1. Standard Deviation
Lecture 20
Sec
Fri, Feb 23, 2007

2. Variability
Our ability to estimate a parameter accurately depends on the variability of the population.
What do we mean by variability in the population?
How do we measure it?

3. Deviations from the Mean
Each unit of a sample or population deviates from the mean by a certain amount.
Define the deviation of x to be (x –x). 1 2 3 4 5 6 7 8
x = 4

4. Deviations from the Mean
Each unit of a sample or population deviates from the mean by a certain amount.
deviation = –4 1 2 3 4 5 6 7 8
x = 4

5. Deviations from the Mean
Each unit of a sample or population deviates from the mean by a certain amount.
dev = –1 1 2 3 4 5 6 7 8
x = 4

6. Deviations from the Mean
Each unit of a sample or population deviates from the mean by a certain amount.
dev = +2 1 2 3 4 5 6 7 8
x = 4

7. Deviations from the Mean
Each unit of a sample or population deviates from the mean by a certain amount.
deviation = +3 1 2 3 4 5 6 7 8
x = 4

8. Deviations from the Mean
How do we obtain one number that is representative of the whole set of individual deviations?
Normally we use an average to summarize a set of numbers.
Why will the average not work in this case?

9. Sum of Squared Deviations
We will square them all first. That way, there will be no canceling.
So we compute the sum of the squared deviations, called SSX.
Procedure
Find the average
Find the deviations from the average
Square them all
Add them up

10. Sum of Squared Deviations
SSX = sum of squared deviations
For example, if the sample is {0, 3, 6, 7}, then
SSX = (0 – 4)2 + (3 – 4)2 + (6 – 4)2 + (7 – 4)2
= (–4)2 + (–1)2 + (2)2 + (3)2 =
= 30.

11. The Population Variance
Variance of the population
The population variance is denoted by 2.

12. The Population Standard Deviation
The population standard deviation is the square root of the population variance.

13. The Sample Variance Variance of a sample
The sample variance is denoted by s2.

14. The Sample Variance
Theory shows that if we divide by n – 1 instead of n, we get a better estimator of 2.
Therefore, we do it.

15. Example
In the example, SSX = 30.
Therefore,
s2 = 30/3 = 10.

16. The Sample Standard Deviation
The sample standard deviation is the square root of the sample variance.
We will interpret this as being representative of deviations in the sample.

17. Example In our example, we found that s2 = 10.
Therefore, s = 10 =

18. Example
Use Excel to compute the mean and standard deviation of the sample
{0, 3, 6, 7}.
Do it once using basic operations.
Do it again using special functions.

19. Alternate Formula for the Standard Deviation
An alternate way to compute SSX is to compute
Then, as before

20. Example Let the sample be {0, 3, 6, 7}. Then  x = 16 andSo
SSX = 94 – (16)2/4
= 94 – 64
= 30,
as before.

21. TI-83 – Standard Deviations
Follow the procedure for computing the mean.
The display shows Sx and x.
Sx is the sample standard deviation.
x is the population standard deviation.
Using the data of the previous example, we have
Sx =
x =

22. Interpreting the Standard Deviation
Both the standard deviation and the variance are measures of variation in a sample or population.
The standard deviation is expressed in the same units as the data.
Therefore, the standard deviation is directly comparable to actual deviations.

23. Interpreting the Standard Deviation
The variance is not comparable to deviations.

24. Interpreting the Standard Deviation
Observations that deviate fromx by much more than s are unusually far from the mean.
Observations that deviate fromx by much less than s are unusually close to the mean.

25. Interpreting the Standard Deviationx

26. Interpreting the Standard Deviationx

27. Interpreting the Standard Deviation
x – s x
x + s

28. Interpreting the Standard Deviation
A little closer than normal tox
but not unusual
x – s x
x + s

29. Interpreting the Standard Deviation
Unusually close tox
x – s x
x + s

30. Interpreting the Standard Deviation
A little farther than normal fromx
but not unusual
x – 2s
x – s x
x + s
x + 2s

31. Interpreting the Standard Deviation
Unusually far fromx
x – 2s
x – s x
x + s
x + 2s