Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
33 views

Section 5: Estimation: Part 1 The Distribution of The Sample Mean

This document discusses the distribution of sample means. It explains that the mean of all possible sample means from a population will equal the population mean. The standard deviation of sample means is the population standard deviation divided by the square root of the sample size. If the underlying data is normally distributed or the sample size is large, the central limit theorem states that the distribution of sample means will be normal. It provides examples of calculating probabilities related to sample means and individual values when the population parameters are known.

Uploaded by

Tian Ze
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
33 views

Section 5: Estimation: Part 1 The Distribution of The Sample Mean

This document discusses the distribution of sample means. It explains that the mean of all possible sample means from a population will equal the population mean. The standard deviation of sample means is the population standard deviation divided by the square root of the sample size. If the underlying data is normally distributed or the sample size is large, the central limit theorem states that the distribution of sample means will be normal. It provides examples of calculating probabilities related to sample means and individual values when the population parameters are known.

Uploaded by

Tian Ze
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 26

Section 5

Estimation: Part 1
The Distribution of the Sample Mean

Learning Objectives
Finding the centre, variability, and
shape of a set of sample MEANS.

Distribution
Recall that the DISTRIBUTION of a
data set includes its measures of

Eventual Goal: Estimate


In this section, we want to estimate the
POPULATION mean based on a
SAMPLE.
We use the SAMPLE mean x
to estimate the POPULATION mean .
This is called a POINT ESTIMATE
because the sample mean is only one
number (or point).
Well come back to this in part 2.
4

Many Possible Sample Means


Given a population, there are many
different possible samples of size n that
you can take.
Each one of these samples will have its
own (likely different) sample mean.
Therefore, there are LOTS of sample
means that can be observed from the
DIFFERENT SAMPLES of the SAME
POPULATION.
5

Goal for Part 1


For all the possible sample means of
samples of size n, determine
The centre (mean) of these sample means.
Their spread (standard deviation).
Their shape (normal, right skewed, etc).

Many Possible Sample Means


Example: Suppose the ENTIRE
population consists of only five
observations: 76, 78, 79, 81, 86.
In this case, the population mean is
easy to calculate:
= (76+78+79+81+86)/5 = 80
7

Example (continued)
Data was: 76, 78, 79, 81, 86.
List ALL POSSIBLE samples of size 4
and calculate their sample means.
Sample

Sample Mean

76, 78, 79, 86

79.75

76, 78, 81, 86

80.25

76, 79, 81, 86

80.5

78, 79, 81, 86

81
8

Example (continued)
Sample

Sample Mean

76, 78, 79, 81

78.5

76, 78, 79, 86

79.75

76, 78, 81, 86

80.25

76, 79, 81, 86

80.5

78, 79, 81, 86

81

So, for each sample of size 4, we got a


different sample mean.
Each of these is an estimate (good or
bad) for the population mean ( = 80).
9

Example (continued)
Sample

Sample Mean

76, 78, 79, 81

78.5

76, 78, 79, 86

79.75

76, 78, 81, 86

80.25

76, 79, 81, 86

80.5

78, 79, 81, 86

81

Interesting fact: What is the mean of


these sample means?
(78.5+79.75+80.25+80.5+81)/5 = 80
So the mean of these SAMPLE means
equals the POPULATION mean!
10

Distribution of Sample Means


This happens in general: If you take all
possible sample means (samples of
any size) and take the mean of those,
you always get the population mean.

11

Distribution of Sample Means


Thus, the CENTRE of the sample
means is, not surprisingly, the
population mean.
POPULATION mean
Symbolically:

x
Mean of the SAMPLE means

12

Distribution of Sample Means


What about the standard deviation of all
the means of samples of size n?
Hard to prove, but the standard
deviation of the sample means is the
POPULATION standard deviation,
divided by the square root of the
sample size:
Also called the
x
standard error
n of the
13 means

Distribution of Sample Means


Finally, what is the shape of the means
of samples of size n?
There are a couple of considerations:
The shape of the underlying data.
The size of the samples that gave you
these sample means.

These are summarized in the Central


Limit Theorem.
14

Central Limit Theorem


The sample means of samples of size
n are NORMALLY DISTRIBUTED if
EITHER of the following conditions are
satisfied:
The underlying data is normally distributed
itself (the data that gave you the sample
means).
O R:
The SAMPLE SIZE is large enough
(magic number is 30 or more).
15

Summary
Sample means of samples of size n are
normally distributed (under the
conditions of the central limit theorem),
with
x

x
n

16

Example
IQ scores are normally distributed with
population mean 100 and population
standard deviation 16. Find the
distribution (that is: mean, standard
deviation, and shape) of the sample
means of samples of size 4.

17

Example (continued)
The IQ score data was: = 100, =
16.
Thus, the mean of the sample means
of size 4 (or, any size for that matter) is

The standard deviation of the means of


samples of size 4 is
18

Example (continued)
Since the underlying data is normally
distributed, so are sample means of
samples of size 4 (and, any other size).
Thus, for this example:

19

Z Scores for Sample Means


Usual Z score for raw data is:

x
Z

For sample means, the x becomes


the SAMPLE MEAN and the becomes
the standard deviation OF THE
SAMPLE MEANS.
20

Z Scores for Sample Means


Therefore, the Z score for sample
means becomes

21

Z Score Decision Table


Use this table to decide which Z
score equation to use.
Given x, convert
to Z - score

Given Z score,
convert back to x

Individual
observations (raw
data)

x
Z

x Z

Means of samples
of size n

x
Z
/ n


x Z

n
22

Example
Heights of UPEI students are normally
distributed with mean 162.5 cm and
standard deviation 6 cm. Find the
probability that
1. A student is taller than 171 cm.
(answer: 0.0778).
2. A sample of 90 students has a mean
height of more than 171 cm.
(answer: about 0).
23

Example (heights; = 162.5,


= 6)
3. Find the percentage of samples of 12
students having a mean height between
158.2 and 163 cm. (Answer: 60.75 %)
4. In a group of 270 students, roughly how
many will be shorter than 153 cm?
(Answer: about 15).
5. If a sample of 45 students has mean
height which was higher than 65% of all
other sample means (of same size), what
was the sample mean? (Answer: 162.85)
24

In Class Exercise 5.1.1


The time required for students at UPEI
to finish a statistics final is normally
distributed with mean 7.3 hours and
standard deviation 1.1 hours. Find the
probability that
1. A student takes longer than 8.1 hours.
2. A sample of 17 students has a mean
finishing time between 6.5 and 7.5 hours.
25

In Class Exercise 5.1.1


(continued; = 7.3, = 1.1)
3. For samples of size 6, what is the mean
finishing time that bounds the top 10% of
all mean finishing times (of samples of
the same size)?

26

You might also like