Dr.

Abu Awwad - AAUP

Arab American University, Palestine (AAUP)

HEALTHCARE STATISTICS- 152526030

Topic 6: Inferences From Two Samples

Instructor: Dr. Abdul Fattah Abu Awwad

[Link]@[Link]

1 / 39
 Dr. Abu Awwad - AAUP

Outline

1 Learning Objectives

2 Two Proportions

3 Two Means: Independent Samples

4 Two Dependent Samples (Matched Pairs)

2 / 39
 Dr. Abu Awwad - AAUP Learning Objectives

Learning Objectives

After studying this chapter, the student will:

1 Distinguish between dependent and independent samples.

2 Construct and interpret confidence intervals about the difference between

two population proportions.

3 Test hypotheses regarding paired (dependent) samples.

4 Construct and interpret confidence intervals about the mean difference of

paired samples.

5 Test hypotheses regarding independent samples.

6 Construct and interpret confidence intervals about the mean difference of

independent samples.

7 Test the equality of two variances.

3 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Two Proportions

Notations:
X1 x1
1 For population 1: p1 = N1
is the population proportion and p̂1 = n1
is the
sample proportion.
X2 x2
2 For population 2: p2 = N2
is the population proportion and p̂2 = n2
is the
sample proportion.

3 The pooled sample proportion is denoted by

n1 p̂1 + n2 p̂2 x1 + x2
p̂ = = .
n1 + n2 n1 + n2
Requirements:
1 The sample proportions are from two simple random samples.

2 The two samples are independent.

3 For each of the two samples, there are at least 5 successes and at least 5
failures. (That is, np̂ ≥ 5 and n(1 − p̂) ≥ 5 for each of the two samples).

4 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Two Proportions: Hypothesis Testing

Example: The table below lists results from a simple random sample of
front-seat occupants involved in car crashes. Use a 0.05 significance level to
test the claim that the fatality rate of occupants is lower for those in cars
equipped with airbags.

41
p̂1 = 11541 ≈ 0.0036: is the estimated proportion of fatality of occupants
in cars equipped with airbags.

52
p̂2 = 9853 ≈ 0.0053: is the estimated proportion of fatality of occupants in
cars without airbags.

41+52
p̄ = 11541+9853 ≈ 0.0043: is the overall (common) estimated proportion of
fatality.

5 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Two Proportions: Hypothesis Testing

6 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Two Proportions: Hypothesis Testing

The null and alternative hypotheses are:

The level of significance is

The p-value is

The conclusion is

7 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Two Proportions: Confidence Interval

Example: Use the sample data given in the previous example to construct a
90% confidence interval estimate of the difference between the two population
proportions. What does the result suggest about the claim that "the fatality
rate of occupants is lower for those in cars equipped with airbags"?
We are 90% confident that the amount of the difference between the two
proportions of fatality lies within the interval (−0.003, −0.0002).

We can translate these numbers into rates by multiplying by some

population size, say, 10000. Using the 90% confidence interval, we
estimate that drivers without airbags are at increased risk of a fatal car
accident by somewhere between 2 and 30 additional fatalities per 10000
occupants.

The confidence interval limits do not contain 0, suggesting that there is a

significant difference between the two proportions.

The confidence interval suggests that the fatality rate is lower for
occupants in cars with airbags than for occupants in cars without airbags.

8 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Two Proportions: Confidence Interval

9 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Application
Example: Refer to Data Set 1 "Body Data" test whether the proportion of
obese female (BMI ≥ 30) is not different form males. Use a 0.05 significance
level. What do you conclude?

SPSS : Transform ⇒ Recode into different variables.

Underweight BMI ≤ 18.4
Normal 18.5 ≤ BMI ≤ 24.9
Overweight 25 ≤ BMI ≤ 29.9
Obese BMI ≥ 30

SPSS : Analyze ⇒ Descriptive Statistics ⇒ Crosstabs

10 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Application

11 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Application

12 / 39
 Dr. Abu Awwad - AAUP Two Proportions

Chi-square test of independence

13 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Notations:
PN1
X1i
i=1
1 For population 1: µ1 = N1
is the population mean and
Pn1
X1i
i=1
X̄1 = n1
is the sample mean.
PN2
X2i
i=1
2 For population 2: µ2 = N2
is the population mean and
Pn2
X2i
i=1
X̄2 = n2
is the sample mean.

Requirements:
1 Both samples are simple random samples.

2 The two samples are independent.

3 Either or both of these conditions are satisfied: The two sample sizes are
both large (with n1 > 30 and n2 > 30) or both samples come from
populations having normal distributions.

14 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Example: Two competing headache remedies claim to give fast-acting relief.

An experiment was performed to compare the mean lengths of time required
for bodily absorption of brand A and brand B headache remedies. Twelve
people were randomly selected and given an oral dosage of brand A. Another
12 were randomly selected and given an equal dosage of brand B. The lengths
of time in minutes for the drugs to reach a specified level in the blood were
recorded. The means, standard deviations, and sizes of the two samples follow.

Past experience with the drug composition of the two remedies permits
researchers to assume that both distributions are approximately normal with
equal variances. Let us use a 5% level of significance to test the claim that
there is no difference in the mean time required for bodily absorption.

15 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

16 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

The null and alternative hypotheses are:

The level of significance is

The p-value is

The conclusion is

Q. Construct a 95% confidence interval for the difference between the two
mean lengths of time for brand A versus brand B. Interpret your result.

17 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Example: Suppose in Data Set 1 "Body Data" a researcher wishes to test

whether the mean pulse rate for female is higher than for male.

In SPSS : Analyze ⇒ Compare means ⇒ Independent-Samples T-Test.

Descriptive statistics

18 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Equality of variances test (Levene’s Test)

The null and alternative hypotheses are:

The level of significance is

The p-value is

The conclusion is

19 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

t-test for equality of means

The null and alternative hypotheses are:

The level of significance is

The p-value is

The conclusion is

20 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

21 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Example: Can we conclude that, on the average, lymphocytes and tumor cells
differ in size? The following are the cell diameters (µm) of 40 lymphocytes and
50 tumor cells obtained from biopsies of tissue from patients with melanoma:

22 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Example: Can we conclude that patients with primary hypertension (PH), on

the average, have higher total cholesterol levels than normotensive (NT)
patients? This was one of the inquiries of interest for Rossi et al., 2003. The
data provide the total cholesterol measurements (mg/dl) for 133 PH patients
and 41 NT patients. Can we conclude that PH patients have, on average,
higher total cholesterol levels than NT patients? Let α = 0.05.

Rossi, G. P., Taddei, S., Virdis, A., Cavallin, M., Ghiadoni, L., Favilla, S., ... &
Salvetti, A. (2003). The T-786C and Glu298Asp polymorphisms of the
endothelial nitric oxide gene affect the forearm blood flow responses of
Caucasian hypertensive patients. Journal of the American College of
Cardiology, 41(6), 938-945.

23 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Example: Reichman et al. (1996) conducted a study with the purpose of

demonstrating that negative symptoms are prominent in patients with
Alzheimer’s disease and are distinct from depression. The following are scores
made on the Scale for the Assessment of Negative Symptoms in Alzheimer’s
Disease by patients with Alzheimer’s disease (PT) and normal elderly,
cognitively intact, comparison subjects (C). Let α = 0.05.
(Discussion: Slides 24 - 26)

Reichman, W. E., Coyne, A. C., Amirneni, S., Molino, B., & Egan, S. (1996).
Negative symptoms in Alzheimer’s disease. The American journal of psychiatry.

24 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Assessing Normality

25 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Descriptive Analysis- Graphical Methods

26 / 39
 Dr. Abu Awwad - AAUP Two Means: Independent Samples

Two Means: Independent Samples

Testing

27 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Two samples are said to be paired when each data point in the first sample
is matched and is related to a unique data point in the second sample.

Paired t-Test is used to determine whether there is statistical evidence

that the mean difference between paired observations is significantly
different from zero.

Requirements:
1 The sample data are dependent (matched pairs).

2 The matched pairs are a simple random sample.

3 Either or both of these conditions are satisfied: The number of pairs of

sample data is large (n > 30) or the pairs of values have differences that are
from a population having a distribution that is approximately normal.

28 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Example: Woo and McKenna (2003) investigated the effect of broadband

ultraviolet B (UVB) therapy and topical calcipotriol cream used together on
areas of psoriasis. One of the outcome variables is the Psoriasis Area and
Severity Index (PASI). The following table gives the PASI scores for 20 subjects
measured at baseline and after eight treatments. Do these data provide
sufficient evidence, at the .01 level of significance, to indicate that the
combination therapy reduces PASI scores?
(Discussion: Slides 29 - 31)

Woo, W. K., & McKenna, K. E. (2003). Combination TL01 ultraviolet B

phototherapy and topical calcipotriol for psoriasis: a prospective randomized
placebo-controlled clinical trial. British Journal of Dermatology, 149(1),
146-150.

29 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Data

30 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Assessing Normality

31 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Testing

32 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Example: Hartard et al. (1996) conducted a study to determine whether a

certain training regimen can counteract bone density loss in women with
postmenopausal osteopenia. The following are strength measurements for five
muscle groups (Leg Press, Hip Flexor, Hip Extensor, Arm Abductor, and
Arm Adductor) taken on 15 subjects before (B) and after (A) 6 months of
training. Let α = 0.05.
(Discussion: Slides 33 - 34)

Hartard, M., Haber, P., Ilieva, D., Preisinger, E., Seidl, G., & Huber, J. (1996).
SYSTEMATIC STRENGTH TRAINING AS A MODEL OF THERAPEUTIC
INTERVENTION: A Controlled Trial in Postmenopausal Women with
Osteopenia: 1. American journal of physical medicine & rehabilitation, 75(1),
21-28.

33 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Assessing Normality

34 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Testing

35 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

Example: Table below lists body temperatures of five subjects at 8 AM and at
12 AM. With a 0.05 significance level test the claim that there is no difference
in body temperatures measured at 8 AM and at 12 AM.

In SPSS : Analyze ⇒ Compare means ⇒ Paired-Samples T-Test.

Descriptive statistics

36 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

37 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

Two Dependent Samples (Matched Pairs)

The null and alternative hypotheses are:

The level of significance is

The p-value is

The conclusion is

Q. Construct a 95% confidence interval estimate of the mean of the

temperature differences. Interpret your result.

38 / 39
 Dr. Abu Awwad - AAUP Two Dependent Samples (Matched Pairs)

The End

39 / 39