in Exercise 2.8 on page 42 are equally likely to occur, find
(a) the probability of event A;
P(A) = 5/18
(b) the probability of event C;
P(C) = 1/3
(c) the probability of event A ∩ C.
P(A ∩ C) = 7/36
2.95 In a certain region of the country it is known from past experience that the probability of
selecting an adult over 40 years of age with cancer is 0.05. If the probability of a doctor correctly
diagnosing a person with cancer as having the disease is 0.78 and the probability of incorrectly
diagnosing a person without cancer as having the disease is 0.06, what is the probability that an
adult over 40 years of age is diagnosed as having cancer?
Consider the events: C: an adult selected has cancer, D: the adult is diagnosed as
having cancer. P(C) = 0.05, P(D | C) = 0.78, P(C ′ ) = 0.95 and P(D | C ′ ) = 0.06. So,
P(D) = P(C ∩ D) + P(C ′ ∩ D) = (0.05)(0.78) + (0.95)(0.06) = 0.096
2.96 Police plan to enforce speed limits by using radar traps at four different locations within the
city limits. The radar traps at each of the locations L1, L2, L3, and L4 will be operated 40%,
30%, 20%, and 30% of the time. If a person who is speeding on her way to work has
probabilities of 0.2, 0.1, 0.5, and 0.2, respectively, of passing through these locations, what is the
probability that she will receive a speeding ticket?
Let S1, S2, S3, and S4 represent the events that a person is speeding as he passes
through the respective locations and let R represent the event that the radar traps is
operating resulting in a speeding ticket. Then the probability that he receives a
speeding ticket: P(R) = ∑ 4 i=1 P(R | Si)P(Si) = (0.4)(0.2) + (0.3)(0.1) + (0.2)(0.5) +
(0.3)(0.2) = 0.27
Chapter 11 – Simple Linear Regression and Correlation
Exercises
11.1. A study was conducted at Virginia Tech to determine if certain static arm-strength
measures have an influence on the “dynamic lift” characteristics of an individual. Twenty-five
individuals were subjected to strength tests and then were asked to perform a weightlifting test in
which weight was dynamically lifted overhead. The data are given here.
11.2 The grades of a class of 9 students on a midterm report (x) and on the final examination (y)
are as follows:
(b) Estimate the final examination grade of a student who received a grade of 85 on the
midterm report.
11.3 The amounts of a chemical compound y that dissolved in 100 grams of water at various
temperatures x were recorded as follows:
(c) Estimate the amount of chemical that will dissolve in 100 grams of water at 50◦C.
For x = 50, ˆy = 5.8254 + (0.5676)(50) = 34.205 grams
STAT 121 – STATISTICAL
ANALYSIS