Compare means (paremetric tests)

Содержание

Слайд 2

Test of population mean vs. hypothesized value,
population standard deviation unknown

Test of population mean vs. hypothesized value, population standard deviation unknown

Слайд 5

Exercise (13)

Exercise (13)

Слайд 6

Exercise (13):
It is known that the mean Haemoglobin percent (Hb%) of adult

Exercise (13): It is known that the mean Haemoglobin percent (Hb%) of
females in a community is 89%. A researcher wanted to test whether pregnancy has a significant effect on hemoglobin level. He randomly selected 25 pregnant females and conducted measurement of their Hb level. The mean Hb% for the sample was of 86 ± 7%. The researcher selected level of significance α = 0.05. The critical value at df of 24 and level of significance 0.05 is 2.064.

Слайд 7

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 8

I. Formulate general Research Question
Does pregnancy have significant effect on mean Hb%?

I. Formulate general Research Question Does pregnancy have significant effect on mean Hb%?

Слайд 9

I. Formulate general Research Question
Does pregnancy have significant effect on mean Hb%?

I. Formulate general Research Question Does pregnancy have significant effect on mean Hb%?

Слайд 10

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 11

II. State Research Hypotheses
Ho: µ = X (pregnancy has no significant effect

II. State Research Hypotheses Ho: µ = X (pregnancy has no significant
on mean Hb%)
Ho: µ ≠ X (pregnancy has significant effect on mean Hb%)
III. What is the appropriate test statistic?
One sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 12

II. State Research Hypotheses
Ho: µ = X (pregnancy has no significant effect

II. State Research Hypotheses Ho: µ = X (pregnancy has no significant
on mean Hb%)
Ho: µ ≠ X (pregnancy has significant effect on mean Hb%)
III. What is the appropriate test statistic?
One sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 13

II. State Research Hypotheses
Ho: µ = X (pregnancy has no significant effect

II. State Research Hypotheses Ho: µ = X (pregnancy has no significant
on mean Hb%)
Ho: µ ≠ X (pregnancy has significant effect on mean Hb%)
III. What is the appropriate test statistic?
One sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 14

II. State Research Hypotheses
Ho: µ = X (pregnancy has no significant effect

II. State Research Hypotheses Ho: µ = X (pregnancy has no significant
on mean Hb%)
Ho: µ ≠ X (pregnancy has significant effect on mean Hb%)
III. What is the appropriate test statistic?
One sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 15

Design Research Hypotheses and Experiment

Design Research Hypotheses and Experiment

Слайд 16

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 17

V. Calculate test statistic
VI. Make a Decision regarding Research Hypotheses
(Specify the

V. Calculate test statistic VI. Make a Decision regarding Research Hypotheses (Specify
Decision Method)
Reject Null Hypothesis Ho
Test statistic (t=2.14) > critical value (2.064)

Слайд 18

V. Calculate test statistic
VI. Make a Decision regarding Research Hypotheses
(Specify the

V. Calculate test statistic VI. Make a Decision regarding Research Hypotheses (Specify
Decision Method)
Reject Null Hypothesis Ho
Test statistic (t=2.14) > critical value (2.064)


(86 - 89)
7
√ 25

t =

= 2.14

Слайд 19

V. Calculate test statistic
VI. Make a Decision regarding Research Hypotheses
(Specify the

V. Calculate test statistic VI. Make a Decision regarding Research Hypotheses (Specify
Decision Method)
Reject Null Hypothesis Ho
Test statistic (t=2.14) > critical value (2.064)
(Critical value method)


(86 - 89)
7
√ 25

t =

= 2.14

Слайд 20

Collect and Analyze Experimental Data

Collect and Analyze Experimental Data

Слайд 22

Reject Ho

Fail to Reject Ho

Fail to Reject Ho

Reject Ho

+1.96

-1.96

µ
89%

X
86%

Reject Ho Fail to Reject Ho Fail to Reject Ho Reject Ho

Слайд 23

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 24

VII. Report a conclusion
Pregnancy has a significant effect on mean Hb%.
The mean

VII. Report a conclusion Pregnancy has a significant effect on mean Hb%.
Hb% of pregnant females (86%) was significantly lower than the mean Hb% of adult females in the community (89%).

Слайд 25

VII. Report a conclusion
Pregnancy has a significant effect on mean Hb%.
The mean

VII. Report a conclusion Pregnancy has a significant effect on mean Hb%.
Hb% of pregnant females (86%) was significantly lower than the mean Hb% of adult females in the community (89%).

Слайд 26

Comparing Means
(Parametric tests)
One Population Inference
Two Population Inference
Independent Sampling Model
Dependent Sampling Model
Analysis

Comparing Means (Parametric tests) One Population Inference Two Population Inference Independent Sampling
of Variance (ANOVA)

Слайд 27

Sample 1

Sample 2


The data is collected by two simple random samples

Sample 1 Sample 2 The data is collected by two simple random
from separate and unrelated populations. This data will then be used to compare the two population means. This is typical of an experimental or treatment population versus a control population.

Слайд 28

Sample 1

Sample 2

n1
X1
S1

n2
X2
S2

Sample 1 Sample 2 n1 X1 S1 n2 X2 S2

Слайд 29

Exercise (14)

Exercise (14)

Слайд 30

Researchers were motivated to test a new antihypertensive drug (A) on a

Researchers were motivated to test a new antihypertensive drug (A) on a
group of patients. They needed to know whether Drug (A) achieves significant reduction in the systolic blood pressure compared with the conventional antihypertensive drug (B).
In the current research, 200 randomly selected patients suffering from essential hypertension and fulfilled the inclusion and exclusion criteria were included. The participants were randomly allocated into two groups; 100 patients were given drug (A) and 100 were given the drug (B). The researchers selected level of significance α = 0.05.
After a period of 10 weeks, the mean systolic blood pressure of the first group receiving drug (A) decreased by 12 + 2.36 mm Hg while that of the second group decreased by 9 + 5.69. The collected data were typed onto computer and analyzed using SPSS software program. The program revealed the value of the test statistic=2.56 (p=0.06)

Слайд 31

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 32

I. Formulate general Research Question
Does pregnancy have significant effect on mean Hb%?

I. Formulate general Research Question Does pregnancy have significant effect on mean Hb%?

Слайд 33

I. Formulate general Research Question
Is there any significant difference in the mean

I. Formulate general Research Question Is there any significant difference in the
reduction of systolic blood pressure achieved by drug (A) and drug (B)?

Слайд 34

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 35

II. State Research Hypotheses
Ho: X1 = X2
There is no significant difference

II. State Research Hypotheses Ho: X1 = X2 There is no significant
in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
Ha: X1 ≠ X2
There is a significant difference in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
III. What is the appropriate test statistic?
Two Independent sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 36

II. State Research Hypotheses
Ho: X1 = X2
There is no significant difference

II. State Research Hypotheses Ho: X1 = X2 There is no significant
in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
Ha: X1 ≠ X2
There is a significant difference in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
III. What is the appropriate test statistic?
Two Independent sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 37

II. State Research Hypotheses
Ho: X1 = X2
There is no significant difference

II. State Research Hypotheses Ho: X1 = X2 There is no significant
in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
Ha: X1 ≠ X2
There is a significant difference in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
III. What is the appropriate test statistic?
Two Independent sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model


(X1 – X2)
S2p + S2p
n1 n2

t =

Слайд 38

II. State Research Hypotheses
Ho: X1 = X2
There is no significant difference

II. State Research Hypotheses Ho: X1 = X2 There is no significant
in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
Ha: X1 ≠ X2
There is a significant difference in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)
III. What is the appropriate test statistic?
Two Independent sample t test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model


(X1 – X2)
S2p + S2p
n1 n2

t =

Слайд 39

Design Research Hypotheses and Experiment

Design Research Hypotheses and Experiment

Слайд 40

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 41

V. Make a Decision regarding Research Hypotheses
(Specify the Decision Method)
Fail

V. Make a Decision regarding Research Hypotheses (Specify the Decision Method) Fail
to Reject Null Hypothesis Ho
P-value (0.6) > α (0.05)
(p-value method)

Слайд 42

Collect and Analyze Experimental Data

Collect and Analyze Experimental Data

Слайд 43

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 44

VI. Report a conclusion
Pregnancy has a significant effect on mean Hb%.
The mean

VI. Report a conclusion Pregnancy has a significant effect on mean Hb%.
Hb% of pregnant females (86%) was significantly lower than the mean Hb% of adult females in the community (89%).

Слайд 45

VI. Report a conclusion
There is insufficient evidence to support the claim that

VI. Report a conclusion There is insufficient evidence to support the claim
there is a significant difference in the mean reduction of systolic blood pressure achieved by drug (A) and drug (B)

Слайд 46

Drug (A)

Drug (B)

n1
X1
S1

n2
X2
S2

Drug (A) Drug (B) n1 X1 S1 n2 X2 S2

Слайд 47

+1.96

-1.96

X1
9

X2
12

Fail to Reject Ho

Reject Ho

Reject Ho

Fail to Reject Ho

+1.96 -1.96 X1 9 X2 12 Fail to Reject Ho Reject Ho

Слайд 48

Comparing Means
(Parametric tests)
One Population Inference
Two Population Inference
Independent Sampling Model
Dependent Sampling Model
Analysis

Comparing Means (Parametric tests) One Population Inference Two Population Inference Independent Sampling
of Variance (ANOVA)

Слайд 49

Sample


The data consists of a single population and two measurements. A

Sample The data consists of a single population and two measurements. A
simple random sample is taken from the population and pairs of measurement are collected. This is also called related sampling or matched pair design.

Слайд 50

Xd is the sample mean of the differences of each pair
Sd is

Xd is the sample mean of the differences of each pair Sd
the sample standard deviation of the differences of each pair

Слайд 51


Matched pairs t-test
compares the means for two dependent populations
(paired

Matched pairs t-test compares the means for two dependent populations (paired difference
difference t-test)
Model Assumptions
Variable is quantitative continuous
Data is normally distributed
Dependent sampling

Слайд 52


Matched pairs t-test
compares the means for two dependent populations
(paired

Matched pairs t-test compares the means for two dependent populations (paired difference
difference t-test)
Test Statistic

Xd sample mean of the differences of each pair
Sd sample standard deviation of the differences of each pair

Слайд 53

Exercise (15)

Exercise (15)

Слайд 54


Exercise (15):
An instructor of Anatomy course wants to know if student

Exercise (15): An instructor of Anatomy course wants to know if student
marks are different on the second midterm compared to the first exam after implementation of a new teaching intervention; TBL (team-based learning). The first and second midterm marks for 35 students were taken and the mean difference in marks is determined.
Data were typed and analyzed using SPSS software program. The level of significance was 0.05. The appropriate statistical test was conducted and revealed test statistic = 3.23 (p=0.004). The following is SPSS output tables

Слайд 55

Xd is the sample mean of the differences of each pair (2.05)
Sd

Xd is the sample mean of the differences of each pair (2.05)
is the sample standard deviation of the differences of each pair

Before mark

After mark

Слайд 57


Data were typed and analyzed using SPSS software program. The level

Data were typed and analyzed using SPSS software program. The level of
of significance was 0.05. The appropriate statistical test was conducted and revealed test statistic = 3.23 (p=0.004). The followings are SPSS output tables.

35
35

Слайд 58

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 59

I. Formulate general Research Question
Does pregnancy have significant effect on mean Hb%?

I. Formulate general Research Question Does pregnancy have significant effect on mean Hb%?

Слайд 60

I. Formulate general Research Question
Is there a difference in students’ marks following

I. Formulate general Research Question Is there a difference in students’ marks
implementation of TBL (Team-based Learning)?

Слайд 61

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 62

II. State Research Hypotheses
Ho: There is no difference in mean pre- and

II. State Research Hypotheses Ho: There is no difference in mean pre-
post-TBL marks
H1: There is a difference in mean pre- and post-TBL marks
III. What is the appropriate test statistic?
Matched pairs t-test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 63

II. State Research Hypotheses
Ho: There is no difference in mean pre- and

II. State Research Hypotheses Ho: There is no difference in mean pre-
post-TBL marks
H1: There is a difference in mean pre- and post-TBL marks
III. What is the appropriate test statistic?
Matched pairs t-test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 64

II. State Research Hypotheses
Ho: There is no difference in mean pre- and

II. State Research Hypotheses Ho: There is no difference in mean pre-
post-TBL marks
H1: There is a difference in mean pre- and post-TBL marks
III. What is the appropriate test statistic?
Matched pairs t-test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 65

II. State Research Hypotheses
Ho: There is no difference in mean pre- and

II. State Research Hypotheses Ho: There is no difference in mean pre-
post-TBL marks
H1: There is a difference in mean pre- and post-TBL marks
III. What is the appropriate test statistic?
Matched pairs t-test
IV. What is the appropriate test Model? (One or Two tailed)
Two Tailed Test Model

Слайд 66

Design Research Hypotheses and Experiment

Design Research Hypotheses and Experiment

Слайд 67

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 68

Collect and Analyze Experimental Data

Collect and Analyze Experimental Data

Слайд 69

V. Make a Decision regarding Research Hypotheses
(Specify the Decision Method)
Reject Null

V. Make a Decision regarding Research Hypotheses (Specify the Decision Method) Reject
Hypothesis Ho
P-value (0.004) < α (0.05)
(p-value method)

Слайд 70


Data were typed and analyzed using SPSS software program. The level

Data were typed and analyzed using SPSS software program. The level of
of significance was 0.05. The appropriate statistical test was conducted and revealed test statistic = 3.23 (p=0.004). The followings are SPSS output tables.

35
35

Слайд 71

V. Make a Decision regarding Research Hypotheses
(Specify the Decision Method)
Reject Null

V. Make a Decision regarding Research Hypotheses (Specify the Decision Method) Reject
Hypothesis Ho
P-value (0.004) < α (0.05)
(p-value method)

Слайд 72

Formulate general Research Question

Procedures of Hypotheses Testing and the Scientific Method

Design

Formulate general Research Question Procedures of Hypotheses Testing and the Scientific Method
research hypotheses and Experiment

Collect and analyze Experimental data

Report conclusions in Non-statistical languages

Слайд 73

VI. Report a conclusion
Pregnancy has a significant effect on mean Hb%.
The mean

VI. Report a conclusion Pregnancy has a significant effect on mean Hb%.
Hb% of pregnant females (86%) was significantly lower than the mean Hb% of adult females in the community (89%).

Слайд 74

VI. Report a conclusion
There is a strong evidence (t = 3.23, p

VI. Report a conclusion There is a strong evidence (t = 3.23,
= 0.004) that TBL as a teaching intervention improves students’ marks. In this data set, it improved marks, on average, by approximately 2 points (mean paired difference =2.05).

Слайд 75

Xd is the sample mean of the differences of each pair (2.05)
Sd

Xd is the sample mean of the differences of each pair (2.05)
is the sample standard deviation of the differences of each pair

Before mark

After mark

Слайд 76

+1.96

-1.96

X1
18.40

X2
20.45

Fail to Reject Ho

Reject Ho

Reject Ho

Fail to Reject Ho

+1.96 -1.96 X1 18.40 X2 20.45 Fail to Reject Ho Reject Ho

Слайд 77

Comparing Means
(Parametric tests)
One Population Inference
Two Population Inference
Independent Sampling Model
Dependent Sampling Model
Analysis

Comparing Means (Parametric tests) One Population Inference Two Population Inference Independent Sampling
of Variance (ANOVA)

Слайд 78

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Suppose we wanted to compare the means of more than two (k) independent populations and want to test the null hypothesis ??: ?? = ?? = ⋯ = ??.
If we can assume all population variances are equal, we can expand the pooled variance t-test for two populations to one factor ANOVA for k populations.

Слайд 79

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

If ??: ?1 = ?2 = ?3 is true, then each population would have the same distribution and the variance of the combined data would be approximately the same.
If the ?? is false, then the difference between centers would cause the combined data to have an increased variance.

Слайд 80

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Data Requirements
Dependent variable that is continuous.
Independent variable (Factor) that is categorical (≥ 3 groups)
Cases that have values on both the dependent and independent variables

Слайд 81

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Data Requirements
Independent samples/groups (i.e., independence of observations); there is no relationship between the subjects in each sample.
This means that:
subjects in the first group cannot also be in the second group
no subject in either group can influence subjects in the other group
no group can influence the other group

Слайд 82

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Data Requirements
Random sample of data from the population.
Normal distribution (approximately) of the dependent variable for each group (i.e., for each level of the factor)
Homogeneity of variances (i.e., variances approximately equal across groups)
No outliers

Слайд 83

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations
Model Assumptions
The populations being sampled are normally distributed
The populations have equal standard deviations
The samples are randomly selected and are independent

Слайд 84

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Research Hypotheses
H0: µ1 = µ2 = µ3  = ... = µk
("all k population means are equal")
Ha: At least one µi different  
("at least one of the k population means is not equal to the others")

Слайд 85

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Test Statistic
SSFactor = the regression sum of squares; SSError = the error sum of squares
SSTotal = the total sum of squares (SST = SSR + SSE)
k = the total number of groups; n = the total number of valid observations
MSFactor = the regression mean square; MSError = the mean square error

Слайд 86

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 87

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Examine descriptive statistics
Check for outliers
Check that the normality assumption is met
Verify that there are mean differences between groups to justify ANOVA

Conduct an exploratory analysis

Слайд 88

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 89

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Determine whether group means are different form one another (warranting post hoc comparison tests)
Check that the homogeneity of variance assumption is met

Conduct One Way ANOVA

Слайд 90

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 91

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
To confirm where the differences occurred between groups

Conduct Post hoc comparison test

Слайд 92

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 93

Exercise (16)

Exercise (16)

Слайд 94

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Exercise (16)
A research was conducted to examine if there a difference in mean score of medical students at 4 universities located at different geographical locations (North-East, North-Central, South, and West region).
In order to conduct the research, 400 randomly selected students from the 4 universities were included, and their scores were reported. Data were typed and analyzed by SPSS software program. The statistician selected level of significance = 0.05.

Слайд 95

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

I. State the Research Question
Is there any difference in the mean score of medical students at North-East, North-Central, South, and West Universities?

Слайд 96

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

I. State the Research Question
Are there any differences in the mean score of medical students enrolled at North-East, North-Central, South, and West Universities?

Слайд 97

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

II. State the Research Hypotheses
H0: µ1 = µ2 = µ3 = µ4  
µ1: mean score of students at North-East University
µ2: mean score of students at North-Central University
µ3: mean score of students at North-East University
µ4: mean score of students at North-East University
Ha: At least one µi different  
("at least one of the 4 population means is not equal to the others")

Слайд 98

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

II. State the Research Hypotheses
H0: µ1 = µ2 = µ3 = µ4  
µ1: mean score of students at North-East University
µ2: mean score of students at North-Central University
µ3: mean score of students at North-East University
µ4: mean score of students at North-East University
Ha: At least one µi different  
("at least one of the 4 population means is not equal to the others")

Слайд 99

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

III. Specify the Dependent and Independent variables
Mention the type of variable and number of groups

Слайд 100

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

III. Specify the Dependent and Independent variables
Mention the type of variable and number of groups

Слайд 101

Exercise (17)

Exercise (17)

Слайд 102

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Exercise (17)
A manager of a pharmaceutical company wants to raise the productivity at his company by increasing the speed at which his pharmacists carry out pharmaceutical formulation.
As he does not have the skills in-house, he employs an external agency which provides training in pharmaceutical formulation Development. They offer 3 packages - a beginner, intermediate and advanced course.
He is unsure which course is needed for the type of work they do at his company so he sends 10 pharmacists on the beginner course, 10 on the intermediate and 10 on the advanced course.

Слайд 103

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Exercise (17)
When they all return from the training he gives them a task to formulate certain parenteral drug produced by the company and times how long it takes them to complete the task.
He wishes to then compare the three courses (beginner, intermediate, advanced) to see if there are any differences in the average time it took to complete the task.
The statistician selected level of significance = 0.05

Слайд 104

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

I. State the Research Question
Are there any differences in the average time taken by the pharmacists who attended the three training courses (Beginner, Intermediate, and Advanced courses), to complete the drug formulation task?

Слайд 105

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

I. State the Research Question
Are there any differences in the average time taken by the pharmacists who attended the three training courses (Beginner, Intermediate, and Advanced courses), to complete the drug formulation task?

Слайд 106

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

II. State the Research Hypotheses
H0: µ1 = µ2 = µ3  
µ1: Mean time (hour) taken by the pharmacists who attended the beginner course to complete the task.
µ2: Mean time (hour) taken by the pharmacists who attended the intermediate course to complete the task.
µ3: Mean time (hour) taken by the pharmacists who attended the advanced course to complete the task.
Ha: At least one µi different  
("at least one of the 3 population means is not equal to the others")

Слайд 107

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

II. State the Research Hypotheses
H0: µ1 = µ2 = µ3  
µ1: Mean time (hour) taken by the pharmacists who attended the beginner course to complete the task.
µ2: Mean time (hour) taken by the pharmacists who attended the intermediate course to complete the task.
µ3: Mean time (hour) taken by the pharmacists who attended the advanced course to complete the task.
Ha: At least one µi different  
("at least one of the 3 population means is not equal to the others")

Слайд 108

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

III. Specify the Dependent and Independent variables
Mention the type of variable and number of groups

Слайд 109

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

III. Specify the Dependent and Independent variables
Mention the type of variable and number of groups

Слайд 110

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 111

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Examine descriptive statistics
Check for outliers
Check that the normality assumption is met
Verify that there are mean differences between groups to justify ANOVA

Conduct an exploratory analysis

Слайд 112

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 113

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

100
100
100
100
400

Слайд 114

Box plot graph

Box plot graph

Слайд 116

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Examine descriptive statistics
Check for outliers
Check that the normality assumption is met
Verify that there are mean differences between groups to justify ANOVA

Conduct an exploratory analysis

Слайд 119

Tests of Normality

Tests of Normality

Слайд 120

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 122

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 123

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Examine descriptive statistics
Check for outliers
Check that the normality assumption is met
Verify that there are mean differences between groups to justify ANOVA

Conduct an exploratory analysis

Слайд 124

Error Bar graph

Error Bar graph

Слайд 126

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Examine descriptive statistics
Check for outliers
Check that the normality assumption is met
Verify that there are mean differences between groups to justify ANOVA

Conduct an exploratory analysis

Слайд 127

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 128

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Determine whether group means are different form one another (warranting post hoc comparison tests)
Check that the homogeneity of variance assumption is met

Conduct One Way ANOVA

Слайд 136

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
Determine whether group means are different form one another (warranting post hoc comparison tests)
Check that the homogeneity of variance assumption is met

Conduct One Way ANOVA

Слайд 137

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 138

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
To confirm where the differences occurred between groups

Conduct Post hoc comparison test

Слайд 144

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Why?
To confirm where the differences occurred between groups

Conduct Post hoc comparison test

Слайд 145

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 146

Return to Exercise (16)

Return to Exercise (16)

Слайд 149

100
100
100
100
400

100 100 100 100 400

Слайд 150

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

IV. Make your Decision regarding Research Hypotheses
kjbj

Слайд 152

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

IV. Make your Decision regarding Research Hypotheses
Reject Null Hypothesis H0 
p-value of F statistic = 0.002
P-value < α
0.002 < 0.05

Слайд 153

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

V. Report a conclusion
The mean score of medical students at North-East University was significantly higher (55.89±9.86) than the mean score of students at North-Central, West, and South Universities (50.83±10.01, 49.03±10.00, 51.33±9.8 respectively).

Слайд 155

100
100
100
100
400

100 100 100 100 400

Слайд 156

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

V. Report a conclusion
The mean score of medical students at North-East University was significantly higher (55.89±9.86) than the mean score of students at North-Central, West, and South Universities (50.83±10.01, 49.03±10.00, 51.33±9.8 respectively).

Слайд 157

Return to Exercise (17)

Return to Exercise (17)

Слайд 161

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

IV. Make your Decision regarding Research Hypotheses
Reject Null Hypothesis H0 
p-value of F statistic = 0.021
P-value < α
0.021 < 0.05

Слайд 163

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

IV. Make your Decision regarding Research Hypotheses
Reject Null Hypothesis H0 
p-value of F statistic = 0.021
P-value < α
0.021 < 0.05

Слайд 164

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

V. Report a conclusion
Pharmacists who attended the Beginner course spent significantly longer duration of time to formulate the drug (27.20±3.04 hours) compared with Pharmacists who attended the Intermediate and the Advanced courses (23.60±3.30 hours, and 23.40±3.23 hours respectively).

Слайд 167

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

V. Report a conclusion
Pharmacists who attended the Beginner course spent significantly longer duration of time to formulate the drug (27.20±3.04 hours) compared with Pharmacists who attended the Intermediate and the Advanced courses (23.60±3.30 hours, and 23.40±3.23 hours respectively).

Слайд 168

One Factor ANOVA model - One Way ANOVA Comparing means from more

One Factor ANOVA model - One Way ANOVA Comparing means from more
than two Independent Populations

Run a One Way ANOVA (SPPS’s One Way ANOVA Procedure)

Слайд 169

Comparing Means
(Parametric tests)
One Population Inference
Two Population Inference
Independent Sampling Model
Dependent Sampling Model
Analysis

Comparing Means (Parametric tests) One Population Inference Two Population Inference Independent Sampling
of Variance (ANOVA)
Имя файла: Compare-means-(paremetric-tests).pptx
Количество просмотров: 40
Количество скачиваний: 0