Statistics (DA) Lecture 3: Confidence Intervals

Two types of estimates in statistics:
Point, CI
CI (Confidence Interval)
CI for the mean with known population standard dev.
CI for the mean with unknown .. .. ..
CI for a proportion
CI for the variance and std dev
CI for a population Total

NORMSINV
CHIINV

How do we determine the margin of error?
LOC (Level Of Confidence)

Ch 1-3  discuss how to view data
Ch 4    use sample data to estimate population parameters
Ch 5    drawing conclusions from data

Purpose of sampling:
obtain enough information to draw a valid inference

Examples:
engineering materials testing
wine-tasting
sample continuous processes

1) create a sampling plan (financial feasible, practical)
    -Objectives
    -Target Population
    -Population Frame
    -Sampling method
    -Operational procedures
    -statistical tools
2) select sampling methods
    -subjective methods (person chooses samples)
        -selection bias
    -convenience method (easily collect data)
    -probabilistic method
        -simple random sampling
         selecting items from a population so that every subset of a given size has an equal chance of being selected (for example, by generating random numbers) Sample with or without replacement

When collecting data probability distribution unknown
Determine characteristics of an entire population

Calculate sample mean, how close to the true mean?

Distribution of the sample mean -> approaches normal dist
"Sampling Distribution of the Mean"
Probability distribution of obtaining a certain sample mean
Determined by the sample size (describes all possible samples of that size)

Standard error of the mean

You are asked to manage an internal quality control process.
Continuous customer satisfaction monitoring system.
Is average quality of products sufficient?
Sample, estimate characteristics of population.