0% found this document useful (0 votes)

34 views174 pages

Descriptive Statistics Overview

Uploaded by

keskin

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

34 views174 pages

Descriptive Statistics Overview

Uploaded by

keskin

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

Applied Statistics and Probability for

Engineers

Sixth Edition
Douglas C. Montgomery George C. Runger

Chapter 6
Descriptive Statistics

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Numerical Summaries of Data
• Statistics id the science of data.
• An important aspect of dealing with data is
organizing and summarizing the data in
ways that facilitate its interpretations and
subsequent analysis (descriptive stats).

Sec 6-1 Numerical Summaries of Data 2

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Numerical Summaries of Data
• Data are the numeric observations of a
phenomenon of interest. The totality of all
observations is a population. A portion used
for analysis is a random sample.
• We gain an understanding of this collection,
possibly massive, by describing it numerically
and graphically, usually with the sample data.
• We describe the collection in terms of shape,
outliers, center, and spread (SOCS).
• The center is measured by the mean.
• The spread is measured by the variance.

Sec 6-1 Numerical Summaries of Data 3

Sec 6-1 Numerical Summaries of Data 4

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 6-1: Sample Mean
Consider 8 observations (xi) of pull-off force from
engine connectors as shown in the table.
8

x i
12.6  12.9  ...  13.1
i
1
xi
12.6
x  average  i 1
 2 12.9
8 8 3 13.4
104 4 12.3
  13.0 pounds 5 13.6
8 6 13.5
7 12.6
8 13.1
13.00
= AVERAGE($B2:$B9)

Figure 6-1 The sample mean is the balance point.

Sec 6-1 Numerical Summaries of Data 5
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Variance Defined

Sec 6-1 Numerical Summaries of Data 6

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Standard Deviation Defined
• The standard deviation is the square root
of the variance.
• σ is the population standard deviation
symbol.
• s is the sample standard deviation symbol.

Sec 6-1 Numerical Summaries of Data 7

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 6-2: Sample Variance
Table 6-1 displays the quantities needed to calculate the
sample variance and sample standard deviation.
i xi x i - xbar (x i - xbar) 2
1 12.6 -0.4 0.16
2 12.9 -0.1 0.01
Dimension of: 3 13.4 0.4 0.16
xi is pounds 4 12.3 -0.7 0.49
Mean is pounds. 5 13.6 0.6 0.36
Variance is pounds2. 6 13.5 0.5 0.25
Standard deviation is pounds. 7 12.6 -0.4 0.16
8 13.1 0.1 0.01
Desired accuracy is generally sums = 104.00 0.0 1.60
accepted to be one more place divide by 8 divide by 7
than the data. xbar = 13.00 variance = 0.2286
standard deviation = 0.48

Table 6‐1

Sec 6-1 Numerical Summaries of Data 8

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Computation of s2
The prior calculation is definitional and tedious. A
shortcut is derived here and involves just 2 sums.

Sec 6-1 Numerical Summaries of Data 9


i 1
x    xi 

2
i
i 1 
n i xi xi2
s 
2 1 12.6 158.76
n 1 2 12.9 166.41
3 13.4 179.56
1,353.60  104.0  8
2 4 12.3 151.29
 5 13.6 184.96
7 6 13.5 182.25
7 12.6 158.76
1.60 8 13.1 171.61
  0.2286 pounds 2 sums = 104.0 1,353.60
7

s  0.2286  0.48 pounds

Sec 6-1 Numerical Summaries of Data 10

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
What is this “n–1”?
• The population variance is calculated with N,
the population size. Why isn’t the sample
variance calculated with n, the sample size?
• The true variance is based on data deviations
from the true mean, μ.
• The sample calculation is based on the data
deviations from x-bar, not μ. X-bar is an
estimator of μ; close but not the same. So
the n-1 divisor is used to compensate for the
error in the mean estimation.
Sec 6-1 Numerical Summaries of Data 11
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Degrees of Freedom
• The sample variance is calculated with the
quantity n-1.
• This quantity is called the “degrees of
freedom”.
• Origin of the term:
– There are n deviations from x-bar in the sample.
– The sum of the deviations is zero.
– n-1 of the observations can be freely determined,
but the nth observation is fixed to maintain the
zero sum.

Sec 6-1 Numerical Summaries of Data 12

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Sample Range
If the n observations in a sample are denoted
by x1, x2, …, xn, the sample range is:

r = max(xi) – min(xi) (6-6)

It is the largest observation in the sample minus

the smallest observation.

From Example 6-3:

r = 13.6 – 12.3 = 1.30
Note that: population range ≥ sample range

Sec 6-1 Numerical Summaries of Data 13

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Stem-and-Leaf Diagrams
• Dot diagrams (dotplots) are useful for small
data sets. Stem & leaf diagrams are better
for large sets.
• Steps to construct a stem-and-leaf diagram:
1) Divide each number (xi) into two parts: a stem,
consisting of the leading digits, and a leaf,
consisting of the remaining digit.
2) List the stem values in a vertical column.
3) Record the leaf for each observation beside its
stem.
4) Write the units for the stems and leaves on the
display.

Sec 6-2 Stem-And-Leaf Diagrams 14

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 6-4: Alloy Strength
To illustrate the construction of
a stem‐and‐leaf diagram,
consider the alloy compressive
strength data in Table 6‐2.

Table 6-2 Compressive Strength (psi) of

Aluminum-Lithium Specimens
105 221 183 186 121 181 180 143
97 154 153 174 120 168 167 141
245 228 174 199 181 158 176 110
163 131 154 115 160 208 158 133
207 180 190 193 194 133 156 123
134 178 76 167 184 135 229 146
218 157 101 171 165 172 158 169
199 151 142 163 145 171 148 158
160 175 149 87 160 237 150 135
196 201 200 176 150 170 118 149

Figure 6-4 Stem-and-leaf diagram for Table 6-2

data. Center is about 155 and most data is
between 110 and 200. Leaves are unordered.
Sec 6-2 Stem-And-Leaf Diagrams 15
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Quartiles
• The three quartiles partition the data into four equally sized counts
or segments.
– First or lower quartile : 25% of the data is less than q1.
– Second quartile : 50% of the data is less than q2, the median.
– Third or upper quartile : 75% of the data is less than q3.

• For the Table 6-2 data:

Value of indexed
f Index item quartile

i th (i+1)th
0.25 20.25 143 144 143.25
0.50 40.50 160 163 161.50
0.75 60.75 181 181 181.00

Sec 6-2 Stem-And-Leaf Diagrams 16

The sample median is a

measure of central
tendency that divides
the data into two equal
parts.

The sample mode is the

most frequently
occurring data.

Sec 6-2 Stem-And-Leaf Diagrams 17

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Percentiles and Interquartile Range
• Percentiles are a special case of the quartiles.
• Percentiles partition the data into 100 segments.

• The interquartile range (IQR) is defined as:

IQR = q3 – q1.
• From the Quartiles example:
IQR = 181.00 – 143.25 = 37.75 = 37.8
• Impact of outlier data:
– IQR is not affected
– Range is directly affected.

Sec 6-2 Stem-And-Leaf Diagrams 18

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Frequency Distributions
• A frequency distribution is a compact
summary of data, expressed as a table,
graph, or function.
• The data is gathered into bins or cells,
defined by class intervals.
• The number of classes, multiplied by the
class interval, should exceed the range of the
data. The square root of the sample size is a
guide.
• The boundaries of the class intervals should
be convenient values, as should the class
width.

Sec 6-3 Frequency Distributions And Histograms 19

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Frequency Distribution Table
Frequency Distribution for Table 6-4 Frequency Distribution of Table 6-2 Data
the data in Table 6‐2 Cumulative
Relative Relative
Considerations: Class Frequency Frequency Frequency
Range = 245 – 76 = 169 70 ≤ x < 90 2 0.0250 0.0250
90 ≤ x < 110 3 0.0375 0.0625
Sqrt(80) = 8.9 110 ≤ x < 130 6 0.0750 0.1375
130 ≤ x < 150 14 0.1750 0.3125
Trial class width = 18.9 150 ≤ x < 170 22 0.2750 0.5875
170 ≤ x < 190 17 0.2125 0.8000
Decisions: 190 ≤ x < 210 10 0.1250 0.9250
Number of classes = 9 210 ≤ x < 230 4 0.0500 0.9750
230 ≤ x < 250 2 0.0250 1.0000
Class width = 20 80 1.0000

Range of classes = 20 * 9 = 180

Starting point = 70

Sec 6-3 Frequency Distributions And Histograms 20

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Histograms
• A histogram is a visual display of a frequency
distribution, similar to a bar chart or a stem-and-leaf
diagram.

• Steps to construct a histogram with equal bin widths:

1) Label the bin boundaries on the horizontal scale.

2) Mark & label the vertical scale with the frequencies
or relative frequencies.
3) Above each bin, draw a rectangle whose height is
equal to the frequency corresponding to that bin.

Sec 6-3 Frequency Distributions And Histograms 21

Figure 6-7 Histogram of compressive strength of 80 aluminum-lithium alloy

specimens. Note these features – (1) horizontal scale bin boundaries & labels with
units, (2) vertical scale measurements and labels, (3) histogram title at top or in
legend.

Sec 6-3 Frequency Distributions And Histograms 22

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Histograms with Unequal Bin Widths
• If the data is tightly clustered in some
regions and scattered in others, it is
visually helpful to use narrow class widths
in the clustered region and wide class
widths in the scattered areas.
• In this approach, the rectangle area, not
the height, must be proportional to the
class frequency.
bin frequency
Rectangle height =
bin width

Sec 6-3 Frequency Distributions And Histograms 23

Figure 6-8 Histogram of compressive strength of 80 aluminum-

lithium alloy specimens. Errors: too many bins (17) create
jagged shape, horizontal scale not at class boundaries,
horizontal axis label does not include units.
Sec 6-3 Frequency Distributions And Histograms 24
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Cumulative Frequency Plot

Figure 6-10 Cumulative histogram of compressive strength of 80

aluminum-lithium alloy specimens. Comment: Easy to see
cumulative probabilities, hard to see distribution shape.

Sec 6-3 Frequency Distributions And Histograms 25

Figure 6-11 Histograms of symmetric and skewed distributions.

(b) Symmetric distribution has identical mean, median and mode measures.

(a & c) Skewed distributions are positive or negative, depending on the

direction of the long tail. Their measures occur in alphabetical order as the
distribution is approached from the long tail.

Sec 6-3 Frequency Distributions And Histograms 26

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Histograms for Categorical Data
• Categorical data is of two types:
– Ordinal: categories have a natural order, e.g.,
year in college, military rank.
– Nominal: Categories are simply different, e.g.,
gender, colors.
• Histogram bars are for each category, are of
equal width, and have a height equal to the
category’s frequency or relative frequency.
• A Pareto chart is a histogram in which the
categories are sequenced in decreasing
order. This approach emphasizes the most
and least important categories.

Sec 6-3 Frequency Distributions And Histograms 27

Figure 6-12 Airplane production in 1985. (Source: Boeing

Company) Comment: Illustrates nominal data in spite of the
numerical names, categories are shown at the bin’s midpoint, a
Pareto chart since the categories are in decreasing order.

Sec 6-3 Frequency Distributions And Histograms 28

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Box Plot or Box-and-Whisker Chart
• A box plot is a graphical display showing center,
spread, shape, and outliers (SOCS).
• It displays the 5-number summary: min, q1,
median, q3, and max.

Figure 6-13 Description of a box plot.

Sec 6-4 Box Plots 29
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Box Plot of Table 6-2 Data

Figure 6-14 Box plot of compressive strength of 80 aluminum-lithium

alloy specimens. Comment: Box plot may be shown vertically or
horizontally, data reveals three outliers and no extreme outliers.
Lower outlier limit is: 143.5 – 1.5*(181.0-143.5) = 87.25.
Sec 6-4 Box Plots 30
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Time Sequence Plots
• A time series plot shows the data value, or statistic, on
the vertical axis with time on the horizontal axis.
• A time series plot reveals trends, cycles or other time-
oriented behavior that could not be seen in the data.

Figure 6-16 Company sales by year (a). By quarter (b).

Sec 6-5 Time Sequence Plots 31

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Digidot Plot
Combining a time series plot with some of the other graphical displays that we
have considered previously will be very helpful sometimes. The stem-and-
leaf plot combined with a time series Plot forms a digidot plot.

Figure 6-17 A digidot plot of the compressive strength data in Table 6-2.

Sec 6-5 Time Sequence Plots 32

In order to find a
relationship between two
variables

Sec 6-6 Scatter Diagrams 33

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Constructing a Probability Plot
• To construct a probability plot:
– Sort the data observations in ascending order:
x(1), x(2),…, x(n).
– The observed value x(j) is plotted against the
observed cumulative frequency (j – 0.5)/n.
– The paired numbers are plotted on the
probability paper of the proposed distribution.
• If the paired numbers form a straight line,
then the hypothesized distribution
adequately describes the data.

Sec 6-6 Probability Plots 34

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 6-7: Battery Life
The effective service life (Xj in minutes) of batteries used in a laptop are given in
the table. We hypothesize that battery life is adequately modeled by a normal
distribution. To this hypothesis, first arrange the observations in ascending order
and calculate their cumulative frequencies and plot them.

Table 6-6 Calculations for Constructing

a Normal Probability Plot

j x (j ) (j -0.5)/10 100(j -0.5)/10

1 176 0.05 5
2 183 0.15 15
3 185 0.25 25
4 190 0.35 35
5 191 0.45 45
6 192 0.55 55
7 201 0.65 65
8 205 0.75 75
9 214 0.85 85
10 220 0.95 95

Figure 6-22 Normal probability plot for battery life.

Sec 6-6 Probability Plots 35

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Probability Plot on Standardized Normal Scores
A normal probability plot can be plotted on ordinary axes using z-values. The
normal probability scale is not used.

Table 6-6 Calculations for

Constructing a Normal
Probability Plot
j x (j ) (j -0.5)/10 zj
1 176 0.05 -1.64
2 183 0.15 -1.04
3 185 0.25 -0.67
4 190 0.35 -0.39
5 191 0.45 -0.13
6 192 0.55 0.13
7 201 0.65 0.39
8 205 0.75 0.67
9 214 0.85 1.04
10 220 0.95 1.64

Figure 6-23 Normal Probability plot obtained

from standardized normal scores. This is
equivalent to Figure 6-19.
Sec 6-6 Probability Plots 36
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Probability Plot Variations

Figure 6-24 Normal probability plots indicating a non-normal distribution.

(a) Light tailed distribution
(b) Heavy tailed distribution
(c) Right skewed distribution

Sec 6-6 Probability Plots 37

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Probability Plots with Minitab
• Obtained using Minitab menu: Graphics > Probability Plot. 14 different
distributions can be used.
• The curved bands provide guidance whether the proposed distribution is
acceptable – all observations within the bands is good.

Probability Plot of Battery Life

Normal - 95% CI
99
Mean 195.7
StDev 14.03
95 N 10
AD 0.257
90
P-Value 0.636
80
70
Percent

60
50
40
30
20

1
150 175 200 225 250
Battery Life (x) in Hours

Sec 6-6 Probability Plots 38

Sixth Edition
Douglas C. Montgomery George C. Runger

Chapter 7
Point Estimation of Parameters and Sampling Distributions

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Point Estimation
• A point estimate is a reasonable value of a
population parameter.
• X1, X2,…, Xn are random variables.
• Functions of these random variables, x-bar
and s2, are also random variables called
statistics.
• Statistics have their unique distributions
which are called sampling distributions.

Sec 7-1 Point Estimation 40

As an example,suppose the random variable X is normally distributed with

an unknown mean μ. The sample mean is a point estimator of the unknown
population mean μ. That is, μ  X . After the sample has been selected,
the numerical value x is the point estimate of μ.
Thus if x1  25, x2  30, x3  29, and x4  31, the point estimate of μ is

25  30  29  31
x  28.75
4

Sec 7-1 Point Estimation 41

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Some Parameters & Their Statistics
Parameter Measure Statistic
μ Mean of a single population x‐bar
σ2 Variance of a single population s2
σ Standard deviation of a single population s
p Proportion of a single population p ‐hat
μ1 ‐ μ2 Difference in means of two populations x bar1 ‐ x bar2
p1 ‐ p2 Difference in proportions of two populations p hat1 ‐ p hat2

• There could be choices for the point estimator of a parameter.

• To estimate the mean of a population, we could choose the:
– Sample mean.
– Sample median.
– Average of the largest & smallest observations in the sample.

Sec 7-1 Point Estimation 42

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Some Definitions
• The random variables X1, X2,…,Xn are a
random sample of size n if:
a) The Xi ‘s are independent random variables.
b) Every Xi has the same probability
distribution.
• A statistic is any function of the
observations in a random sample.
• The probability distribution of a statistic is
called a sampling distribution.

Sec 7-2 Sampling Distributions and the Central Limit Theorem 43

Sec 7-2 Sampling Distributions and the Central Limit Theorem 44

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-2: Central Limit Theorem
Suppose that a random variable X has a continuous
uniform distribution: 1 2, 4  x  6
f  x  
0, otherwise
Find the distribution of the sample mean of a random
sample of size n = 40.

By the CLT the distribution X is normal .

ba 64
  5
2 2
 b  a  6  4
2 2

 2
 13
12 12
 2
13 1
 X2    Figure 7‐5 The distribution of
n 40 120
X and X for Example 7‐2.
Sec 7-2 Sampling Distributions and the Central Limit Theorem 45
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Sampling Distribution of a Difference in Sample Means
• If we have two independent populations with means μ1
and μ2, and variances σ12 and σ22, and
• If X-bar1 and X-bar2 are the sample means of two
independent random samples of sizes n1 and n2 from
these populations:
• Then the sampling distribution of:

is approximately standard normal, if the conditions of the

central limit theorem apply.
• If the two populations are normal, then the sampling
distribution of Z is exactly standard normal.
Sec 7-2 Sampling Distributions and the Central Limit Theorem 46
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-3: Aircraft Engine Life
The effective life of a component used
in jet-turbine aircraft engine is a
random variable with mean 5000 and
SD 40 hours and is close to a normal
distribution. The engine manufacturer Figure 7‐6 The sampling distribution of
introduces an improvement into the X2 − X1 in Example 7‐3.
Manufacturing process for this
component that changes the Process
parameters to 5050 and 30. Random Old (1) New (2) Diff (2‐1)
samples of size 16 and 25 are x ‐bar = 5,000 5,050 50
selected. s= 40 30
n= 16 25
What is the probability that the Calculations
difference in the two sample means is s / √n = 10 6 11.7
at least 25 hours? z= ‐2.14
P(xbar2‐xbar1 > 25) = P(Z > z) = 0.9840
= 1 ‐ NORMSDIST(z)
Sec 7-2 Sampling Distributions and the Central Limit Theorem 47
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Unbiased Estimators Defined

Sec 7-3.1 Unbiased Estimators 48

• X is a random variable with mean μ and variance σ2. Let

X1, X2,…,Xn be a random sample of size n.
• Show that the sample mean (X-bar) is an unbiased
estimator of μ.

Sec 7-3.1 Unbiased Estimators 49

Show that the sample variance (S2) is a unbiased estimator of σ2.

Sec 7-3.1 Unbiased Estimators 50

• If we consider all unbiased estimators of θ,

the one with the smallest variance is called
the minimum variance unbiased estimator
(MVUE).
• If X1, X2,…, Xn is a random sample of size
n from a normal distribution with mean μ
and variance σ2, then the sample X-bar is
the MVUE for μ.

Sec 7-3.2 Variance of a Point Estimate 51

Sec 7-3.3 Standard Error Reporting a Point Estimate 52

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-5: Thermal Conductivity
• These observations are 10 xi
measurements of thermal 41.60
conductivity of Armco iron. 41.48
42.34
• Since σ is not known, we use s to 41.95
calculate the standard error. 41.86
42.18
• Since the standard error is 0.2% 41.72
of the mean, the mean estimate is 42.26
41.81
fairly precise. We can be very 42.04
confident that the true population 41.924 = Mean
mean is 41.924 ± 2(0.0898) or 0.284 = Std dev (s )
0.0898 = Std error
between 41.744 and 42.104.
Sec 7-3.3 Standard Error Reporting a Point Estimate 53
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Mean Squared Error

Conclusion: The mean squared error (MSE) of

the estimator is equal to the variance of the
estimator plus the bias squared.

Sec 7-3.4 Mean Squared Error of an Estimator 54

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Relative Efficiency
• The MSE is an important criterion for
comparing two estimators.

• If the relative efficiency is less than 1, we

conclude that the 1st estimator is superior
than the 2nd estimator.

Sec 7-3.4 Mean Squared Error of an Estimator 55

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Optimal Estimator
• A biased estimator can
be preferred than an
unbiased estimator if it
has a smaller MSE.
• Biased estimators are
occasionally used in
Figure 7-8 A biased estimator that
linear regression. has a smaller variance than the
unbiased estimator .
• An estimator whose MSE
is smaller than that of
any other estimator is
called an optimal
estimator.
Sec 7-3.4 Mean Squared Error of an Estimator 56
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Methods of Point Estimations

• Moments
• Maximum likelihood
• Bayesian Estimation

Sec 7.4. Methods of Point Estimations 57

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Moments Defined
• Let X1, X2,…,Xn be a random sample from the
probability distribution f(x), where f(x) can be either
a:
– Discrete probability mass function, or
– Continuous probability density function
• The kth population moment (or distribution
moment) is E(Xk), k = 1, 2, ….
n
T he k th
s a m p le m o m e n t i s 1 / n   X i
k
, k  1, 2 , .. .
i 1

• If k = 1 (called the first moment), then:

– Population moment is μ.
– Sample moment is x-bar.
• The sample mean is the moment estimator of the
population mean.
Sec 7-4.1 Method of Moments 58
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Moment Estimators

Sec 7-4.1 Method of Moments 59

Suppose that X1, X2, …, Xn is a random sample

from a normal distribution with parameter μ and
σ2 where E(X) = μ and E(X2) = μ2 + σ2.
1 n 1 n
  X   Xi and  2   2   X i2
n i 1 n i 1
2
1
n
 n

2 1 n 2
 X i
2
 n n iX
 i 1 
   X i  X 2  i 1
n i 1 n
  n
 
2
n

  X X
2
 n  Xi  
1 i
   X i2   i 1    i 1
(biased)
n  i 1 n  n
 
 
Sec 7-4.1 Method of Moments 60
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-9: Gamma Distribution Moment Estimators-1
Suppose that X1, X2, …, Xn is a random sample from a gamma distribution with
parameter r and λ where E(X) = r/ λ and E(X2) = r(r+1)/ λ2 .

r
 E  X   X is the mean

 E  X 2   E  X  is the variance or
r 2

2
r  r 1
 E  X 2  and now solving for r and  :
2
X2
r  n
1/ n  Xi2  X 2
i 1

X
  n
1/ n  Xi2  X 2
i 1

Sec 7-4.1 Method of Moments 61

2
Using the time to failure data in xi
11.96
xi
143.0416
the table. We can estimate the 5.03 25.3009
67.40 4542.7600
parameters of the gamma 16.07 258.2449
31.50 992.2500
distribution. x‐bar = 21.646 7.73 59.7529
11.10 123.2100
2
ΣX = 6645.4247 22.38 500.8644

X2 21.6462
r    1.29
n
1/ n   X i2  X 2 1 8 6645.4247  21.646 2

i 1

X 21.646
    0.0598
n
1/ n   X i2  X 2 1 8 6645.4247  21.646 2

i 1
Sec 7-4.1 Method of Moments 62
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Maximum Likelihood Estimators
• Suppose that X is a random variable with
probability distribution f(x;θ), where θ is a single
unknown parameter. Let x1, x2, …, xn be the
observed values in a random sample of size n.
Then the likelihood function of the sample is:

L(θ) = f(x1;θ) ∙ f(x2; θ) ∙…∙ f(xn; θ)

• Note that the likelihood function is now a

function of only the unknown parameter θ. The
maximum likelihood estimator (MLE) of θ is the
value of θ that maximizes the likelihood function
L(θ).
Sec 7-4.2 Method of Maximum Likelihood 63
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-10: Bernoulli Distribution MLE
Let X be a Bernoulli random variable. The probability mass function is
f(x;p) = px(1-p)1-x, x = 0, 1 where P is the parameter to be estimated.
The likelihood function of a random sample of size n is:
L  p   p x1 1  p   p x2 1  p   ...  p xn 1  p 
1 x1 1 x2 1 xn

n
n  xi n
 xi
  p xi 1  p  1  p 
1 xi n
 p i1 i 1

i 1

 n   n

ln L  p     xi  ln p   n   xi  ln 1  p 
 i 1   i 1 
n
 n

d ln L  p  
xi 

n  i 1
xi

 i 1

dp p 1  p 

Sec 7-4.2 Method of Maximum Likelihood 64

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-11: Normal Distribution MLE for μ
Let X be a normal random variable with unknown mean μ and
known variance σ2. The likelihood function of a random sample of
size n is: n
1  x   2  2 2

L    e i

i1  2
n
1
 xi 
2
1 2
 e2 i1

 2 2 n2

n 1 n
ln L   ln 2   2  xi  
2 2

2 2 i1
d ln L  1 n
 2  xi  
d  i1

Sec 7-4.2 Method of Maximum Likelihood 65

Let X be a exponential random variable with parameter λ.

The likelihood function of a random sample of size n is:
n
n   xi
L      e   xi
 e n i 1

i 1
n
ln L     n ln       xi
i 1

d ln L    n n
   xi
d  i 1

Equating the above to zero we get

n
  n x
i 1
i  1 X (same as moment estimator)

Sec 7-4.2 Method of Maximum Likelihood 66

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-13: Normal Distribution MLEs for μ & σ2
Let X be a normal random variable with both unknown mean μ
and variance σ2. The likelihood function of a random sample of
size n is: 1  x     2 
L   ,  
n 2 2
2
e 
i 1 2
i

n
1
  xi   
2
1 2
 e 2 i 1

 2  2 n2

n 1
ln L   ,    ln  2 2   2
n

  xi   
2 2

2 2 i 1

 ln L   ,  2  1 n



2
x     0
i 1
i

 ln L   ,  2  n 1 n

  xi   
2
  0
  2
 2 2 2 4 i 1

  xi  X 
2

2
  X and   i 1

Sec 7-4.2 Method of Maximum Likelihood n 67

Notes:
• Mathematical statisticians will often prefer MLEs because of
these properties. Properties (1) and (2) state that MLEs are
MVUEs.
• To use MLEs, the distribution of the population must be known
or assumed.

Sec 7-4.2 Method of Maximum Likelihood 68

This property is illustrated in Example 7‐13.

Sec 7-4.2 Method of Maximum Likelihood 69

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-14: Invariance
For the normal distribution, the MLEs were:

Sec 7-4.2 Method of Maximum Likelihood 70

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Complications of the MLE Method
The method of maximum likelihood is an
excellent technique, however there are two
complications:
1. It may not be easy to maximize the likelihood
function because the derivative function set
to zero may be difficult to solve algebraically.
2. It may not always be possible to use
calculus methods directly to determine the
maximum of L(ѳ).

The following example illustrate this.

Sec 7-4.2 Method of Maximum Likelihood 71
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-16: Gamma Distribution MLE-1
Let X1, X2, …, Xn be a random sample from a gamma
distribution. The log of the likelihood function is:
 n  r xir 1e   x 
ln L  r ,    ln  
i


 i 1  r  
n n
 nr ln      r  1  ln  xi   n ln   r      xi
i 1 i 1

 ln L  r ,   n
 'r 
 n ln      ln  xi   n
r i 1  r 
 ln L  r ,   nr n
   xi
  i 1
Equating the above derivative to zero we get

Sec 7-4.2 Method of Maximum Likelihood 72

Figure 7‐11 Log likelihood for the gamma distribution using the failure
time data. (a) Log likelihood surface. (b) Contour plot.

Sec 7-4.2 Method of Maximum Likelihood 73

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Bayesian Estimation of Parameters-1
• The moment and likelihood methods interpret
probabilities as relative frequencies and are called
objective frequencies.
• The random variable X has a probability
distribution of parameter θ called f(x|θ).
• Additional information about θ is that it can be
summarized as f(θ), the prior distribution, with
mean μ0 and variance σ02. Probabilities associated
with f(θ) are subjective probabilities.
• The joint distribution is f(x1, x2, …, xn|θ).
• The posterior distribution is f(θ|x1, x2, …, xn) is our
degree of belief regarding θ after observing the
sample data.

7-4.3 Bayesian Estimation of Parameters 74

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Bayesian Estimation of Parameters-2
• Now the joint probability distribution of the sample is
f(x1, x2, …, xn, θ) = f(x1, x2, …, xn |θ) ∙ f(θ)
• The marginal distribution is:
  f  x1 , x2 ,..., xn ,θ  , for θ discrete
 θ
f  x1 , x2 ,..., xn    
  f  x1 , x2 ,..., xn ,θ  dθ, for θ continuous
 
• The desired posterior distribution is:

f  x1 , x2 ,..., xn ,θ 
f  θ | x1 , x2 ,..., xn  
f  x1 , x2 ,..., xn 

The Bayesian estimator of θ is θ, the mean of the posterior distribution.

7-4.3 Bayesian Estimation of Parameters 75

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-16: Bayes Estimator for the mean of a Normal Distribution -1

Let X1, X2, …, Xn be a random sample from a normal distribution unknown mean μ and
known variance σ2. Assume that the prior distribution for μ is:

1 1 
  2  2 0  02  2 02
f μ  
    0 
2
2 02
e  e
2 0 2 2
0

The joint probability distribution of the sample is:

  ( x   )
n
2 2
 1 2
1
f  x1 , x2 ,..., xn |   
i
i 1
e
 2  2 n2

 n 2 
 
n

1  1 2 2
 xi  2 
  xi  n 2 

 e  i 1 i 1 

 2  2 n2

7-4.3 Bayesian Estimation of Parameters 76

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-17: Bayes Estimator for the mean of a Normal Distribution-2

Now the joint probability distribution of the sample and μ is:

f  x1 , x2 ,..., xn ,    f  x1 , x2 ,..., xn |    f  μ 
  1 n  
1/2   2  2  2   2   02  2
xi  
 xi2 02 
 
1 
  0     0    2  02 
 e    

 2  2 n2
2 0
    
1
1/2   2  2  2
1   2   0  x   h ( x , x ,..., x , 2 ,  , 2 )
    2  2  1 1 2 n 0 0
  0   0 
e n   n 

Upon completing the square in the exponent,

2
 1 1   2  ( 2 / n )  0 x  02  2
1/ 2  2  2      2    h 2 ( x1 , x 2 ,..., x n ,  ,  0 ,  02 )

 x1 , x 2 , ..., x n ,   
2 2 2
 0  /n     0  / n  0  / n  
f e
w h e re h i ( x 1 , x 2 , ..., x n ,  2 ,  0 ,  02 )is a fu n c tio n o f th e o b s e rv e d
v a lu e s a n d th e p a ra m e te r s  2 ,  0 a n d  02 .
S in c e f  x1 , x 2 , ..., x n ,   does not depend on 
 1 1   2  ( 2 / n )  0  x  2  2
 1/ 2  2  2      0
  h 3 ( x1 , x 2 ,..., x n ,  ,  0 ,  02 )

f  | x 1 , x 2 , ..., x n  e  0  /n     02   2 / n  

7-4.3 Bayesian Estimation of Parameters 77

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-17: Bayes Estimator for the mean of a Normal Distribution-3

which is recognized as a normal probability density

function with posterior mean

and posterior variance

 0  n 
1 2 2
 1 1 
V    2  2   2

 0  n   0   2
n

7-4.3 Bayesian Estimation of Parameters 78

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 7-17: Bayes Estimator for the mean of a Normal Distribution-4

To illustrate:
– The parameters are: μ0 = 0, σ02= 1
– Sample: n = 10, x-bar = 0.75, σ2 = 4


  n  2
0   2
0 x

 02   2 n


 4 10  0  1 0.75 
 0.536
1   4 10 

7-4.3 Bayesian Estimation of Parameters 79

Sixth Edition
Douglas C. Montgomery George C. Runger

Chapter 8
Statistical Intervals for a Single Sample

A confidence interval estimate for  is an interval of the form

l ≤  ≤ u,

where the end-points l and u are computed from the sample

data.

There is a probability of 1  α of selecting a sample for which

the CI will contain the true value of .

The endpoints or bounds l and u are called lower- and upper-

confidence limits ,and 1  α is called the confidence
coefficient.

Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 81

If X is the sample mean of a random sample of

size n from a normal population with known
variance 2, a 100(1  α)% CI on  is given by

x  z/ 2 / n    x  z/ 2/ n (8-1)

where zα/2 is the upper 100α/2 percentage point of

the standard normal distribution.

Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 82

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
EXAMPLE 8-1 Metallic Material Transition
Ten measurements of impact energy (J) on specimens of A238 steel cut at
60°C are as follows: 64.1, 64.7, 64.5, 64.6, 64.5, 64.3, 64.6, 64.8, 64.2,
and 64.3. The impact energy is normally distributed with  = 1J. Find a
95% CI for , the mean impact energy.

The required quantities are zα/2 = z0.025 = 1.96, n = 10,  = l, and x  64.46 .

The resulting 95% CI is found from Equation 8-1 as follows:

 
x  z     x  z / 2
n n
1 1
64 . 46  1 . 96    64 . 46  1 . 96
10 10
63 . 84    65 . 08

Interpretation: Based on the sample data, a range of highly plausible

values for mean impact energy for A238 steel at 60°C is
63.84J ≤  ≤ 65.08J
Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 83
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
8.1.2 Sample Size for Specified Error on the Mean, Variance Known

If x is used as an estimate of , we can be

100(1 − α)% confident that the error | x  will not
exceed a specified amount E when the sample
size is

2
z 
n     (8-2)
 E 

Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 84

Consider the CVN test described in Example 8-1.

Determine how many specimens must be tested to ensure
that the 95% CI on  for A238 steel cut at 60°C has a length
of at most 1.0J.

The bound on error in estimation E is one-half of the length of

the CI.
Use Equation 8-2 to determine n with E = 0.5,  = 1, and
zα/2 = 1.96.
2
 z / 2  
n     
1 . 96 1 
2
 15.37
 0 .5  
 E 
Since, n must be an integer, the required sample size is
n = 16.
Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 85
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
8-1.3 One-Sided Confidence Bounds

A 100(1 − α)% upper-confidence bound for  is

  x  z  / n (8-3)

and a 100(1 − α)% lower-confidence bound for 

x  z  / n  l   (8-4)

Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 86

The same data for impact testing from Example 8-1 are used
to construct a lower, one-sided 95% confidence interval for
the mean impact energy.
Recall that zα = 1.64, n = 10,  = l, and x  64.46 .

A 100(1 − α)% lower-confidence bound for  is


x  z 
n

   
10
  
Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 87
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
8-1.4 A Large-Sample Confidence Interval for 

When n is large, the quantity

X  
S / n
has an approximate standard normal distribution.
Consequently,
s s
x  z/2    x  z / 2 (8-5)
n n
is a large sample confidence interval for , with confidence
level of approximately 100(1  ).

Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 88

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 8-5 Mercury Contamination
A sample of fish was selected from 53 Florida lakes, and mercury
concentration in the muscle tissue was measured (ppm). The mercury
concentration values were
1.230 1.330 0.040 0.044 1.200 0.270
0.490 0.190 0.830 0.810 0.710 0.500
0.490 1.160 0.050 0.150 0.190 0.770
1.080 0.980 0.630 0.560 0.410 0.730
0.590 0.340 0.340 0.840 0.500 0.340
0.280 0.340 0.750 0.870 0.560 0.170
0.180 0.190 0.040 0.490 1.100 0.160
0.100 0.210 0.860 0.520 0.650 0.270
0.940 0.400 0.430 0.250 0.270

Find an approximate 95% CI on .

Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 89

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 8‐5 Mercury Contamination (continued)
The summary statistics for the data are as follows:

Variable N Mean Median StDev Minimum Maximum Q1 Q3

Concentration 53 0.5250 0.4900 0.3486 0.0400 1.3300 0.2300 0.7900

Because n > 40, the assumption of normality is not necessary to use in Equation
8-5. The required values are n  53, x  0.5250, s  0.3486 , and z0.025 = 1.96.

The approximate 95 CI on  is

s s
x  z 0.025  μ  x  z 0.025
n n
0 .3486 0 .3486
0 .5250  1 .96  μ  0 .5250  1 .96
53 53
0 .4311  μ  0 .6189
Interpretation: This interval is fairly wide because there is variability in the
mercury concentration measurements. A larger sample size would have produced
a shorter interval.
Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 90
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Large-Sample Approximate Confidence Interval

Suppose that θ is a parameter of a

probability distribution, and let ̂ be an
estimator of θ. Then a large-sample
approximate CI for θ is given by

ˆ  z  /2  ˆ    ˆ  z  /2  ˆ

Sec 8-1 Confidence Interval on the Mean of a Normal, σ2 Known 91

Let X1, X2, , Xn be a random sample from a

normal distribution with unknown mean  and
unknown variance 2. The random variable

X  
T  (8-6)
S/ n

has a t distribution with n  1 degrees of freedom.

Sec 8-2 Confidence Interval on the Mean of a Normal, σ2 Unknown 92

x
If and s are the mean and standard deviation of a random sample
from a normal distribution with unknown variance 2, a 100(1  )
confidence interval on  is given by

x  t  /2, n  1 s / n    x  t  /2, n  1 s / n (8-7)

where t2,n1 the upper 1002 percentage point of the t distribution with
n  1 degrees of freedom.

One-sided confidence bounds on the mean are found by replacing

t/2,n-1 in Equation 8-7 with t ,n-1.

Sec 8-2 Confidence Interval on the Mean of a Normal, σ2 Unknown 93

Construct a 95% CI on  to the following data.

19.8 10.1 14.9 7.5 15.4 15.4
15.4 18.5 7.9 12.7 11.9 11.4
11.4 14.1 17.6 16.7 15.8
19.5 8.8 13.6 11.9 11.4

The sample mean is x 13.71 and sample standard deviation is s = 3.55.

Since n = 22, we have n  1 =21 degrees of freedom for t, so t0.025,21 = 2.080.

The resulting CI is
x  t  /2, n  1 s / n    x  t  /2, n  1 s / n
13 . 71  2 . 080 ( 3 . 55 ) / 22    13 . 71  2 . 080 3 . 55  / 22
13 . 71  1 . 57    13 . 71  1 . 57
12 . 14    15 . 28
Interpretation: The CI is fairly wide because there is a lot of variability in the
measurements. A larger sample size would have led to a shorter interval.

Sec 8-2 Confidence Interval on the Mean of a Normal, σ2 Unknown 94

Let X1, X2, , Xn be a random sample from a normal

distribution with mean  and variance 2, and let S2
be the sample variance. Then the random variable

2

n  1 S 2
X
2
(8-8)

has a chi-square (2) distribution with n  1 degrees

of freedom.

Sec 8-3 Confidence Interval on σ2 & σ of a Normal Distribution 95

If s2 is the sample variance from a random sample of n

observations from a normal distribution with unknown variance
2, then a 100(1 – )% confidence interval on 2 is

( n  1) s 2  ( n  1) s 2 (8-9)

  
  n  1     n 1

where  
and
 n  1 n 1    
are the upper and lower
100/2 percentage points of the chi-square distribution with
n – 1 degrees of freedom, respectively.

A confidence interval for  has lower and upper limits that

are the square roots of the corresponding limits in
Equation 8–9.
Sec 8-3 Confidence Interval on σ2 & σ of a Normal Distribution 96
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
One‐Sided Confidence Bounds

The 100(1 – )% lower and upper confidence

bounds on 2 are

(n 1)s 2   ( n  1) s 2


  and    (8-10)
n 1 n 1

Sec 8-3 Confidence Interval on σ2 & σ of a Normal Distribution 97

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 8-7 Detergent Filling
An automatic filling machine is used to fill bottles with liquid detergent. A random sample
of 20 bottles results in a sample variance of fill volume of s2 = 0.01532. Assume that the fill
volume is approximately normal. Compute a 95% upper confidence bound.

A 95% upper confidence bound is found from Equation 8‐10 as follows:

 
n  1 s 2
 2    

 2  2 0  1  0 .0 1 5 3
1 0 .1 1 7
 2  0 .0 2 8 7
A confidence interval on the standard deviation  can be obtained by taking the square
root on both sides, resulting in

  0.17
Sec 8-3 Confidence Interval on σ2 & σ of a Normal Distribution 98
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
8-4 A Large-Sample Confidence Interval For a Population Proportion

Normal Approximation for Binomial Proportion

If n is large, the distribution of

X  np Pˆ  p
Z  
np (1  p ) p (1  p )
n

is approximately standard normal.

The quantity p(1  p) /isncalled the standard error of the point estimator . P̂

Sec 8-4 Large-Sample Confidence Interval for a Population Proportion 99

If p̂ is the proportion of observations in a random

sample of size n, an approximate 100(1  )%
confidence interval on the proportion p of the
population is

pˆ (1  pˆ ) pˆ (1  pˆ )
pˆ  z  p  pˆ  z (8-11)
n n

where z/2 is the upper /2 percentage point of the

standard normal distribution.

Sec 8-4 Large-Sample Confidence Interval for a Population Proportion 100

In a random sample of 85 automobile engine crankshaft bearings, 10 have a

surface finish that is rougher than the specifications allow. Construct a 95%
two-sided confidence interval for p.

A point estimate of the proportion of bearings in the population that exceeds the
roughness specification is pˆ  x / n  10 / 85  0 ..12

A 95% two-sided confidence interval for p is computed from Equation 8-11 as

pˆ 1  pˆ  pˆ 1  pˆ 
pˆ  z0.025  p  pˆ  z0.025
n n
0.12  0.88  0.12  0.88 
0.12  1.96  p  0.12  1.96
85 85
0.0509  p  0.2243

Interpretation: This is a wide CI. Although the sample size does not appear to
be small (n = 85), the value of is fairly small, which leads to a large standard
error for contributing to the wide CI.

Sec 8-4 Large-Sample Confidence Interval for a Population Proportion 101

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Choice of Sample Size
Sample size for a specified error on a binomial proportion :

If we set E  z p 1  p  / n and solve for n, the appropriate sample size is

2
 z  
n  p 1  p  (8‐12)
 E 
The sample size from Equation 8-12 will always be a maximum for
p = 0.5 [that is, p(1 − p) ≤ 0.25 with equality for p = 0.5], and can be
used to obtain an upper bound on n.

2
 z  
n  0 .25  (8‐13)
 E 

Sec 8-4 Large-Sample Confidence Interval for a Population Proportion 102

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 8-9 Crankshaft Bearings
Consider the situation in Example 8-8. How large a sample is required if we want
to be 95% confident that the error in using p̂ to estimate p is less than 0.05?

Using pˆ  0 . 12 as an initial estimate of p, we find from Equation 8-12 that the

required sample size is
2 2
z   1 .96 
n   0 .025  pˆ 1  pˆ     0 .12 0 .88   163
 E   0 . 05 
If we wanted to be at least 95% confident that our estimate p̂ of the true proportion
p was within 0.05 regardless of the value of p, we would use Equation 8-13 to find
the sample size
2 2
 z 0 .025   1 .96 
n  0 .25     0 .25   385
 E   0 .05 
Interpretation: If we have information concerning the value of p, either from a
preliminary sample or from past experience, we could use a smaller sample while
maintaining both the desired precision of estimation and the level of confidence.

Sec 8-4 Large-Sample Confidence Interval for a Population Proportion 103

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Approximate One-Sided Confidence Bounds on a Binomial Proportion

The approximate 100(1  )% lower and upper confidence

bounds are

pˆ 1  pˆ  pˆ1  pˆ  (8-14)
pˆ  z p and p  pˆ  z
n n

respectively.

Sec 8-4 Large-Sample Confidence Interval for a Population Proportion 104

Reconsider the crankshaft bearing data introduced in Example 8-8. In that

example we reported that pˆ  0.12 and n  85. The 95% CI was 0.0509  p  0.2243
.
Construct the new Agresti-Coull CI.

z2 2 pˆ 1  pˆ  z2 2 z2 2 pˆ 1  pˆ  z2 2

pˆ   z 2  2 pˆ   z 2  2
2 n n 4n LCL  2 n n 4n
UCL 
1  z 2 n
2
1  z 2 n
2

1.96 2 0.12  0.88  1.96 2 1.96 2 0.12  0.88  1.96 2

0.12   1.96  0.12   1.96 
2(85) 85 4(85) 2 2(85) 85 4(85) 2
 
1  (1.96 2 / 85) 1  (1.96 2 / 85)
 0.2024  0.0654

Interpretation: The two CIs would agree more closely if the sample size were
larger.

Sec 8-4 Large-Sample Confidence Interval for a Population Proportion 105

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
8-5 Guidelines for Constructing Confidence Intervals
Table 8-1 provides a simple road map for appropriate calculation of a
confidence interval.

Sec 8-5 Guidelines for Constructing Confidence Intervals 106

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
8-6 Tolerance and Prediction Intervals
8‐6.1 Prediction Interval for Future Observation

A 100 (1  )% prediction interval (PI) on a single future observation

from a normal distribution is given by

1 1
x  t  n  1 s 1   X n  1  x  t  n  1 s 1  (8-15)
n n

The prediction interval for Xn+1 will always be longer than the confidence
interval for .

Sec 8-6 Tolerance & Prediction Intervals 107

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 8-11 Alloy Adhesion
The load at failure for n = 22 specimens was observed, and found that x  13 . 71
and s  3.55. The 95% confidence interval on  was 12.14    15.28. Plan to test
a 23rd specimen.

A 95% prediction interval on the load at failure for this specimen is

1 1
x  t  n 1 s 1   X n 1  x  t  n 1 s 1 
n n
1 1
13 .71  2 .080 3 .55 1   X 23  13 .71  2 .080 3 .55 1 
22 22
6 .16  X 23  21 .26

Interpretation: The prediction interval is considerably longer than the CI. This is
because the CI is an estimate of a parameter, but the PI is an interval estimate of a
single future observation.

Sec 8-6 Tolerance & Prediction Intervals 108

A tolerance interval for capturing at least % of the

values in a normal distribution with confidence level
100(1 – )% is

x  ks, x  ks

where k is a tolerance interval factor found in

Appendix Table XII. Values are given for  = 90%,
95%, and 99% and for 90%, 95%, and 99%
confidence.

Sec 8-6 Tolerance & Prediction Intervals 109

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
Example 8-12 Alloy Adhesion
The load at failure for n = 22 specimens was observed, and found that x  13 . 71
and s = 3.55. Find a tolerance interval for the load at failure that includes 90% of
the values in the population with 95% confidence.

From Appendix Table XII, the tolerance factor k for n = 22,  = 0.90, and 95%
confidence is k = 2.264.

The desired tolerance interval is

x  ks , x  ks 
[1 3 . 7 1  2 .2 6 4 3 .5 5 , 1 3 .7 1  2 .2 6 4 3 .5 5 ]
(5 .6 7 , 2 1 .7 4 )

Interpretation: We can be 95% confident that at least 90% of the values of load at
failure for this particular alloy lie between 5.67 and 21.74.

Sec 8-6 Tolerance & Prediction Intervals 110

Sixth Edition
Douglas C. Montgomery George C. Runger

Chapter 9
Tests of Hypotheses for a Single Sample

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
9-1.1 Statistical Hypotheses
A statistical hypothesis is a statement about the parameters of one or
more populations.
Let H0 : μ = 50 centimeters per second and H1 : μ 50 centimeters per second
The statement H0 : μ = 50 is called the null hypothesis.
The statement H1 : μ 50 is called the alternative hypothesis.

One-sided Alternative Hypotheses

H0 : μ = 50 centimeters per second H0 : μ = 50 centimeters per second

or
H1 : μ 50 centimeters per second H1 : μ 50 centimeters per second

Sec 9-1 Hypothesis Testing 112

• A procedure leading to a decision about a

particular hypothesis

• Hypothesis-testing procedures rely on using the

information in a. random sample from the
population of interest

• If this information is consistent with the hypothesis,

then we will conclude that the hypothesis is true;
if this information is inconsistent with the hypothesis,
we will conclude that the hypothesis is false.
Sec 9-1 Hypothesis Testing 113
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1.2 Tests of Statistical Hypotheses

H0 : μ = 50 centimeters per second

H1 : μ 50 centimeters per second

Figure 9-1 Decision criteria for testing H0: = 50 centimeters per second
versus H1:  50 centimeters per second.

Sec 9-1 Hypothesis Testing 114

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
9-1.2 Tests of Statistical Hypotheses
Table 9-1 Decisions in Hypothesis Testing

Decision H0 i S True H0 i S False

Fail to reject H0 No error Type II error

Reject H0 Type I error No error

Probability of Type I and Type II Error

 = P(type I error) = P(reject H0 when H0 is true)

β = P(type II error) = P(fail to reject H0 when H0 is false)

Sometimes the type I error probability is called the significance level, or

the -error, or the size of the test.

Sec 9-1 Hypothesis Testing 115

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
Computing the Probability of Type I Error

  P ( X  48 .5 when   50 )  P ( X  51 .5 when   50 )

The z-values that correspond to the critical values 48.5 and 51.5 are

48 .5  50 51 .5  50
z1    1 .90 and z2   1 .90
0 .79 0 .79
Therefore

 P(Z  1.90)  P(Z  1.90)  0.028717  0.028717  0.057434

which implies 5.74% of all random samples would lead to rejection of the
hypothesis H0: μ = 50.

Sec 9-1 Hypothesis Testing 116

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
Computing the Probability of Type II Error

  P ( 48 . 5  X  51 . 5 when   52 )
The z-values corresponding to 48.5 and 51.5 when   52 are

48.5  52 51.5  52
z1   4.43 and z2   0.63
0.79 0.79

Hence,
  P(4.43  Z  0.63)  P(Z  0.63)  P(Z  4.43)
= 0.2643  0.0000
 0.2643

which means that the probability that we will fail to reject the false
null hypothesis is 0.2643.

Sec 9-1 Hypothesis Testing 117

The power of a statistical test

The power of a statistical test is the probability of rejecting the null
hypothesis H0 when the alternative hypothesis is true.

• The power is computed as 1 - β, and power can be interpreted as the probability

of correctly rejecting a false null hypothesis.

• For example, consider the propellant burning rate problem when we are testing
H 0 : μ = 50 centimeters per second against H 1 : μ not equal 50 centimeters per
second . Suppose that the true value of the mean is μ = 52.

When n = 10, we found that β = 0.2643, so the power of this test is

1 – β = 1 - 0.2643
= 0.7357

Sec 9-1 Hypothesis Testing 118

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
9-1.3 One-Sided and Two-Sided Hypotheses
Two-Sided Test:
H0:   0
H1:  ≠ 0
One-Sided Tests:

H0:   0 H0:   0
H1:  > 0 or H1:  < 0

In formulating one‐sided alternative hypotheses, one should remember that

rejecting H0 is always a strong conclusion. Consequently, we should put the
statement about which it is important to make a strong conclusion in the
alternative hypothesis.

Sec 9-1 Hypothesis Testing 119

The P-value is the smallest level of

significance that would lead to rejection of
the null hypothesis H0 with the given data.

P‐value is the observed significance level.

Sec 9-1 Hypothesis Testing 120

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
9-1.4 P-Values in Hypothesis Tests
Consider the two-sided hypothesis test H0:  50 against H1:  50
with n  16 and   2.5. Suppose that the observed sample mean is
centimeters per second. Figure 9-6 shows a critical region for this test with
critical values at 51.3 and the symmetric value 48.7.
The P-value of the test is the  associated with the critical region. Any smaller
value for  expands the critical region and the test fails to reject the null
hypothesis when x  51 . 3 .
The P-value is easy to compute after the test statistic is observed.

x  51 . 3

P - value  1  P ( 48.7  X  51.3)

 48.7  50 51.3  50 
 1  P Z 
 2.5 / 16 2.5/ 16 
 1  P ( 2.08  Z  2.08)
 1  0.962  0.038 Figure 9-6 P-value is area of shaded region when x  51. . 3
Sec 9-1 Hypothesis Testing 121
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
9-1.5 Connection between Hypothesis Tests and
Confidence Intervals
A close relationship exists between the test of a hypothesis for , and
the confidence interval for .

If [l, u] is a 100(1  ) confidence interval for the parameter , the

test of size  of the hypothesis

H0:   0
H1:   0

will lead to rejection of H0 if and only if 0 is not in the 100(1  )

CI [l, u].

Sec 9-1 Hypothesis Testing 122

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-1 Hypothesis Testing
9-1.6 General Procedure for Hypothesis Tests

1. Identify the parameter of interest.

2. Formulate the null hypothesis, H0 .
3. Specify an appropriate alternative hypothesis, H1.
4. Choose a significance level, .
5. Determine an appropriate test statistic.
6. State the rejection criteria for the statistic.
7. Compute necessary sample quantities for calculating the test statistic.
8. Draw appropriate conclusions.

Sec 9-1 Hypothesis Testing 123

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-2 Tests on the Mean of a Normal Distribution,
Variance Known

9-2.1 Hypothesis Tests on the Mean

Consider the two‐sided hypothesis test
H0:   0
H1:  ≠ 0

The test statistic is:

X  0
Z0  (9-1)
/ n
Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 124
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-2 Tests on the Mean of a Normal Distribution,
Variance Known

9-2.1 Hypothesis Tests on the Mean

Reject H0 if the observed value of the test statistic

z0 is either:
z0 > z/2 or z0 < -z/2

Fail to reject H0 if the observed value of the test

statistic z0 is
-z/2 < z0 < z/2

Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 125
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
EXAMPLE 9-2 Propellant Burning Rate

Air crew escape systems are powered by a solid propellant. The

burning rate of this propellant is an important product characteristic.
Specifications require that the mean burning rate must be 50
centimeters per second and the standard deviation is   2
centimeters per second. The significance level of   0.05 and a
random sample of n  25 has a sample average burning rate of
x  51 . 3 centimeters per second. Draw conclusions.

The seven-step procedure is

1. Parameter of interest: The parameter of interest is , the mean

burning rate.
2. Null hypothesis: H0:   50 centimeters per second
3. Alternative hypothesis: H1:   50 centimeters per second

Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 126
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
EXAMPLE 9-2 Propellant Burning Rate
4. Test statistic: The test statistic is
x   0
z 0 
/ n

5. Reject H0 if: Reject H0 if the P-value is less than 0.05. The boundaries of
the critical region would be z0.025  1.96 and z0.025  1.96.

6. Computations: Since x  51 . 3 and   2,

51 . 3  50
z0   3 . 25
2 / 25
7. Conclusion: Since z0  3.25 and the p-value is  2[1  (3.25)]  0.0012,
we reject H0:   50 at the 0.05 level of significance.

Interpretation: The mean burning rate differs from 50 centimeters per

second, based on a sample of 25 measurements.

Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 127
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-2 Tests on the Mean of a Normal Distribution,
Variance Known
9-2.2 Type II Error and Choice of Sample Size
Finding the Probability of Type II Error 
Consider the two-sided hypotheses test H0:   0 and H1:   0

Suppose the null hypothesis is false and the true value of the mean is
  0  , where   0.
X  0 X   0     n
The test statistic Z0 is Z 0   
/ n / n 

Hence, the distribution of Z0 when H1 is true is

  n 
Z 0 ~ N  , 1  (9-2)
  

  n   n
    z /2       z /2 
 


(9-3)
     
Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 128
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-2 Tests on the Mean of a Normal Distribution,
Variance Known
Sample Size for a Two‐Sided Test

For a two-sided alternative hypothesis:

( z / 2  z ) 2  2
n ~ where     0 (9-4)
2

Sample Size for a One‐Sided Test

For a one-sided alternative hypothesis:

( z  z ) 2 2
n where     0 (9-5)
2

Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 129
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
EXAMPLE 9-3 Propellant Burning Rate Type II Error
Consider the rocket propellant problem of Example 9-2. The true burning rate
is 49 centimeters per second. Find  for the two-sided test with   0.05,
  2, and n  25?

Here   1 and z/2  1.96. From Equation 9-3,

 25   25 

   1.96       1 . 96  
   
  
   0.54     4.46   0.295
The probability is about 0.3 that the test will fail to reject the null hypothesis
when the true burning rate is 49 centimeters per second.

Interpretation: A sample size of n  25 results in reasonable, but not great

power  1  β  1  0.3  0.70.

Suppose that the analyst wishes to design the test so that if the true mean
burning rate differs from 50 centimeters per second by as much as 1 centimeter
per second, the test will detect this (i.e., reject H0:   50) with a high probability,
say, 0.90. Now, we note that   2,   51  50  1,   0.05, and   0.10.

Since z/2  z0.025  1.96 and z  z0.10  1.28, the sample size required to detect
this departure from H0:   50 is found by Equation 9-4 as

( z/ 2  z ) 2 2 (1.96  1.28) 2 22

n ~  ~ 42
2 12
The approximation is good here, since
   
  z  /2   n /     1 . 96  1  42 /2    5 . 20  ~ 0 , which is small relative
to .

Interpretation: To achieve a much higher power of 0.90 we need a

considerably large sample size, n  42 instead of n  25.

Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 131
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-2 Tests on the Mean of a Normal Distribution,
Variance Known
9-2.3 Large Sample Test
A test procedure for the null hypothesis H0:   0 assuming that the
population is normally distributed and that 2 known is developed. In
most practical situations, 2 will be unknown. Even, we may not be
certain that the population is normally distributed.

In such cases, if n is large (say, n  40) the sample standard deviation s

can be substituted for  in the test procedures. Thus, while we have
given a test for the mean of a normal distribution with known 2, it can
be easily converted into a large-sample test procedure for unknown
2 regardless of the form of the distribution of the population.

Exact treatment of the case where the population is normal, 2 is

unknown, and n is small involves use of the t distribution.

Sec 9-2 Tests on the Mean of a Normal Distribution, Variance Known 132
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-3 Tests on the Mean of a Normal Distribution,
Variance Unknown
9-3.1 Hypothesis Tests on the Mean
One-Sample t-Test
Consider the two‐sided hypothesis test

H0:   0 and H1:  ≠ 0

Test statistic: X  0
T0 
S/ n

Alternative hypothesis Rejection criteria

H1:   0 t0  t/2,n-1 or t0  t/2,n1

H1:   0 t0  t,n1
H1:   0 t0  t,n1

Sec 9-3 Tests on the Mean of a Normal Distribution, Variance Unknown 133
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
EXAMPLE 9-6 Golf Club Design
An experiment was performed in which 15 drivers produced by a particular club maker
were selected at random and their coefficients of restitution measured. It is of interest to
determine if there is evidence (with   0.05) to support a claim that the mean coefficient
of restitution exceeds 0.82.
The observations are:
0.8411 0.8191 0.8182 0.8125 0.8750
0.8580 0.8532 0.8483 0.8276 0.7983
0.8042 0.8730 0.8282 0.8359 0.8660

The sample mean and sample standard deviation are x  0 . 83725 and s = 0.02456. The
objective of the experimenter is to demonstrate that the mean coefficient of restitution
exceeds 0.82, hence a one-sided alternative hypothesis is appropriate.
The seven‐step procedure for hypothesis testing is as follows:

1. Parameter of interest: The parameter of interest is the mean coefficient of

restitution, .

2. Null hypothesis: H0:   0.82

3. Alternative hypothesis: H1:   0.82

4. Test Statistic: The test statistic is

5. Reject H0 if: Reject H0 if the P‐value is less than 0.05.

6. Computations: Since x  0 . 83725

, s = 0.02456,  = 0.82, and n  15, we have
x   0 0 .83725  0.82
t0  t0   2 .72
s/ n 0.02456 / 15
7. Conclusions: From Appendix A Table II, for a t distribution with 14 degrees
of freedom, t0  2.72 falls between two values: 2.624, for which   0.01, and
2.977, for which   0.005. Since, this is a one-tailed test the P-value is
between those two values, that is, 0.005 < P < 0.01. Therefore, since P < 0.05,
we reject H0 and conclude that the mean coefficient of restitution exceeds
0.82.

Interpretation: There is strong evidence to conclude that the mean coefficient

of restitution exceeds 0.82.

Sec 9-3 Tests on the Mean of a Normal Distribution, Variance Unknown 135
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-3 Tests on the Mean of a Normal Distribution,
Variance Unknown
9-3.2 Type II Error and Choice of Sample Size
The type II error of the two-sided alternative would be

  P (  t  /2 , n  1  T 0  t  /2 , n  1 |   0 )
 P (  t  /2 , n  1  T 0  t  /2 , n  1 )
Curves are provided for two‐sided alternatives on Charts VIIe and VIIf . The abscissa scale
factor d on these charts is defined as

  0 
d  
 
For the one‐sided alternative  0 or 
, use0 charts VIIg and VIIh. The abscissa scale
factor d on these charts is defined as

  0 
d  
 
Sec 9-3 Tests on the Mean of a Normal Distribution, Variance Unknown 136
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-3 Tests on the Mean of a Normal Distribution,
Variance Unknown
EXAMPLE 9-7 Golf Club Design Sample Size
Consider the golf club testing problem from Example 9-6. If the mean
coefficient of restitution exceeds 0.82 by as much as 0.02, is the
sample size n  15 adequate to ensure that H0:   0.82 will be rejected
with probability at least 0.8?

To solve this problem, we will use the sample standard deviation

s  0.02456 to estimate . Then d  /  0.02/0.02456  0.81. By
referring to the operating characteristic curves in Appendix Chart VIIg
(for  = 0.05) with d = 0.81 and n = 15, we find that  = 0.10,
approximately. Thus, the probability of rejecting H0:  = 0.82 if the true
mean exceeds this by 0.02 is approximately 1   = 1  0.10 = 0.90,
and we conclude that a sample size of n = 15 is adequate to provide
the desired sensitivity.

Sec 9-3 Tests on the Mean of a Normal Distribution, Variance Unknown 137
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and
Standard Deviation of a Normal Distribution
9-4.1 Hypothesis Test on the Variance
Suppose that we wish to test the hypothesis that the variance of a normal
population 2 equals a specified value, say , or equivalently, that the
standard deviation  is equal to 0. Let X1, X2,... ,Xn be a random sample
of n observations from this population. To test
2 2
H 0 :   0
2 2 (9-6)
H 1 :   0

we will use the test statistic:

( n  1) S 2
X 02  (9-7)
 02

Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 138
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and Standard
Deviation of a Normal Distribution

9-4.1 Hypothesis Test on the Variance

If the null hypothesis H0 : 2  02 is true, the test statistic X 02 defined in
Equation 9-7 follows the chi-square distribution with n  1 degrees of
freedom. This is the reference distribution for this test procedure.

Therefore, we calculate   , the value of the test statistic X 02 , and the null
hypothesis H0 : 2  02 would be rejected if

    
2
n 1 or if    2  n 1
2 2
where   n  1 and     n  1 are the upper and lower 100  /2
percentage points of the chi-square distribution with n 1 degrees of
freedom, respectively. Figure 9‐17(a) shows the critical region.

Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 139
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and
Standard Deviation of a Normal Distribution
9-4.1 Hypothesis Test on the Variance
The same test statistic is used for one-sided alternative hypotheses. For
the one-sided hypotheses.

H 0 :  2   02 (9-8)
H 1 :  2   02

we would reject H0 if     
2
n  1, whereas for the other one-sided
hypotheses

H 0 :  2   02
(9-9)
H 1 :  2   02

we would reject H0 if     2  n 1 . The one-sided critical regions are

shown in Fig. 9-17(b) and (c).
Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 140
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and
Standard Deviation of a Normal Distribution

9-4.1 Hypothesis Tests on the Variance

Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 141
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and Standard
Deviation of a Normal Distribution
EXAMPLE 9-8 Automated Filling
An automated filling machine is used to fill bottles with liquid detergent.
A random sample of 20 bottles results in a sample variance of fill
volume of s2 = 0.0153 (fluid ounces)2. If the variance of fill volume
exceeds 0.01 (fluid ounces)2, an unacceptable proportion of bottles will
be underfilled or overfilled. Is there evidence in the sample data to
suggest that the manufacturer has a problem with underfilled or
overfilled bottles? Use  = 0.05, and assume that fill volume has a
normal distribution.

Using the seven-step procedure results in the following:

1. Parameter of Interest: The parameter of interest is the population
variance 2.

2. Null hypothesis: H0: 2 = 0.01

3. Alternative hypothesis: H1: 2 > 0.01

Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 142
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and
Standard Deviation of a Normal Distribution
Example 9-8
4. Test statistic: The test statistic is

  
n  1s 2
 02
 2
5. Reject H0: Use  = 0.05, and reject H0 if       30 . 14.

6. Computations:
19 (0.0153 )
    29 .07
0.01
 2
7. Conclusions: Since    29.07     30.14 , we conclude that
there is no strong evidence that the variance of fill volume exceeds 0.01
(fluid ounces)2. So there is no strong evidence of a problem with
incorrectly filled bottles.

Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 143
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and
Standard Deviation of a Normal Distribution

9-4.2 Type II Error and Choice of Sample Size

For the two-sided alternative hypothesis:


 
0

Operating characteristic curves are provided in

Charts VIi and VIj:

Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 144
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-4 Hypothesis Tests on the Variance and
Standard Deviation of a Normal Distribution
EXAMPLE 9‐9 Automated Filling Sample Size
Consider the bottle‐filling problem from Example 9‐8. If the variance of the filling
process exceeds 0.01 (fluid ounces)2, too many bottles will be underfilled. Thus, the
hypothesized value of the standard deviation is 0 = 0.10. Suppose that if the true
standard deviation of the filling process exceeds this value by 25%, we would like to
detect this with probability at least 0.8. Is the sample size of n = 20 adequate?
To solve this problem, note that we require

 0 .125

  1 .25
This is the abscissa parameter forChart
0 .10 From this chart, with n = 20 and 
0VIIk.
 = 1.25, we find that . Therefore, there is only about a 40% chance that the null
 ~ 0if.6the true standard deviation is really as large as  = 0.125
hypothesis will be rejected
fluid ounce.
To reduce the ‐error, a larger sample size must be used. From the operating
characteristic curve with  = 0.20 and  = 1.25, we find that n = 75, approximately.
Thus, if we want the test to perform as required above, the sample size must be at
least 75 bottles.

Sec 9-4 Tests of the Variance & Standard Deviation of a Normal Distribution 145
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-5 Tests on a Population Proportion
9-5.1 Large-Sample Tests on a Proportion

Many engineering decision problems include hypothesis testing about p.

H 0 : p  p0
H 1 : p  p0
An appropriate test statistic is

X  np 0
Z0  (9-10)
np 0 (1  p 0 )

Sec 9-5 Tests on a Population Proportion 146

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-5 Tests on a Population Proportion
EXAMPLE 9-10 Automobile Engine Controller A semiconductor
manufacturer produces controllers used in automobile engine applications.
The customer requires that the process fallout or fraction defective at a critical
manufacturing step not exceed 0.05 and that the manufacturer demonstrate
process capability at this level of quality using  = 0.05. The semiconductor
manufacturer takes a random sample of 200 devices and finds that four of
them are defective. Can the manufacturer demonstrate process capability for
the customer?

We may solve this problem using the seven-step hypothesis-testing procedure

as follows:
1. Parameter of Interest: The parameter of interest is the process fraction
defective p.

2. Null hypothesis: H0:p = 0.05

3. Alternative hypothesis: H1: p < 0.05

This formulation of the problem will allow the manufacturer to make a strong
claim about process capability if the null hypothesis H0: p = 0.05 is rejected.
Sec 9-5 Tests on a Population Proportion 147
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-5 Tests on a Population Proportion
Example 9-10
4. The test statistic is (from Equation 9‐10)
x  np 0
z0 
np 0 (1  p 0 )

where x = 4, n = 200, and p0 = 0.05.

5. Reject H0 if: Reject H0: p = 0.05 if the p‐value is less than 0.05.

6. Computations: The test statistic is

4  200 ( 0 . 05 )
z0    1 . 95
200 ( 0 . 05 )( 0 . 95 )

7. Conclusions: Since z0 = 1.95, the P‐value is  (1.95) = 0.0256, so we reject H0

and conclude that the process fraction defective p is less than 0.05.
Practical Interpretation: We conclude that the process is capable.

Sec 9-5 Tests on a Population Proportion 148

Another form of the test statistic Z0 is

X/n  p0 Pˆ  p 0
Z0  or Z0 
p0 (1  p0 )/n p 0 (1  p 0 ) /n

Sec 9-5 Tests on a Population Proportion 149

9-5.2 Type II Error and Choice of Sample Size

For a two-sided alternative
 p 0  p  z  /2 p 0 (1  p 0 ) / n   p  p  z  /2 p 0 (1  p 0 ) / n  (9-11)
     0 
 p (1  p ) / n   p (1  p ) / n 
   

If the alternative is p < p0

 p 0  p  z  p 0 (1  p 0 ) /n  (9-12)
  1   
 p (1  p ) /n 
 

If the alternative is p > p0

 p 0  p  z  p 0 (1  p 0 ) / n  (9-13)
   
 p (1  p ) / n 
 

Sec 9-5 Tests on a Population Proportion 150

9-5.3 Type II Error and Choice of Sample Size

For a two-sided alternative

2
 z  /2 p 0 (1  p 0 )  z  p (1  p ) 
n  (9-14)
 p  p0 

For a one-sided alternative

2
 z p 0 (1  p 0 )  z  p (1  p ) 
n    (9-15)
 p  p0 

Sec 9-5 Tests on a Population Proportion 151

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-5 Tests on a Population Proportion
Example 9-11 Automobile Engine Controller Type II Error
Consider the semiconductor manufacturer from Example 9-10. Suppose that its
process fallout is really p = 0.03. What is the -error for a test of process capability
that uses n = 200 and a = 0.05?

The -error can be computed using Equation 9-12 as follows:

 0 . 05  0 . 03  (1 . 645 ) 0 . 05 0 . 95  / 200 
 1   
 0 . 03 1  0 . 03  / 200 
 1    0 . 44   0 . 67

Thus, the probability is about 0.7 that the semiconductor manufacturer will fail to
conclude that the process is capable if the true process fraction defective is
p = 0.03 (3%). That is, the power of the test against this particular alternative is only
about 0.3. This appears to be a large -error (or small power), but the difference
between p = 0.05 and p = 0.03 is fairly small, and the sample size n = 200 is not
particularly large.

Sec 9-5 Tests on a Population Proportion 152

Suppose that the semiconductor manufacturer was willing to accept a -error

as large as 0.10 if the true value of the process fraction defective was
p = 0.03. If the manufacturer continues to use a = 0.05, what sample size
would be required?

The required sample size can be computed from Equation 9-15 as follows:
2
 1 . 645 0 . 05 0 . 95   1 . 28 0 . 03 0 . 97  
n  
 0 . 03  0 . 05 

~ 832
where we have used p = 0.03 in Equation 9-15.

Conclusion: Note that n = 832 is a very large sample size. However, we are
trying to detect a fairly small deviation from the null value p0 = 0.05.

Sec 9-5 Tests on a Population Proportion 153

• Testing for Goodness of Fit

• Contingency Table Tests
• Nonparametric Procedures

154
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-7 Testing for Goodness of Fit
• The test is based on the chi-square distribution.
• Assume there is a sample of size n from a population whose probability
distribution is unknown.
• Let Oi be the observed frequency in the ith class interval.
• Let Ei be the expected frequency in the ith class interval.
The test statistic is

k
(O i  E i ) 2
X 2
0   Ei
(9-16)
i 1

Sec 9-7 Testing for Goodness of Fit 155

Poisson Distribution
The number of defects in printed circuit boards is hypothesized to follow a
Poisson distribution. A random sample of n = 60 printed boards has been
collected, and the following number of defects observed.

Number of Observed
Defects Frequency
0 32
1 15
2 9
3 4

Sec 9-7 Testing for Goodness of Fit 156

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-7 Testing for Goodness of Fit
Example 9-12
The mean of the assumed Poisson distribution in this example is unknown and
must be estimated from the sample data. The estimate of the mean number of
defects per board is the sample average, that is, (32·0 + 15·1 + 9·2 + 4·3)/60 =
0.75. From the Poisson distribution with parameter 0.75, we may compute pi, the
theoretical, hypothesized probability associated with the ith class interval. Since
each class interval corresponds to a particular number of defects, we may find the
pi as follows:
e  0 .75 0 . 75 0
p1  P ( X  0 )   0 . 472
0!
e  0 .75 0 . 75 1
p 2  P ( X  1)   0 . 354
1!
e  0 .75 0 . 75 2
p3  P ( X  2)   0 . 133
2!
p 4  P ( X  3 )  1   p1  p 2  p 3   0 . 041

Sec 9-7 Testing for Goodness of Fit 157

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-7 Testing for Goodness of Fit
Example 9-12
The expected frequencies are computed by multiplying the sample size n = 60
times the probabilities pi. That is, Ei = npi. The expected frequencies follow:

Expected
Number of Defects Probability Frequency
0 0.472 28.32
1 0.354 21.24
2 0.133 7.98
3 (or more) 0.041 2.46

Sec 9-7 Testing for Goodness of Fit 158

Number of Observed Expected

Defects Frequency Frequency
0 32 28.32
1 15 21.24
2 (or more) 13 10.44

The chi-square test statistic in Equation 9-16 will have k p  1 = 3 1  1 = 1

degree of freedom, because the mean of the Poisson distribution was
estimated from the data.

Sec 9-7 Testing for Goodness of Fit 159

The seven-step hypothesis-testing procedure may now be applied, using

 = 0.05, as follows:

1. Parameter of interest: The variable of interest is the form of the distribution

of defects in printed circuit boards.

2. Null hypothesis: H0: The form of the distribution of defects is Poisson.

3. Alternative hypothesis: H1: The form of the distribution of defects is not

Poisson.

4. Test statistic: The test statistic is

k
o i  E i 2
    Ei
i 1

Sec 9-7 Testing for Goodness of Fit 160

6. Computations:

 

32  28 . 32 2 
15  21 . 24 2

28 . 32 21 . 24


13  10 . 44 2  2 . 94
10 . 44
2
7. Conclusions: We find from Appendix Table III that    2 . 71 and
2
   3.84 . Because    2 . 94 lies between these values, we conclude
that the P-value is between 0.05 and 0.10. Therefore, since the P-value
exceeds 0.05 we are unable to reject the null hypothesis that the distribution
of defects in printed circuit boards is Poisson. The exact P-value computed
from Minitab is 0.0864.

Sec 9-7 Testing for Goodness of Fit 161

9-8 Contingency Table Tests

Many times, the n elements of a sample from a population

may be classified according to two different criteria. It is
then of interest to know whether the two methods of
classification are statistically independent;

Table 9-2 An r  c Contingency Table

Columns
1 2  c
1 O11 O12  O1c
2 O21 O22  O2c
Rows

r Or1 Or2  Orc

Sec 9-8 Contingency Table Tests 162
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-8 Contingency Table Tests
We are interested in testing the hypothesis that the row-and-column methods of
classification are independent. If we reject this hypothesis, we conclude there is
some interaction between the two criteria of classification. The exact test
procedures are difficult to obtain, but an approximate test statistic is valid for
large n. Let pij be the probability that a randomly selected element falls in the ijth
cell, given that the two classifications are independent. Then pij=uivj, where ui is
the probability that a randomly selected element falls in row class i and vj is the
probability that a randomly selected element falls in column class j. Now.
assuming independence, the estimators of ui and vj are

c
1
uˆ i 
n
 O ij
j 1
r
1 (9-17)
vˆ j 
n
 O ij
i 1

Sec 9-8 Contingency Table Tests 163

Then, for large n, the statistic

r c ( O ij  E ij ) 2 (9-19)
     E ij
i 1 j 1

has an approximate chi-square distribution with (r 1)(c  1) degrees of freedom if

the null hypothesis is true. We should reject the null hypothesis if the value of the

test statistic   is too large. The P-value would be calculated as the probability

 
beyond   on the r  1 c  1  distribution, or P  p  r  1 c  1     . For a


fixed-level test, we would reject the hypothesis of independence if the observed

 
value of the test statistic   exceeded   r  1 c  1 .

Sec 9-8 Contingency Table Tests 164

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9‐8 Contingency Table Tests
A company has to choose among three health insurance plans. Management
wishes to know whether the preference for plans is independent of job
classification and wants to use α= 0.05.

The opinions of a random sample of 500 employees are shown in Table 9-3.
Table 9-3 Observed Data for Example 9-14
Health Insurance Plan
Job
1 2 3 Totals
Classification
Salaried workers 160 140 40 340
Hourly workers 40 60 60 160
Totals 200 200 100 500

To find the expected frequencies, we must first compute uˆ1  (340/500)  0.68,
uˆ2  (160/500)  0.32, vˆ1  (200/500)  0.40, vˆ2  (200/500)  0.40, and
vˆ 3  (100 /500)  0 . 20

Sec 9-8 Contingency Table Tests 165

The expected frequencies may now be computed from Equation 9-18.

For example, the expected number of salaried workers favoring health insurance
plan 1 is
E 11  n uˆ1 vˆ1  500 0 . 68 0 . 40   136

The expected frequencies are shown in below table

Health Insurance Plan

Job
1 2 3 Totals
Classification
Salaried workers 160 140 40 340
Hourly workers 40 60 60 160
Totals 200 200 100 500

Sec 9-8 Contingency Table Tests 166

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-8 Contingency Table Tests
Example 9-14
The seven-step hypothesis-testing procedure may now be applied to this
problem.
1. Parameter of Interest: The variable of interest is employee preference
among health insurance plans.
2. Null hypothesis: H0: Preference is independent of salaried versus hourly
job classification.
3. Alternative hypothesis: H1: Preference is not independent of salaried
versus hourly job classification.
4. Test statistic: The test statistic is
r c o ij  E ij 2
     E ij
i 1 j 1
5. Reject H0 if: We will use a fixed-significance level test with a = 0.05.
Therefore, since r = 2 and c = 3, the degrees of freedom for chi-square are
(r – 1)(c – 1) = (1)(2) = 2, and we would reject H0 if           5 .9 9 .

Sec 9-8 Contingency Table Tests 167


160  136 2 
140  136 2 
40  68 2
136 136 68


40  64 2 
60  64 2 
60  32 2
64 64 32
 49 . 63

7. Conclusions: Since    49 .63   


 5.99 , we reject the hypothesis of
independence and conclude that the preference for health insurance plans

is not independent of job classification. The P-value for    49.63 is
P = 1.671  10–11. (This value was computed from computer software.)

Sec 9-8 Contingency Table Tests 168

• The sign test is used to test hypotheses about the median  of a continuous distribution.

• Suppose that the hypotheses are H 0 :    0 H 1 :    0

• Test procedure: Let X1, X2,... ,Xn be a random sample from the population of interest.
Form the differences Xi  0 , i =1,2,…,n.

• An appropriate test statistic is the number of these differences that are positive, say R+.

• P-value for the observed number of plus signs r+ can be calculated directly from the
binomial distribution.

• If the computed P-value is less than or equal to the significance level α, we will reject H0 .

• For two-sided alternative H 0 :    0 H 1 :    0

• If the computed P-value is less than the significance level α, we will reject H0 .

Sec 9-9 Nonparametric Procedures 169

Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
EXAMPLE 9-15 Propellant Shear Strength Sign Test
Montgomery, Peck, and Vining (2012) reported on a study in which a rocket motor is
formed by binding an igniter propellant and a sustainer propellant together inside a metal
housing. The shear strength of the bond between the two propellant types is an important
characteristic. The results of testing 20 randomly selected motors are shown in Table 9-5.
Test the hypothesis that the median shear strength is 2000 psi, using α = 0.05.
Table 9-5 Propellant Shear Strength Data
Shear
Observation Differences
Strength Sign
i xi - 2000
xi
1 2158.70 158.70 +
2 1678.15 –321.85 –
3 2316.00 316.00 +
4 2061.30 61.30 +
5 2207.50 207.50 +
6 1708.30 –291.70 –
7 1784.70 –215.30 –
8 2575.10 575.10 +
9 2357.90 357.90 +
10 2256.70 256.70 +
11 2165.20 165.20 +
12 2399.55 399.55 +
13 1779.80 –220.20 –
14 2336.75 336.75 +
15 1765.30 –234.70 –
16 2053.50 53.50 +
17 2414.40 414.40 +
18 2200.50 200.50 +
19 2654.20 654.20 +
20 1753.70 –246.30 –
Sec 9-9 Nonparametric Procedures 170
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
EXAMPLE 9-15 Propellant Shear Strength Sign Test - Continued

The seven-step hypothesis-testing procedure is:

1. Parameter of Interest: The variable of interest is the median of the distribution

of propellant shear strength.
2. Null hypothesis: H 0 :   2 0 0 0 p si

3. Alternative hypothesis: H 1 :   2 0 0 0 p si

4. Test statistic: The test statistic is the observed number of plus differences in
Table 9-5, i.e., r+ = 14.
5. Reject H0 : If the P-value corresponding to r+ = 14 is less than or equal to
α= 0.05
6. Computations : r+ = 14 is greater than n/2 = 20/2 = 10.
 1 
P-value : P  2 P  R 
 14 w hen p  
 2 
20
 20 
 0 . 5  0 . 5 
20 r
 2
r
 
r 14  r 
 0 .1 1 5 3
7. Conclusions: Since the P-value is greater than α= 0.05 we cannot reject
the null hypotheses that the median shear strength is 2000 psi.
Sec 9-9 Nonparametric Procedures 171
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.
9-9 Nonparametric Procedures
9-9.2 The Wilcoxon Signed-Rank Test

• A test procedure that uses both direction (sign) and magnitude.

• Suppose that the hypotheses are H 0 :   0 H1 :   0

Test procedure : Let X1, X2,... ,Xn be a random sample from continuous and symmetric
distribution with mean (and Median) μ. Form the differences X i  0 .

• Rank the absolute differences X i  0 in ascending order, and give the ranks
to the signs of their corresponding differences.

• Let W+ be the sum of the positive ranks and W– be the absolute value of the sum of the
negative ranks, and let W = min(W+, W−).
Critical values of W, can be found in Appendix Table IX.

• If the computed value is less than the critical value, we will reject H0 .

• For one-sided alternatives H1 :   0 reject H0 if W– ≤ critical value

H1 :    0 reject H0 if W+ ≤ critical value

Sec 9-9 Nonparametric Procedures 172

Let’s illustrate the Wilcoxon signed rank test by applying it to the propellant shear strength
data from Table 9-5. Assume that the underlying distribution is a continuous symmetric
distribution. Test the hypothesis that the median shear strength is 2000 psi, using
α = 0.05.

The seven-step hypothesis-testing procedure is:

1. Parameter of Interest: The variable of interest is the mean or median of the

distribution of propellant shear strength.
2. Null hypothesis: H 0 :   2 0 0 0 p si

3. Alternative hypothesis: H 1 :   2 0 0 0 p si

4. Test statistic: The test statistic is W = min(W+, W−)

5. Reject H0 if: W ≤ 52 (from Appendix Table IX).

6. Computations : The sum of the positive ranks is

w+ = (1 + 2 + 3 + 4 + 5 + 6 + 11 + 13 + 15 + 16 + 17 + 18 + 19 + 20) = 150, and the
sum of the absolute values of the negative ranks is
w- = (7 + 8 + 9 + 10 + 12 + 14) = 60.

Sec 9-9 Nonparametric Procedures 173

Observation Differences
Signed Rank
i xi - 2000

16 53.50 1
4 61.30 2
1 158.70 3
11 165.20 4
18 200.50 5
5 207.50 6
7 –215.30 –7
13 –220.20 –8
15 –234.70 –9
20 –246.30 –10
10 256.70 11
6 –291.70 –12
3 316.00 13
2 –321.85 –14
14 336.75 15
9 357.90 16
12 399.55 17
17 414.40 18
8 575.10 19
19 654.20 20

W = min(W+, W−) = min(150 , 60) = 60

7. Conclusions: Since W = 60 is not ≤ 52 we fail to reject the null hypotheses that the mean
or median shear strength is 2000 psi.
Sec 9-9 Nonparametric Procedures 174
Copyright © 2014 John Wiley & Sons, Inc. All rights reserved.

Understanding Measures of Dispersion
No ratings yet
Understanding Measures of Dispersion
22 pages
Measures of Dispersion Explained
No ratings yet
Measures of Dispersion Explained
7 pages
Understanding Statistics: Key Concepts
No ratings yet
Understanding Statistics: Key Concepts
10 pages
Basic Statistics for Engineering Students
No ratings yet
Basic Statistics for Engineering Students
14 pages
Measures of Dispersion in Statistics
No ratings yet
Measures of Dispersion in Statistics
10 pages
Six Sigma and Acceptance Sampling Guide
No ratings yet
Six Sigma and Acceptance Sampling Guide
87 pages
Measures of Variability and Position
No ratings yet
Measures of Variability and Position
7 pages
Descriptive Statistics Overview for Engineers
No ratings yet
Descriptive Statistics Overview for Engineers
19 pages
Understanding Measures of Central Tendency
No ratings yet
Understanding Measures of Central Tendency
64 pages
Understanding Measures of Dispersion
No ratings yet
Understanding Measures of Dispersion
12 pages
Understanding Measures of Dispersion
No ratings yet
Understanding Measures of Dispersion
33 pages
Fluid Mechanics Lab Report Overview
No ratings yet
Fluid Mechanics Lab Report Overview
14 pages
Basic Statistics Assignment Solutions
100% (1)
Basic Statistics Assignment Solutions
4 pages
Central Tendency and Dispersion Measures
No ratings yet
Central Tendency and Dispersion Measures
16 pages
Fundamental Statistics for Olympiads
No ratings yet
Fundamental Statistics for Olympiads
68 pages
Understanding Descriptive Statistics
No ratings yet
Understanding Descriptive Statistics
15 pages
Determining Blocks in Randomized Design
No ratings yet
Determining Blocks in Randomized Design
152 pages
Numerical Measures in Data Analysis
No ratings yet
Numerical Measures in Data Analysis
51 pages
Measures of Variability in Statistics
No ratings yet
Measures of Variability in Statistics
4 pages
Numerical Measures of Data Analysis
No ratings yet
Numerical Measures of Data Analysis
23 pages
Measures of Central Tendency Answers
100% (1)
Measures of Central Tendency Answers
11 pages
Body Weight Data Analysis at UNRI
No ratings yet
Body Weight Data Analysis at UNRI
21 pages
Fluid Mechanics Lab Report Overview
No ratings yet
Fluid Mechanics Lab Report Overview
14 pages
Measures of Central Tendency Explained
No ratings yet
Measures of Central Tendency Explained
5 pages
Statistical Concepts in Modern Math
No ratings yet
Statistical Concepts in Modern Math
9 pages
Sampling Distribution Explained
No ratings yet
Sampling Distribution Explained
25 pages
Point Estimation and Sampling Distributions
No ratings yet
Point Estimation and Sampling Distributions
43 pages
Variability Measures in Statistics
No ratings yet
Variability Measures in Statistics
11 pages
Central Tendency and Variation Measures
No ratings yet
Central Tendency and Variation Measures
14 pages
Measures of Variability Explained
No ratings yet
Measures of Variability Explained
7 pages
Standard Deviation in Excel Guide
No ratings yet
Standard Deviation in Excel Guide
16 pages
Central Tendency in Statistics Explained
No ratings yet
Central Tendency in Statistics Explained
23 pages
Understanding Measures of Data Spread
No ratings yet
Understanding Measures of Data Spread
36 pages
Mean, Median, SD, Correlation in Excel
No ratings yet
Mean, Median, SD, Correlation in Excel
13 pages
Understanding Measures of Dispersion
No ratings yet
Understanding Measures of Dispersion
10 pages
Descriptive Statistics Overview
No ratings yet
Descriptive Statistics Overview
42 pages
Describing Data: Measures & Techniques
No ratings yet
Describing Data: Measures & Techniques
64 pages
Measures of Variation in Statistics
No ratings yet
Measures of Variation in Statistics
12 pages
Analytical Chemistry: Standard Deviations
No ratings yet
Analytical Chemistry: Standard Deviations
8 pages
Central Tendency Measures Explained
No ratings yet
Central Tendency Measures Explained
9 pages
Statistical Analysis Exercises Guide
No ratings yet
Statistical Analysis Exercises Guide
19 pages
Understanding Sample Variance in Statistics
No ratings yet
Understanding Sample Variance in Statistics
27 pages
Understanding Tree Height Dispersion
No ratings yet
Understanding Tree Height Dispersion
12 pages
Comparing Two Means and Chi-Square Tests
No ratings yet
Comparing Two Means and Chi-Square Tests
12 pages
Variance and Standard Deviation Explained
No ratings yet
Variance and Standard Deviation Explained
38 pages
Measures of Dispersion in Statistics
No ratings yet
Measures of Dispersion in Statistics
6 pages
Measures of Central Tendency Explained
No ratings yet
Measures of Central Tendency Explained
18 pages
Estimation and Sampling Distributions
No ratings yet
Estimation and Sampling Distributions
50 pages
Measures of Central Tendency & Variability
100% (1)
Measures of Central Tendency & Variability
20 pages
Measures of Dispersion in Statistics
No ratings yet
Measures of Dispersion in Statistics
4 pages
Measures of Location and Spread in Statistics
No ratings yet
Measures of Location and Spread in Statistics
45 pages
Measures of Dispersion Explained
No ratings yet
Measures of Dispersion Explained
27 pages
Sampling Techniques in Statistics
No ratings yet
Sampling Techniques in Statistics
7 pages
Mathematics and Statistics Question Bank
No ratings yet
Mathematics and Statistics Question Bank
8 pages
ANOVA for Pain Management Interventions
No ratings yet
ANOVA for Pain Management Interventions
8 pages
Measures of Variation PDF
No ratings yet
Measures of Variation PDF
15 pages
CS1 Actuarial Statistics Study Guide
No ratings yet
CS1 Actuarial Statistics Study Guide
34 pages
Understanding Probability Theory Basics
No ratings yet
Understanding Probability Theory Basics
15 pages
Understanding Probability Theory Concepts
No ratings yet
Understanding Probability Theory Concepts
49 pages
Predictive Liability Distributions in Insurance
No ratings yet
Predictive Liability Distributions in Insurance
50 pages
Ukuran Risiko Model Kerugian Agregat
No ratings yet
Ukuran Risiko Model Kerugian Agregat
12 pages
Stat Prob q3 Module 1
No ratings yet
Stat Prob q3 Module 1
31 pages
Central Limit Theorem Applications
No ratings yet
Central Limit Theorem Applications
19 pages
Sample Size and Sampling Techniques Guide
No ratings yet
Sample Size and Sampling Techniques Guide
18 pages
Session 6
No ratings yet
Session 6
9 pages
Fitting Probability Distributions to Rainfall Data
No ratings yet
Fitting Probability Distributions to Rainfall Data
43 pages
Probability and Statistics Course Syllabus
No ratings yet
Probability and Statistics Course Syllabus
1 page
40 Simplification and Ratio Problems
No ratings yet
40 Simplification and Ratio Problems
3 pages
Understanding Statistics Basics
No ratings yet
Understanding Statistics Basics
32 pages
Kiyosi Itô's Appointment Year in Bureau
No ratings yet
Kiyosi Itô's Appointment Year in Bureau
9 pages
STA3702 Statistical Inference Exam 2021
No ratings yet
STA3702 Statistical Inference Exam 2021
4 pages
Markov Chain Solutions for MMAT 5340
No ratings yet
Markov Chain Solutions for MMAT 5340
5 pages
Statistical Inference Solutions Tutorial
No ratings yet
Statistical Inference Solutions Tutorial
6 pages
On The Exact Variance of Products Goodman
No ratings yet
On The Exact Variance of Products Goodman
7 pages
Data Science Concepts in Math & Stats
No ratings yet
Data Science Concepts in Math & Stats
2 pages
Validity, Reliability, and Probability Concepts
No ratings yet
Validity, Reliability, and Probability Concepts
14 pages
Statistical Inference For Data Science Compress
No ratings yet
Statistical Inference For Data Science Compress
78 pages
Expectation in Probability Distributions
No ratings yet
Expectation in Probability Distributions
53 pages
Understanding Random Variables and Their Functions
No ratings yet
Understanding Random Variables and Their Functions
13 pages
Conditional Probability Examples and Solutions
No ratings yet
Conditional Probability Examples and Solutions
18 pages
Input Modeling for Simulation Analysis
No ratings yet
Input Modeling for Simulation Analysis
19 pages
Data Science and Visualization Exam Guide
No ratings yet
Data Science and Visualization Exam Guide
2 pages
BSc Actuarial Science Exam Paper
No ratings yet
BSc Actuarial Science Exam Paper
9 pages
Intraday Portfolio Risk: VaR & CVaR Analysis
No ratings yet
Intraday Portfolio Risk: VaR & CVaR Analysis
11 pages
Permutations and Combinations Explained
No ratings yet
Permutations and Combinations Explained
23 pages
Understanding Probability Distributions
No ratings yet
Understanding Probability Distributions
46 pages