"Year of Universal Health Care"

NATIONAL UNIVERSITY OF TRUJILLO

FACULTY OF MEDICINE
SCHOOL OF MEDICINE

Practice Nº 07: Hypothesis testing

STUDENTS
Baltodano Caipo, Renzo
Castillo Bustamante, Lourdes
Castillo Pacheco, Andrew
Mejía Marín, Diego

GROUP
9

TEACHER
Dr. Carlos Risco Dávila
Ms. María Isabel Izquierdo Henríquez

SUBJECT
Biostatistics II

TRUJILLO - PERU
2020

1. Investigators Bertino et al (A - 3) conducted a study to examine data collected pertaining to the

pharmacokinetics of gentimycin in three 18-year populations: patients with acute leukemia, patients
with other non-leukemic malignancies, and patients without occult malignant disease or
pathophysiologies other than renal failure known to alter the pharmacokinetics of gentamicin. Among
the statistics reported by the investigators was a value of 59.1 as the initial calculated mean creatine
clearance, with a standard deviation of 25.6 for a sample of 211 patients with malignant disease other
than leukemia. The aim is to know if it is possible to conclude that the mean for the population of
individuals presenting the same pathological condition is less than 60. Let α = 0.10.
HYPOTHESIS TEST FOR THE POPULATION MEAN
We put forward the statistical hypotheses:
Ho: µ ≥ 60
H1: µ < 60
Level of significance:
α= 0.10 = 10%
Sample and population data
n= 211
x-bar= 59.1
s= 25.6
Z test statistic
CRITICAL VALUE METHOD
Z0 = (59.1 - 60)/(25.6/RAIZ(211)) = -0.51067
Z0.1 = DISTR.NORM.STANDARD.INV (0.1) = - 1.282
Decision: As Z0 > Z0.1 therefore we do not reject H0 (LOWER TAIL TEST).
P-VALUE METHOD
p-value = 0.3048 > 0.1
Decision: As the p-value > α, therefore H0 is not rejected.

2. One of the studies by researchers Klesges et al (A - 4) aims to find out the factors associated with
discrepancies between carboxyhemoglobin levels and self-reported smoking status. A sample of 3918
self-reported nonsmokers had a mean carboxyhemoglobin level of 0.9 with a standard deviation of 0.96.
The aim is to know if it is possible to conclude that the population mean is less than 1.0. Let α= 0.01.
HYPOTHESIS TEST FOR THE POPULATION MEAN
We put forward the statistical hypotheses:
Ho: µ ≥ 1
H1: µ < 1
Level of significance:
α= 0.01 = 1%
Sample data and population data n= 3918 x-bar= 0.9 s= 0.96

Z test statistic
CRITICAL VALUE METHOD
Z0 = (0.9 - 1)/(0.96/RAIZ(3918)) = -6.5202
Z0.01 = DISTR.NORM.STANDARD.INV (0.01) = - 2.326
Decision: As Z0 < Z0.01 we therefore reject H0 (LOWER TAIL TEST).
P-VALUE METHOD
p-value = 0.000000000035107 < 0.01
Decision: As the p-value < α, therefore H0 is rejected.

3. Dr. Jeffrey M. Barret (A - 5) from Lakeland, Florida, United States, reported data corresponding to
8 cases of umbilical cord prolapse. Mothers' ages 25, 28, 17, 26, 27, 22, 25 and 30 years. The aim is to
know if it is possible to conclude that the mean of the population from which the sample is supposed to
be drawn is older than 20 years. Let α= 0.01.
HYPOTHESIS TEST FOR THE POPULATION MEAN
We put forward the statistical hypotheses: 25
Ho: µ ≤ 20 28
17
H1: µ > 20 26
Level of significance: 27
22
α= 0.01 = 1%
25
Sample and population data n= 8 30

X-bar: (=AVERAGE (25,28,17........30)) = 25

s= DESVEST.M (25,28,17........30) = 4
t-test statistic
CRITICAL VALUE METHOD t0 = (25- 20)/(4/RAIZ(8)) = 3.53553
t0.01 = INV.T.2C (2*0.01;8-1) = 2.99795157
Decision: As t0 > t0.01 we therefore reject H0 (UPPER TAIL TEST).
P-VALUE METHOD
p-value = DISTR.T.CD (3.53553;8-1) = 0.00476339 < 0.01
Decision: As the p-value < α, therefore H0 is rejected.
4. A study was made of a sample of 25 files of chronically ill patients treated as outpatients. The mean
number of consultations per patient was 4.8 and a standard deviation of the sample was 2. Is it possible
to conclude from these data that the population mean is greater than 4 visits per patient? Assume that
the probability of making a type I error is 0.05 What are the assumptions that must be met?
HYPOTHESIS TEST FOR THE POPULATION MEAN
We put forward the statistical hypotheses:
Ho: µ ≤ 4
H1: µ > 4
Level of significance:
α= 0.05 = 5%
Sample and population data
n= 25
X-bar= 4.8
s= 2
t-test statistic
CRITICAL VALUE METHOD
t0 = (4.8 - 4)/(2/RAIZ(25)) = 2
t0.05 = = INV.T.2C (2*0.05;8-1) = 1.710882
Decision: As t0 > t0.05 we therefore reject H0 (UPPER TILT TEST).
P-VALUE METHOD
p-value =DISTR.T.CD (2;25-1) = 0.02847 < 0.05
Decision: As the p-value < α, therefore H0 is rejected.
That is, the data do provide sufficient evidence to indicate, at a significance level of 5%, that the mean
number of visits by chronically ill patients is greater than 4 visits per patient.
5. Diana M Bailey conducted a study to examine the causes of certified occupational therapists leaving
the profession temporarily or permanently. Of 696 individuals who responded to a data collection
survey, 63% decided to leave their jobs to have and care for their own children. Based on these data, is it
possible to conclude that, in general, more than 60% of the individuals in the sampled population
decided to leave their jobs to dedicate themselves to having and caring for their own children? Let α =
0.05. What is the sampled population? What assumptions are necessary for the procedure to be valid?
HYPOTHESIS TEST FOR THE POPULATION PROPORTION
We put forward the statistical hypotheses:
Ho: µ ≤ 0.60
H1: µ > 0.60
Level of significance:
α= 0.05 = 5%
Sample and population data
n= 696

>x=
p-bar= 0.63
438.48
Z test statistic
CRITICAL VALUE METHOD
Z0 =(0.63-0.60)/RAIZ((0.60*(1-0.60)/696)) = 1.6155
Z0.05=DISTR.NORM.STANDARD.INV (0.95)= 1.6450
Decision: As Z0 < Z0.05 therefore we do not reject H0 (UPPER TAIL TEST).
P-VALUE METHOD
p-value =DISTR.NORM.STAND (1.6155)= 0.947 > 0.05Decision: As the p-value > α, therefore H0 is not
rejected.
THAT MEANS THAT THE NUMBER OF INVIDUALS (OCCUPATIONAL THERAPISTS)
DECIDED TO LEAVE THEIR JOBS TO HAVE AND CARE FOR THEIR OWN CHILD IS LESS
THAN 60%.
6. In an article published in the American Journal of Public Health, Colsher et al (A . 19) describe
the results of the health survey applied to 119 male convicts aged 50 years and older residing in the
facilities of a state social readaptation center. It was found that 21.6% of respondents reported a
history of venereal disease. Based on these findings, is it possible to conclude that more than 15% of
the sampled population has a history of venereal disease? Let α= 0.05.

Solution:
n = 119 males
n ≥ 30
Age = 50 years
We put forward the statistical hypotheses:
H₀= μ≤ 0.15
H₁= μ> 0.15

Level of significance:
0.05
Calculation of critical test values:
0.066 z= 1.64

0.85
0.00714286 Z0= 2.01633331
0.00107143
x - U0
0.03273268 "=($)

Decision:
As Z0 > Z0.05 we therefore reject H0.
Conclusion:
It is denied that more than 15% of male convicts aged 50 years and older residing in the facilities of a
State social rehabilitation center have a history of venereal diseases.
7. Henning et al. (A - 20) found that 66% of children in a sample of 670 completed the entire
hepatitis B vaccination series. Is it possible to conclude that, based on these data, in the sampled
population, more than 60% have the complete series of hepatitis B vaccines? Let α = 0.05
Solution:

n= 760 children

n ≥ 30

0.05

We put forward the statistical hypotheses:

H₀= μ≤ 0.6
H₁= μ> 0.6
Level of significance:
Calculation of critical test values:
0.06 z= 1.64

0.4
0.00059701 Z0= 3.1701735
0.00035821
0.01892641

x - U0
"=(9)

Decision:
As Z0 > Z0.05 we therefore reject H0.
Conclusion:
It is denied that more than 60% of children have the full series of hepatitis B vaccines.
8. Hemoglobin values (g %) were recorded from a sample of 20 children who were part of an acute
leukemia study. The variance of the observations was 5. Do these data provide sufficient evidence to
indicate that the population variance is greater than 4? Let α= 0.05

Solution:
STATISTICAL ASSUMPTIONS
H0 o2 <= 4
H1 a2 > 4

LEVEL OF SIGNIFICANCE
α= 0.05

SAMPLE DATA
n= 20
S²= 5

Test statistic:
CRITICAL VALUE METHOD
X²0= 23.7500
X²(0,05)= 30.144

DECISION
X²0 < X²(0.05)
23.750 < 30.144

Answer: Therefore, H0 is not rejected, indicating that the variance of the population of children is
less than or equal to 4.
9. A sample of 25 large hospital administrators participated in a study to investigate the nature and
degree of frustration and emotional stress associated with work. Each participant was given a test to
estimate the degree of emotional stress they experienced as a result of the duties and responsibilities
associated with their job. The variance of the results obtained was 30. Is it possible to conclude from
these data that the population variance is greater than 25? Let α= 0.05.

Solution:

STATISTICAL ASSUMPTIONS
Solution:...........................................................................................................................................6
X20= 22.04....................................................................................................................................11
Reply..............................................................................................................................................11
_ (K,-K2)-(4_12 _ C4.5 - 3.4) - 0 _ 1.1.............................................................................................12
s+s11+15...........................................................................................................................................12
Determine significance level:........................................................................................................13
p-bar prom=.................................................................................................................................19

SAMPLE DATA
n= 25
S²= 30

Test statistic:
CRITICAL VALUE METHOD
X²0= 28.8000
X²(0,05)= 30.144

DECISION
X²0 < X²(0.05)
28.800 < 30.144

Answer: Therefore, H0 is not rejected, indicating that the variance of the population subject to
emotional stress from work is less than or equal to 25.
10. In a study of 15 patients with pulmonary sarcoid disease, blood gas concentrations were
measured. The variance of PaO2 values (in mmHg) was 450. Test the null hypothesis that the
population variance is greater than 250. Let α= 0.01.

Solution:
STATISTICAL ASSUMPTIONS
H0 o2 <= 250
H1 o2 > 250

SIGNIFICANCE LEVEL α= 0.01

SAMPLE DATA
n= 15
S²= 450

TEST STATISTIC
CRITICAL VALUE METHOD
X²0= 25.2000
X²(0,05)= 36.191

DECISION
X²0 < X²(0.05)
25.200 < 36.191

Answer: Therefore, H0 is not rejected, indicating that the variance of the population is less than or
equal to 250.
11. An analysis of amniotic fluid from a simple random sample of 15 pregnant women provided the
following amounts of total protein (grams per 100ml):
0.69 1.04 0.39 0.37 0.64 0.73 0.69 1.04
0.83 1.00 0.19 0.61 0.42 0.20 0.79
Do these data provide sufficient evidence to indicate that the population variance is greater than 0.05?
Let α= 0.05. What assumptions are necessary?
Solution:
0.69 1.04 0.39 0.37 0.64 0.73 0.69 1.04
0.83 1 0.19 0.61 0.42 0.2 0.79

15
0.05

S2 =
0.07871714
0.05
H,c2 =
H:02> 0.05

By the critical value method:

, (15-1) *0.079
X20= ------------- ----------
0.05

X20= 22.04

23.6847913

Reply

As, X20 < X20.05

then we cannot reject the hypothesis.
12. A team of researchers wants to know whether the data they have collected provide sufficient
evidence to indicate a difference between the mean serum uric acid concentrations of normal
individuals and individuals with Down syndrome. The data consist of serum uric acid readings from
12 individuals with Down syndrome and 15 healthy individuals. The means are x-bar1= 4.5 mg /100
ml and x-bar2=3.4 mg/ml. The data correspond to two simple and independent random samples, each
drawn from a population following a normal distribution, with a variance equal to 1 for the Down
syndrome population and 1.5 for the healthy population.

1. Express the null hypothesis:

H0: μ1 - μ2 = 0
2. Express the alternative hypothesis:
H1: μ1 - μ2 ≠ 0

3. Specify the level of significance:

α = 0.05
4. Determine the sample size:
n1=12 n2= 15
5. Establish the critical values that establish the rejection regions from the non-rejection regions.
Corresponding to ± 1.96 (Bilateral):
_ _ _ (1-2 )-(u1-u2)
2— -- en* vn1n2

7. Collect data and calculate the sample value of the appropriate statistical test:

_ (K,-K2)-(4_12 _ C4.5 - 3.4) - 0 _ 1.1

" n; fl 1.5 0.4282 - 7
s+s11+15
vn1 n2

8. Determine whether the statistical test has been in the rejection zone to a non-rejection zone.
2.57 falls in the region of rejection.
9. Determine the statistical decision.
H0 is rejected since 2.57 falls in the rejection region as it is greater than 1.96.
10. Express the statistical decision in terms of the problem:
Based on these data, there are indications that the population means are different.
13. Frigerio et al. (A-I0) measured the energy consumed in 32 Gambian women. Sixteen of the
individuals studied were lactating women (L) and the remainder were non-pregnant non-lactating
women (NENL). The following data were reported:

Sample Energy consumption (kj/d)

L 5289, 6209, 6054, 6665, 6343, 7699, 5678, 6954, 6916,

4770, 5979, 6305, 6502, 6113. 6347, 5657
NENL 9920, 8581, 9305, 10765, 8079, 9046, 7134, 8736, 10230, 7121,
8665, 5167, 8527, 7791, 8782, 6883

To pose statistical hypotheses:

H0 = μ1-μ2=0
H1 = μ1-μ2≠0

Determine significance level:

α= 0.05 5%

Data in the sample and population

n1 = 16 n2 = 16
x-bar1 = 6217.5 x-bar2 = 8421.3125
s1 = 693.0005291 s2 = 1396.842092
S1^2 = 480249.7333 S2^2 = 1951167.829
And the following are
σ1² = σ2² unknown

Test statistic: t
v= 30
S2p = 1215708.781

Critical value method p-value method

t0 = -5.6533 p-value = 0.0000037

t0,025 = -2.04

Decision

1) As t0<t 0.025
2) Like
p<α
H0 is rejected

Conclusion

There is sufficient reason to affirm that the mean of the sampled populations differs from the mean of the
sampled populations.
14. Several researchers want to know if it is possible to concludeNENL
that two populations of children
SAMPLE L
SAMPLE
differ with respect to the average age at which they can walk on their own. The researchers
obtained the following data (ages in months).
What can researchers
1 conclude?
5289 Let α=0.05. 17 9920
2 6209 18 8581
Sample population A: 9.5, 10.5, 9.0, 9.75, 10.0, 18.0, 10.0,
13.5, 10.0, 9.5, 10.0, 9.75 12.5, 9.5,
3 6054 19 9305
Sample population B: 13.5, 13.75, 12.0, 13.75, 12.5,9.5, 12.0,
13.5, 12.0, 12.0
4 6665 20 10765

What can researchers conclude? Let α=0.05. 21 8079

5 6343

6 7699 22 9046
State the statistical hypothesis 23 7134
7 5678
Ho: 8 µA-µB ≥ 6954 0 24 8736
H1: µA-µB < 0 25 10230
9 6916
26 7121
10 4770
Level of significance:
α= 0.05 5% 27 8665
11 5979
28 5176
12 6305
29 8527
13 6502
Sample and14
population data 30 7791
6113
AB
n= 12 12 31 8782
15 6347
10.4318182 12.2083333
X=
32 6883
16 5657
Assumed that the age at which children can walk on their own should be normally distributed in both populations.
σ²= ???? ????
s²= 2.101 2.100
s²p= 2.101

Test statistic: t

Critical value method p-value method

t0= t0.025= -3.0023
-2.07
p-value= 0.004 < 0.05

Decision: Decision:
As t0 < t0.025 As the p-value < α
I therefore reject H0 I therefore reject H0

CONCLUSION: It is concluded that the average age of population B at which children

can walk alone is higher in population A.
 SAMPLE A SAMPLE B
1 9.5 1 12.5
2 10.5 2 9.5
3 9 3 13.5
4 13.75
4 9.75
5 12
5 10
6 13.75
6 13 7 12.5
7 10 8 9.5
8 13.5 9 12
9 10 10 13.5
10 9.5 11 12
11 10 12 12

Test of equal variances: Test statistic: F

H0: σ²A = σ²B F0= 1.0004
H1: σ²A ≠ σ²B p-value= 0.4998 > 0.05
α= 0.05

Therefore, the population variances are equal.

15. An article published by Kashima et ai. (A-12) describes

research related to parents of mentally retarded children, in which
information is presented on the teaching of self-care in a program
supported by different media, mainly through videotapes and
instruction manuals. As part of the study, 17 families participated
in the training program provided by personnel with extensive
experience in parent training projects. Before and after the training
 program,BEFORE
a behavioral
(A) AFTER (B)and descriptive test was administered to
State the statistical hypothesis
each head
1 7of household
11 or main parent. The test assesses
2
knowledge of6 the principles
14
of behavior modification. A high score
3 10 16 Ho: µA-µB ≥ 0
indicates
4 greater
16 knowledge.
17 The following scores correspond to
H1: µA-µB < 0
the tests
5 of the
8 heads9 of household, before and after the training
program:6 13 15
Level of significance:
7 8 9
Before:8 714 6 10 1716 8 α= 8
13 14 16 11 120.05
13 9 10 5%
9 17
16 8 5 20
Afterwards: 11 11 14 1216 17 9Sample
10
15 and
9 population
17 20 data14 15 14
12
11 12
15 18 15 149
12 13 15 A B
13 9 14 n= on these data, 17 17
Is it possible to conclude, based that the training
14 10 15
program x-bar= 10.765 14.118
15 increases
17 knowledge
18 regarding the principles of behavior
modification? Let α=0.05
16 8 15
17 5 9

Testthat
Assuming of equal variances:
the variable Test statistic:
"level of knowledge regarding F
principles of behavior modification" should have normal distribution in
1.285
H0: σ²A = σ²B
both populations σ²
H1: σ²A ≠ σ²B α=
= ???? ???? F0= p-value=
0.311 > 0.05
s²= 13.316 10.360
0.05
s²p= Therefore,
11.838the
population variances are equal.
Test statistic: t

Critical value method p-value method

t0= -2.8411
t0,025= -2.04
p-value= 0.004 < 0.05

Decision: Decision:
As t0 < t0.025 As the p-value < α
I therefore reject H0 I therefore reject H0
Conclusion: Parents who received the training increased their knowledge of the principles of
behavior modification.
16. Some research suggests that among patients with primary unipolar depression there is a high rate
of alcoholism. An investigation by Winokur and COrYell (A-24) further explores this possible
relationship. Among 210 families of women with severe primary unipolar depression, the researchers
found that alcoholism was present in 89. Among 299 families categorized as control families in the
research, alcoholism was present in 94. Do these data provide sufficient evidence to conclude that
alcoholism is more likely to be found in families with members who have unipolar depression? Let
α=0.05.

A B
DEPRESSION CONTROL
ALCO 89 94 183
NOT 121 205 326
SOMETHING
TOTAL 210 299 509

Raise ith¡pi^
A=DEPRESSION
Ho: PA-PBS 0 B= NORMAL
Hl: PA-PB> 0

A B
n= 210 299

Ex 89 94

p-har= 0.42 0.31

Designation level:
d= 0.05
Critical value method
Sample and population data
zD= 2.5350

z0,05= 1.645
0.358

Decision: Decision:
SinceZ0>z0.05 As the value-p>a
p-value method
Test statistic:z
p-barprom= o-value= 0.0056216 > 0.05
Therefore, he rejected the Hl I therefore reject the Hl

CONCLUSION: it is rejected that alcoholism is more likely to be found in families with members who have unipolar depression.
17. In a study on obesity, the following results were obtained from samples of men and women
between the ages of 20 and 75 years:
n Overweight

Men 150 21
Women 200 48

Is it possible to conclude from these data that there is a difference in the proportions of overweight
individuals in the sampled populations? Let α= 0.05.

n Overweight
Men 150 21
Women 200 48

ANSWER: There is no difference in the proportions of individuals with

overweight.

To raise the lipopote sisestadisticas

A= Overweight men

He: PA-PB 0 B= Overweight women

H1: PA-PB + 0
Significance level I :
a= 0.05

A B
n= 150 200
Overweight 21 48
Sample and population data
Test statistic;:

p-bar 0.196
prom=
Critical value method p-value method
p-bar= 0.14 0.24

Z= -1 2856
z0.05= 3.290 p- 0.1905951 0.0
value= 5
Decision: Decision:
As the value-p>a
As <0.05

Therefore ,does not reject ¡a H0 Therefore I do not reject the HQ

18. A test was performed to estimate the level of distress of a sample of male patients and a sample
of female patients shortly before the same surgical procedure. The sample sizes and variances
ca1culated from the scores obtained are as follows:
Males: n = 16, s2=150

Women: n = 21, s2 = 275 Do these data provide sufficient evidence to indicate that, in
the populations represented, the scores obtained by women are more variable than those obtained by
men? Let α=0.05.

HYPOTHESIS TEST FOR THE DIFFERENCE IN POPULATION VARIANCES

State the statistical hypothesis
A= female scores
Ho: c'B B=male scores
H1: a2A > a?B
Level of significance:

a 0.05 5%
Sample and population data

A B
n= 21 16
32= 275 150
a2
X X
Test statistic : F

Critical value method

F= 1.833 p-value method
F0.05= 2.33

2.33 > 0.45 R-value= 0.050 = 0.05

Decision:
As FFa Decision:
As the p-value Sa I therefore reject the HO

19. Cell packet volume determinations were performed in two groups of children with congenital
cyanotic heart disease. Sample sizes and variances were:

Solution: 6
X20= 22.04 11
Reply 11
 _ (K,-K2)-(4_12 _ C4.5 - 3.4) - 0 _ 1.1 12
s+s11+15 12
Determine significance level: 13
p-bar prom= 19

Do these data provide sufficient evidence to conclude that the variance of population
2 is greater than the variance of population 1? Let a = . 05.

s2
Group n
1 10 40
2 16 84

ANSWER: The variance of population 2 is greater

than the variance of population 1.

HYPOTHESIS TEST FOR THE DIFFERENCE IN POPULATION VARIANCES

State the statistical hypothesis

Solution:.........................................................................................................................................6
X20= 22.04..................................................................................................................................11
Reply............................................................................................................................................11
_ (K,-K2)-(4_12 _ C4.5 - 3.4) - 0 _ 1.1...........................................................................................12
s+s11+15..........................................................................................................................................12
Determine significance level:.......................................................................................................13
p-bar prom=................................................................................................................................19

Sample andpopulation data

G2 G1

n= 16 10
s2=
84 40
q2 X X

Solution:.........................................................................................................................................6
X20= 22.04..................................................................................................................................11
Reply............................................................................................................................................11
_ (K,-K2)-(4_12 _ C4.5 - 3.4) - 0 _ 1.1...........................................................................................12
s+s11+15..........................................................................................................................................12
Determine significance level:.......................................................................................................13
p-bar prom=................................................................................................................................19

I therefore reject the HO

Decision: Decision:
As F>Fa As the p-value Sa

I therefore reject H0 I therefore reject H0