L14 Estimation

Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 50

ESTIMATION

MATH 30-6
Probability and Statistics
OBJECTIVES
The students at the end of the lesson are expected to:
• Define Estimation
• Identify the different parameters and its sample
point estimation
• List down the steps for interval estimation
• Solve estimation problems
• Relate estimation problems to real life situations
Inferential Statistics

Estimation Hypothesis
Testing

Point Interval
estimation Estimation
Point Estimation
Values to be computed sample population

Mean x μ
Proportion p r
Difference x 1 – x2 μ1 – μ2
Difference Of proportion p 1 – p2 r 1 – r2
St. Dev. s σ
Interval Estimation
Significance level

ˆ  ˆ ˆ
P  L    U  1   

ˆL  ˆ  ˆU  → 100(1-α) % confidence interval

Confidence limits Confidence coefficient/


Degree of confidence

The wider the interval, the higher the


confidence level.
Confidence Intervals

 
2 1-α 2

z / 2 μ z / 2

Critical values
Critical Values
Confidence α α/2 Critical value
Level zα/2

90% 0.10 0.050


91% 0.09 0.045
92% 0.08 0.040
93% 0.07 0.035
94% 0.06 0.030
95% 0.05 0.025
96% 0.04 0.020
97% 0.03 0.015
98% 0.02 0.010
99% 0.01 0.005
Estimation About the Mean

• Large Sample
a. Variance is known
b. n>30
• Assume s = s
• Sample size estimation
• Small Sample
a. Variance is unknown and n<30
Large Sample Estimation
Example 1: An electrical firm manufactures light bulbs that have
a length of life that is approximately normally distributed with
a standard deviation of 40 hours. If a sample of 30 bulbs has
an average life of 780 hours, find the 96% confidence interval
for the population mean of all bulbs produced by this firm
9.10 For a random sample of 50 measurements on the breaking
strength of cotton threads, the mean breaking strength was
found to be 210 grams and the standard deviation 18 grams.
Obtain a confidence interval for the true mean breaking
strength of cotton threads of this type, with confidence
coefficient 0.90.
Large Sample Estimation
9.11 A random sample of 40 engineers was selected from among
the large number employed by a corporation engaged in
seeking new sources of petroleum. The hours worked in a
particular week were determined for each engineer selected.
These data had a mean of 46 hours and a standard deviation
of 3 hours. For that particular week, estimate the mean hours
worked for all engineers in the corporation, with a 95%
confidence coefficient.
9.12 An important property of plastic clays is the percent of
shrinkage on drying. For a certain type of plastic clay, 45 test
specimens showed an average shrinkage percentage of 18.4
and a standard deviation of 1.2. Estimate the true average
percent of shrinkage for specimens of this type in a 98%
confidence interval.
Margin of Error

Sample Size
Sample Size Determination
9.13 The breaking strength of threads has a standard deviation
of 18 grams. How many measures on breaking strength
should be used in the next experiment if the estimate of the
mean breaking strength is to be within 4 grams of the true
mean breaking strength, with confidence coefficient 0.90?
9.14 In the setting of Exercise 9.11, how many engineers should
be sampled if it is desired to estimate the mean number of
hours worked to within 0.5 hour with confidence coefficient
0.95?
9.15 Refer to Exercise 9.12.How many specimens should be
tested if it is desired to estimate the percent of shrinkage to
within 0.2 with confidence coefficient 0.98?
Proportion Testing
Proportion
Possible Values of Proportions:
0<p<1
Interpretation:
p = 1 – p preferred is equal to unpreferred
p=0.5 Undecided when p =0.5 is in the interval
pL > 0.5 Preferred
pu < 0.5 Not preferred
Single Proportion
9.16 Upon testing 100 resistors manufactured by Company A, it
is found that 12 fail to meet the tolerance specifications. Find
a 95% confidence interval for the true fraction of resistors
manufactured by Company A that fail to meet the tolerance
specification. What assumptions are necessary for your
answer to be valid?
9.17 Refer to Exercise 9.16.If it is desired to estimate the true
proportion failing to meet tolerance specifications to within
0.05, with confidence coefficient 0.95, how many resistors
should be tested?
9.18 Careful inspection of 70 precast concrete supports to be
used in a construction project revealed 28 with hairline
cracks. Estimate the true proportion of supports of this type
with cracks in a 98% confidence interval.
Small Sample Estimation
Assumptions:
1. σ is unknown
2. Sample size is small (n < 30)
3. The parent population is approximately normal.
Theorem:
If x and s2 are the mean and variance, respectively of a
random sample of size n taken from a population that is
normally distributed with mean μ and variance σ2 then
x
t
s/ n
is the value of a random variable T having the t-distribution
with v = n – 1 degrees of freedom.
Small Sample Distribution
Small Sample Estimation
9.22 The warpwise breaking strength measured on five
specimens of a certain cloth gave a sample mean of
180 psi and a standard deviation of 5 psi. Estimate
the true mean warpwise breaking strength for cloth
of this type in a 95% confidence interval. What
assumption is necessary for your answer to be valid?

9.23 Answer Exercise 9.22 if the same sample data


resulted from a sample of
a 10 specimens
b 100 specimens
Small Sample Distribution
A machine is producing metal pieces that are cylindrical
in shape. A sample of pieces is taken and the
diameters are
1.01, 0.97, 1.03,1.04,0.99, 0.98, 0.99, 1.01 and 1.03
centimeters. Find a 99% confidence interval for the
mean diameter of pieces from this machine assuming
an approximate normal distribution.

9.21 The Environmental Protection Agency has collected data on


the LC50 (concentration killing 50% of the test animals in a
specified time interval) measurements for certain chemicals likely
to be found in freshwater rivers and lakes. For a certain species of
fish, the LC50 measurements (in parts per million) for DDT in 12
experiments yielded the following:
16, 5, 21, 19, 10, 5, 8, 2, 7, 2, 4, 9
Assuming such LC50 measurements to be approximately normally
distributed, estimate the true mean LC50 for DDT with confidence
Confidence Interval About Mean
Differences
• Large Sample
a. Variance is known
b. n>30
• Assume s = s
• Small Sample (n<30)
a. Variance is unknown and unequal
b. Variance is unknown and equal
• Paired Observation
• Proportions
Interpreting Differences
D = 1-2

Lower Boundary Upper Boundary Interpretation


Negative Negative 1 < 2
Negative Positive 1 =2
Positive Positive 1 >2
Positive Negative
Estimate for the Difference of Means

If x1 and x2 are the means of two independent random


samples of size n1 and n2 from a population with
known variances σ12 and σ22, respectively, a
100(1-α)% confidence interval for μ1 - μ2 is given by

s1 s 2 s1 s 2
2 2 2 2

( x1  x2 )  z / 2   1   2  ( x1  x2 )  z / 2 
n1 n2 n1 n2
Example
A random sample of size n1 = 25 taken from a normal
population with a standard deviation of σ1 = 5 has a
mean x1 = 80. A second random sample of size n2=36
taken from a different normal population with a
standard deviation σ2= 3, has a mean x2 = 75. Find
the 95% CI for the difference of the population means.
Large Sample Confidence Interval
for Mean Differences
Example 9.12
Large Sample Confidence Interval
for Mean Differences
9.36 From prev prob. Suppose a similar experiment is to be run
again with an equal number of specimens from each type of
coupling agent. How many specimens should be used if we
want to estimate the true difference bet. mean resistances to
within 1 unit, with a confidence coefficient of 0.95?
Small Sample Estimate for the
Difference of Means

where
(n1  1) s1  (n2  1) s 2
2 2

sp 
n1  n2  2
Small Sample Estimate for the
Difference of Means
Example 9.14 Copper produced by sintering (heating without
melting) a powder under certain conditions is then measured
for porosity (the volume fraction due to voids) in a certain
laboratory. Two different batches of copper were produced
and the batches are checked for consistency by measuring
their porosity. The porosity measurements on four random
samples from batch 1 show a mean of 0.22 and a variance of
0.0010. The porosity measurements on five random samples
from batch 2 show a mean of 0.17 and a variance 0.0020.
Estimate the 95% confidence interval for difference between
the (population) mean porosity measurements for these two
batches. Assume that the sample is taken from a normal
population with equal variances.
Example
A course in statistics is taught to 12 students by the
conventional classroom procedures. A second group of 10
students was given to the same course by means of a
programmed material. At the end of the term, the same
exam was given to each group. The 12 students gave an
average grade of 85 with a standard deviation of 4, while
the 10 students gave an average of 81 with a standard
deviation of 5. Find the 90% CI for the difference between
the population means, assuming the populations are
normally distributed with equal variances.
Estimate for the Difference of Means
If x1 and x2 are the means of two independent random
samples of size n1 and n2 from a population with
unknown and unequal variances, a 100(1-α)% CI for
μ1 - μ2 is given by
2 2 2 2
S1 S 2 S1 S 2
( x1  x2 )  t / 2   1   2  ( x1  x2 )  t / 2 
n1 n2 n1 n2
2
where  s1 s2 
2 2
  
n n 
v  2
1 2 
2 2
 s1   s2 
2
   
n  n 
 1   2 
n1  1 n2  1
Example
A taxi company is trying to decide whether to purchase brand A
Or brand B tires for its fleet of taxis. To estimate the difference
Of two brands, an experiment is conducted using 12 of each
Brand. The tires are run until they are worn out. The results
Are
Brand A Brand B
Mean 36 300 km 38100 km
Standard Dev 5 000km 6 100 km

Compute for the 95% confidence interval for u1 - u2,assuming


The populations to be normally distributed and variances
unequal
Estimate for Paired Observations
If d and sd are the mean and standard deviation of the
normally distributed differences of n random pairs of
Measurements, a (1-a)100% confidence interval for uD

Sd Sd
d  t / 2   D  d  t / 2
n n
Where
v=n-1
Example
The government awarded grants to the agricultural
departments of 9 universities to test the yield capabilities of
two new varieties of wheat. Each variety was planted on plots
of equal area at each university and the yields, in kg are
recorded as follows:
University
Variety 1 2 3 4 5 6 7 8 9
1 38 23 35 41 44 29 37 31 38
2 45 25 31 38 50 33 36 40 43

Find a 95% confidence interval for


a. the mean difference between the yields of the two varieties
assuming equal variances,
b. Paired difference between two varieties
assuming the yields of the two varieties to be approximately
normally distributed
Estimate for Difference of Proportion
If p is the proportion of success in a random sample of
size n, and q = 1- p, an approximate (1-a) 100%
confidence interval for the binomial parameter p is
given by

pˆ 2 qˆ2 pˆ1qˆ1 pˆ 2 qˆ2 pˆ1qˆ1


( pˆ 2  pˆ1 )  z / 2   p2  p1  ( pˆ 2  pˆ1 )  z / 2 
n2 n1 n2 n1
Example

Ten engineering schools in the US were surveyed.


The sample contained 250 electrical engineers, 80
being women; 175 chemical engineers, 40 being
women. Compute a 90% confidence interval for the
difference between the proportion of women in these
two fields of engineering. Is there a significant
difference between the two proportions?
Estimate for Difference of Proportion
If p is the proportion of success in a random sample of
size n, and q = 1- p, an approximate (1-a) 100%
confidence interval for the binomial parameter p is
given by

pˆ 2 qˆ2 pˆ1qˆ1 pˆ 2 qˆ2 pˆ1qˆ1


( pˆ 2  pˆ1 )  z / 2   p2  p1  ( pˆ 2  pˆ1 )  z / 2 
n2 n1 n2 n1
Example
Ten engineering schools in the US were surveyed.
The sample contained 250 electrical engineers, 80
being women; 175 chemical engineers, 40 being
women. Compute a 90% confidence interval for the
difference between the proportion of women in these
two fields of engineering. Is there a significant
difference between the two proportions?
Estimation about variances

• Single sample
– Chi Squared
• Ratio Of Variance
– F distribution
If s2 is the variance of a random sample size n, from a
Normal population, a (1- )100% confidence interval
for s2 is

(n  1)s 2 2 (n  1)s 2
s  2
  /2
2
 1 / 2

V = n-1
Estimate for Single Sample Variance
Example

A manufacturer of car batteries claims that his


batteries will last, on average, 3 years with variance
of 1 year. If 5 of these batteries have lifetimes of 1.9,
2.4,3.0,3.5, and 4.2 years, construct a 95% confidence
interval for s and decide if the manufacturer’s claim
that s = 1 is valid. Assume the population of battery
lives to be approximately normally distributed.
Estimate for Two Sample:Ratio of Variance
If s12 and s22 are the variances of independent samples
of size n, from a Normal population, a (1- )100%
confidence interval
for s1 2 / s2 2 is

s 1 s1
2 2 2
s1
 2  2 f / 2 (v2 , v1 )
s2 f / 2 (v1 , v2 ) s 2 s2
2
Example
A taxi company is trying to decide whether to purchase brand A or brand B tires for
its fleet of taxis. To estimate the difference of two brands, an experiment is
conducted using 10 of each brand. The tires are run until they are worn out.
The results are
Brand A Brand B
Mean 36 300 km 38100 km
Standard Dev 5 000km 6 100 km

Compute for the 95% confidence interval for u1 - u2,assuming


The populations to be normally distributed and variances unequal

Construct a 90% confidence interval for s1/s2 for the said


problem and can we assume that s1=s2 when we construct our
confidence interval for µ1-µ2?
REFERENCES
• http://en.wikipedia.org/wiki/Statistics
• http://writing.colostate.edu/guides/research/
stats/index.cfm

You might also like