Chapter IV
estimation of parameters
by Group 1
importance:
Estimation parameters in statistics are
numerical values that infer population
characteristics from sample data,
helping in decision-making, prediction,
and reducing uncertainty.
point estimation of
a population
point estimation of a population
A point estimate is a single numerical value used to
approximate an unknown population parameter. It is
derived from sample data and serves as the best
available estimate for characteristics of the
population, such as mean, proportion, or variance.
point estimation types:
Mean Estimation
Proportion Estimation
Variance and Standard
Deviation Estimation
mean estimation
Population Parameter: True population mean (μ)
Point Estimate: Sample mean (x̄ )
Formula:
xirepresents individual data points in the sample
n is the sample size
mean estimation
Example:
Suppose we take a sample of 10 students and their test
scores are: 78, 85, 90, 76, 88, 92, 80, 79, 84, 87
The sample mean is calculated as:
proportion estimation
proportion estimation
Example:
Suppose a survey of 200 voters finds that 120 support a
particular candidate. The estimated proportion of voters who
support the candidate is:
Variance and Standard Deviation Estimation
Variance and Standard Deviation Estimation
Example:
Suppose we have the same test scores as before. We first
find the sample mean then compute the variance
and standard deviation.
3 Methods of Finding Point
Estimates
A. Maximum Likelihood Estimation (MLE)
B. Method of Moments (MME)
C. Least Squares Estimation (LSE)
Maximum Likelihood Estimation (MLE)
This method finds the parameter value that maximizes
the likelihood of observing the sample data.
It is widely used in probability and statistics,
particularly for estimating parameters of probability
distributions.
Method of Moments (MME)
This approach estimates parameters by equating
sample moments (e.g., mean, variance) to population
moments.
Least Squares Estimation (LSE)
Used in regression analysis, LSE estimates parameters
by minimizing the sum of squared residuals.
Properties of a Good
Point Estimator
1. Unbiasedness – The expected value of the
estimator should be equal to the true population
parameter.
2. Efficiency – Among all unbiased estimators, it
should have the smallest variance.
3.Sufficiency – It should use all available
information in the sample.
real world Examples:
A company wants to estimate the average salary of its
employees. They take a random sample of 50 employees and
find that the average salary is $55,000.
Point estimate for population mean salary:
However, we don’t know how accurate this is.
Solution: Use confidence intervals to provide a range around
this estimate.
confidence interval
estimates for the
population mean
confidence interval estimates for the
population mean
A confidence interval estimate for the population
mean is a range of values that likely contains the true
population mean (u) with a certain level of confidence.
It is commonly used in statistics to estimate an
unknown population parameter based on sample data.
Formula for Confidence Interval (CI)
For a population mean (μ), If the population standard
the confidence interval is deviation (σ) is unknown:
calculated as:
a less than 55
55 - a
definitions
= sample mean
z*= critical value from the standard normal
distribution (used when σ is known)
t* = critical value from the t-distribution (used when
σ is unknown and sample size is small)
σ = population standard deviation
s = sample standard deviation
n = sample size
Interpretation
A 95% confidence interval, for example,
means that if we took many random samples
and calculated confidence intervals for
each, about 95% of them would contain the
true population mean.
Example
Suppose a sample of n = 30 students has an average test score of
75 with a standard deviation of 10. To calculate a 95% confidence
interval, assuming the population standard deviation is unknown:
2.
3.
Compute the confidence interval:
The interval is calculated as:
Conclusion: We are 95% confident that the true population
mean falls between 71.26 and 78.74.
confidence interval
estimates for the
population mean (When σ is
Unknown)
Confidence Interval Estimates for the
Population Mean (When σ is Unknown)
When the population standard deviation (𝜎) is
unknown, we use the t-distribution instead of the
normal distribution to calculate the confidence interval
for the population mean. This is because the sample
standard deviation (s) is used as an estimate for 𝜎,
leading to more variability, especially for smaller sample
sizes.
Formula for Confidence
Interval (σ Unknown)
When to Use the
t-Distribution?
Population standard deviation (σ) is unknown.
Small sample size (n < 30).
Data approximately follows a normal distribution.
Steps to Calculate a Confidence
Interval (When σ is Unknown)
Interpretation: We are 95% confident that the true population
mean lies between 46.70 and 53.30.
Example Calculation
Interpreting the Confidence
Interval
Meaning: We are 95% confident that the
population mean is between 46.70 and 53.30.
Increasing confidence level → wider interval.
Larger sample size → narrower interval.
point estimates for
the population
proportion
Confidence Interval Estimates for the
Population Mean (When σ is Unknown)
A point estimate is a single value used to approximate
a population parameter based on sample data.
For example, if you're estimating the population mean
(μ), the sample mean (𝑥̄) is the point estimate.
Confidence Interval Estimates for the
Population Mean (When σ is Unknown)
Here are common population parameters and their point
estimates:
Population mean (μ) → Sample mean (𝑥̄)
Population proportion (p) → Sample proportion (𝑝̂)
Population variance (σ²) → Sample variance (s²)
Population standard deviation (σ) → Sample standard
deviation (s)
1. Point Estimate of Population Mean (μ):
2. Point Estimate of Population Proportion (p):
x is the number of successes in the sample
n is the sample size
3. Point Estimate of Population Variance (σ²):
4. Point Estimate of Population Standard
Deviation (σ):
Interval Estimates
of Population
Proportions
Interval Estimates of Population
Proportions
Internal estimates of population proportion refer to
statistical approximations of the proportion of a
specific characteristic within a population, based on
internal data rather than external sources. These
estimates are often derived from sample data using
statistical techniques.
Methods for Estimating
Population Proportion:
Methods for Estimating
Population Proportion:
Methods for Estimating
Population Proportion:
3. Bayesian Estimation
Uses prior knowledge or assumptions about p
to refine estimates.
The Beta distribution is often used as a
conjugate prior for proportion estimation.
Methods for Estimating
Population Proportion:
4. Maximum Likelihood Estimation (MLE)
The MLE for population proportion in a
Bernoulli process is simply the sample
proportion,
Interpreting Interval
Estimates of
Population Proportions
Interpreting Interval Estimates of
Population Proportions
An interval estimate provides a range of values within
which the true population proportion (p) is likely to fall.
Instead of just giving a single number (point estimate), an
interval gives a margin of error to account for uncertainty
in sampling.
The most common interval estimate is a confidence
interval (CI).
Confidence Interval for
a Population Proportion
How to Interpret a
Confidence Interval
Let’s say we conduct a survey to estimate the
proportion of students in a school who prefer
online learning. We take a random sample and
find that 50% (0.50) of students in the sample
prefer online learning. Suppose our 95%
confidence interval is (0.45, 0.55).
How to Interpret a
Confidence Interval
What This Means:
We are 95% confident that the true proportion of all
students who prefer online learning is between 45%
and 55%.
This means that if we repeated this study multiple
times, 95% of the confidence intervals we calculate
would contain the true population proportion.
How to Interpret a
Confidence Interval
What This Does NOT Mean:
It does not mean that there is a 95% probability that the
true proportion is in the interval. The true proportion is
either in the interval or it isn’t—we just don’t know for
sure.
It does not mean that 95% of students prefer online
learning. Instead, it tells us the range of possible values
for the true proportion.
Effect of Confidence Level and Sample Size
The width of a confidence interval depends on two factors:
a. Confidence Level (e.g., 90%, 95%, 99%)
A higher confidence level (e.g., 99%) gives a wider interval → more
certainty, but less precision.
A lower confidence level (e.g., 90%) gives a narrower interval → more
precision, but less certainty.
b. Sample Size (n)
A larger sample size gives a narrower confidence interval, meaning a
more precise estimate.
A smaller sample size gives a wider confidence interval, meaning a less
precise estimate.
Real-World Applications
Elections: Pollsters estimate the percentage of voters who
support a candidate. If a candidate’s support is between 48%
and 52%, it means they may or may not be ahead.
Medicine: A study might estimate that 60% to 70% of patients
respond to a new drug, helping doctors make treatment
decisions.
Business: A company surveys customers and finds that 30% to
40% would recommend their product. If the lower bound is below
30%, they might need to improve.
Common Mistakes in
Interpretation
Confusing probability with confidence: A 95% confidence
interval does not mean a 95% chance that the true proportion
is in the interval.
Overinterpreting small samples: If the sample is too small,
the confidence interval will be too wide and unreliable.
Ignoring overlap: If two groups' confidence intervals overlap
significantly, it may mean no real difference between them.
