Estimation
Estimation of population’s mean
One aspect of inferential statistics is estimation, which is the process of estimating the
value of a parameter from information obtained from a sample.
Sampling error is the difference between the sample statistic and the corresponding
population parameter due to the fact that the sample is not a perfect representation of the
population.
Point estimate: A single value from sample data to estimate a population parameter. The
point estimate of is x . Associated with this value is the standard error SE x =
√n
Interval estimate: Called also confidence interval: An interval of values with a defined
probability of containing the population’s mean.
estimate + (tabled value) (standard error of the estimate)
The maximum error of estimate (tabled value standard error of the estimate) is the
maximum likely difference between the point estimate of a parameter and the actual value of
the parameter.
The probability that the interval estimate will contain the parameter is the confidence
coefficient (1-). The confidence level associated with a confidence interval states how much
confidence we have that this interval contains the true population parameter. The confidence
level is denoted by (1-)100%.
Three common confidence intervals are used: the 90%, the 95%, and the 99% confidence
intervals. The most commonly used interval estimate is the 95% interval.
95% confidence interval means that we are 95% confident that the mean of the population
from which we selected the sample is included in the interval.
1. Confidence Intervals for the Mean, Known or n 30: z-disrtribution
If n 30, apply Central limit theorem, shape of the sampling distribution of x has
approximately the standard normal distribution, regardless of the population’s distribution.
CI of = x ± (z/2 x )
z/2 calculation: area to be looked in z table = 1-
2
Example:
The average age of a sample of 42 factory workers is 32 years. Assuming the
population SD is 9 years:
a) Find the point estimate for the population mean
b) Construct a 95% confidence interval for the population mean
Solution:
x=32 σ =9 n=42
a) Point Estimate of a Population Mean:
x=μ=32 yrs
b) Construct a 95% confidence interval for the population mean
1) Assumption:
n > 30 Large Sample Use Z-Table
σx
σ
2) Formula: CI =x ±( Z × )
√n
If A = 0.4750 Z = + 1.96
95% CI CI =32± 1.96 × 9/ √ 42
95 To find x 1:
=47.5 % 95
2 =47.5 % CI =32−1.96 × 9/ √ 42
µ 2 x 1=29.28
0.4750 0.4750 To find x 2 :
CI =32+1.96 ×9 / √ 42
x 2=34.72
−z=−1.96 x=0 + z=1.96 Note: x is the midpoint of x1∧ x2
x 1=29.28 x=32 x 2=34.72 29.28+34.72
x= =32
2
−z=−1.96 z¿ 0 + z=1.96
c) Interpretation
We are 95 % confident that the population mean age of factory workers [29.28 ;
34.72]
2. Confidence Intervals for the Mean, n˂ 30 & unknown: Student’s t distribution
Most of the time, is not known. It can be estimated by S, the standard deviation of the
sample. When the sample size is small, S is not a good estimate of .
We use instead a t-distribution, a member of the Gaussian distribution which has more
variance of the standard Normal distribution (variance is greater than 1), and the areas under
the curve are a function of degrees of freedom df = n-1. T-distribution is similar to standard
normal distribution in that it is unimodal bell shaped, symmetrical, mean=median=mode=0
and located at center of distribution, curve approaches but never touches x-axis, and its area
still equal to 1.
Assumptions:
Assume random sample
assume that the population is normally distributed.
As the sample size increases, the t-dist approaches the standard normal distribution.
CI of = x ± (t/2 S x ) where S x = S
√n
% Confidence level
To read + t you need: Degrees of Freedom: df= n-1
Example:
The mean Blood Glucose Level (BGL) of a sample of 12 men in a certain occupation
is 150 with a standard deviation of 5. Assuming that the population distribution is
normal:
a) Find the point estimate for the population mean
b) Construct a 98% confidence interval for the population mean
Solution:
x=150 S=5 n=12
a) Point Estimate of a Population Mean:
x=μ=150 yrs
b) Construct a 98% confidence interval for the population mean
1) Assumption:
n < 30 & is unknown, assume population distribution is normal small sample,
use T-table
2) Formula: CI =x ± ¿
%CI = 98%
To read + t you need:
Degrees of Freedom: df= n-1 = 12-1==11
Look at the following table to read t-values.
The t Distribution Table
The entries in the table give the critical values of t for the specific number of degrees of freedom
80% 90% 95% 98% 99% 99.8%
df One tail α 0.10 0.05 0.025 0.01 0.005 0.001
Two tail α 0.20 0.10 0.05 0.02 0.01 0.002
1 3.078 6.314 12.706 31.820 63.657 318.309
2 1.886 2.920 4.303 6.965 9.925 22.327
3 1.638 2.353 3.182 4.541 5.841 10.215
4 1.533 2.132 2.776 3.747 4.604 7.173
5 1.476 2.015 2.571 3.365 4.032 5.893
6 1.440 1.943 2.447 3.143 3.707 5.208
7 1.415 1.895 2.365 2.998 3.499 4.785
8 1.397 1.860 2.306 2.897 3.355 4.501
9 1.383 1.833 2.262 2.821 3.250 4.297
10 1.372 1.812 2.228 2.764 3.169 4.144
11 1.363 1.796 2.201 2.718 3.106 4.025
12 1.356 1.782 2.179 2.681 3.055 3.930
CI =150 ±¿
To find x 1:150−2.718× 5/ √ 12 x 1=146.08
To find x 2 :150+ 2.718× 5/ √ 12 x 2=153.92
Interpretation: We are 98 % confident that the population mean Blood Gloucose level of men
in a certain occupation [146.08 ; 153.92]
3. Determining the required sample size for the mean
The sample size required is given by the formula
z 2 2
n = ------------- rounded up to the nearest whole number
2
E
E, also called the maximum error of the estimate, is equal half the length of the interval.
Example: what is the sample size required to estimate the mean monthly fuel consumption
per household vehicle, and to insure that we can be 95% confident that the estimate, x , is
within 0.5$ of (from previous studies, = 20.65)
z 2 2
n = ------------- = (1.96 20.65)2 / 0.52 = 6552.58 n = 6553
E2