Estimation

Business Statistic
Theory of Estimation
Theory of Estimation
 The objective of sampling is to use the information
obtained from the sample to learn about the population
from which the sample is taken.
 Estimation : It refers to the procedure where sample
information is used to estimate the numerical value of
some population measure. For example population mean
and Population variance.
 It involves using sample statistic to predict the value of
population parameters.
Types of Estimates
1. Point Estimates
2. Interval Estimates
Point Estimates
 It is a specific value used to estimate population

parameter.
 When a point estimate is found, the sample or statistic
used is called an estimator and the specific number
obtained is called an estimate.
 It can be said that estimate is an educated guess about
some value of population parameter, where an estimator is
the rule used to obtain the guess.
 For example Sample mean is the estimator and its specific
value obtained from a sample is the estimate.
Properties of Good Estimator :
1. Unbiasedness :
 An estimator is said to be unbiased if its expected value
is equal to the population parameter it estimates.
 In other words, If the expected value or mean of the
sampling distribution of a statistic is equal to
corresponding parameter, the statistic is called the
unbiased estimator.
 Otherwise the estimator is said to be biased estimator.
1. Unbiasedness :
 Thus sample mean(X Bar) is an unbiased estimator of
population mean (μ)
 Sample proportion (p) is also an unbiased estimator of
the population proportion(π)
 Note : Sample variance is not an unbiased estimator of
population variance. Sample Variance tends to
underestimate the population variance.
2. Efficiency : An estimator is said to be efficient if it has

relatively small variance.
 If the sampling distribution of two statistic have the same
expected value then of the two, the statistic with the smaller
variance is known as efficient estimator.
 The estimator with the larger variance is called inefficient
estimator.
 The sampling distribution of mean and median both have the
same expected value, equal to the population mean. But the
variance of the sampling distribution of the mean is smaller
then the variance of the sampling distribution of median. Thus
the sampling mean provides a better estimate for population
mean.
2. consistency : An estimator is said to be consistent if its

probability of being close to the parameter it estimate
increases as the sample size increases.
• When a statistic is a consistent estimator, then the tendency is for
larger samples to yield estimates that are closer to the parameter
value than the ones found using smaller samples.
2. Sufficiency :A sufficient estimator is one which utilises all

information a sample contains about the parameter to be
estimated.
 For example, the sample mean (X bar), is a sufficient
estimator of the population mean, (μ).
 This means that no other estimator of (μ) for the same sample
data, such as the sample median or sample mode, can add any
further information to the population parameter being
estimated.
 Like sample mean, Sample proportion(p) is a sufficient
estimator of the population proportion,(π).
A good estimator is one that is unbiased, efficient,
Consistent and sufficient. As an estimator of the population
mean, the sample mean possesses all these properties.
Interval estimates
 An interval estimate gives the estimate in terms of a range of values.
 Internal Estimate :
point estimate (Plus/Minus) (Interval Coefficient * standard Error)
 Internal estimate is obtained by adding and subtracting some quantity to the
point estimate.
 The interval coefficient or confidence coefficient depends on the level of
confidence and the shape of the sampling distribution.
 A level of confidence equals to 95% means that the probability is 0.95 that the
parameter value being estimated is contained, and 0.05 that is not contained,
within the internal we obtained.
 Similarly a 99% level of confidence implies that the probability is 0.99 of the
parameter to be included in the interval.
Confidence Interval for Population
Mean: Large Samples
 When the Population Standard  When the Population Standard
Deviation is known Deviation is not known
CI = x̅ ± zα/2 CI = x̅ ± zα/2
where
1-α = Confidence Level
CI in case of small samples
 When X is normally distributed  When X is not normally

distributed
Case 1 : When σ is known
In this case CI can not be

CI = x̅ ± zα/2
determined
Case 2 : when σ is not known
CI = x̅ ± tα/2
t-distribution
 t-distributions are a family of distributions.

 There is a t-distribution corresponding to a given degrees of freedom(v).
Degrees of freedom (v) is one less than the sample size(n).
 v = n-1
 A t-distribution is symmetrical, bell shaped and extends from minus infinity to
infinity.
 t-distribution is flatter than the normal curve and has fatter tails.
 The degree of flatness decreases with increases in sample size n and it
gradually tends to approach the z-distribution.

Estimation

Uploaded by

Copyright:

Available Formats

Estimation

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Estimation

Uploaded by

Copyright:

Available Formats

Business Statistic

 It is a specific value used to estimate population

2. Efficiency : An estimator is said to be efficient if it has

2. consistency : An estimator is said to be consistent if its

2. Sufficiency :A sufficient estimator is one which utilises all

 When X is normally distributed  When X is not normally

In this case CI can not be

Case 2 : when σ is not known

 t-distributions are a family of distributions.

You might also like