Professional Documents
Culture Documents
TOPIC2 - Statistical Interval
TOPIC2 - Statistical Interval
TOPIC2 - Statistical Interval
Section 8.1:
CI for µ when σ 2 known & drawing from normal distribution
Section 8.1.2:
Sample size calculation for estimating µ with specified error, σ 2 known
Section 8.2:
CI for µ when σ 2 unknown & drawing from normal distribution
1/
Confidence Intervals
Point estimates are a good start, but we should also give the client
some idea of the confidence in our estimate.
2/
Confidence Interval for µ - Normal parent population, known
σ2
We use the observed x¯as the point estimate for µ.
(
4 5 6 7 8 9 10 11 12
)
x
3/
Confidence Interval for µ - Normal parent population, known
σ2
(
4 5 6 7 8 9 10 11 12
)
x
“We are 95% confident that the true mean µ falls in this
interval.” 4/
Confidence Interval for µ - Normal parent population, known
σ2
.(0,1)
95% OF DISTRIBUTION
NOTATION: z0.025 is the z-value such that 97.5% of the distribution is below and 2.5% is
above it (an upper tail z-value).
6 / 31
Confidence Interval for µ - Normal parent population, known
σ2
We can state the lower and upper end-points of the 95% CI for µ
from a random sample of size n drawn from a normally distributed
population with variance σ2 and sample mean of x¯as:
Lower end-point (L) = x¯ − 0.02 √σ
n
z 5
Upper end-point (U) = x¯ + 0.02 √σ
n
zNOTE: x¯ lies in the center of the 2-sided5confidence interval.
95% CI for µ when σ 2 known and drawing from a normally
distributed population:
σ −z
x¯ √ ≤ µ ≤ x¯+ z √ σ
0.025 0.025
n n
Or...
σ σ
x¯− 1.96 √ ≤ µ ≤ x¯+ 1.96 √
n n
7 / 31
Confidence Interval for µ - Normal parent population, known
σ2
Example (Fill weights of boxes)
The sample mean for the fill weights of 100 boxes is x¯ = 12.050. The
population variance of the fill weights is known to be (0.100)2. Find a
95% confidence interval for the population mean µ fill weight of the
boxes.
ANS: L = x¯ −0.025
z ·√σ0.100
= 12.050 − 1.96 √ = 12.030.
· n
100
U = x¯ + 0.025
z √σ = 12.050 + 1.96 √ = 12.070.
0.100
· n
·
100
The 95% confidence interval for µ is [12.030, 12.070].
We are 95% confident that the true parameter value lies in this interval.
NOTE: Because σ2 was very small and n was fairly large, we have a very narrow confidence
interval for µ (which is good).
8 / 31
Confidence Interval for µ - Normal parent population, known
σ2
CI for any choice of confidence level, or 100(1-α)% confidence
The confidence level of choice is stated as 100(1 − α)%.
For a 95% confidence interval, α = 0.05.
For an 80% confidence interval, α =
0.20.
¯ σ ¯ σ
P (X − zα / 2 √ n ≤ µ ≤ X + z α / 2 √ ) = 1 − α
n
In a two-sided confidence interval, the α amount is split between the
two tails, thus we see α/2 or specifically, z α / 2 in the formula.
9 / 31
Confidence Interval for µ - Normal parent population, known
σ2
100(1-α)% Confidence interval on the mean, variance known If
x¯is the sample mean of a random sample of size n from a normal
population with known variance σ 2 , a 100(1 − α)% confidence
interval for µ is given by
σ
x¯− z α / 2 σ√ ≤ µ ≤ x¯+ z α / 2 √
n n
ANS: L = x¯ −0.10
z ·√σ0.100
= 12.050 − 1.28 √ = 12.037.
· n
100
U = x¯ + 0.10z √σ0.100 = 12.050 + 1.28 √ = 12.063.
· n
·
100
The 80% confidence interval for µ is [12.037, 12.063].
We are 80% confident that the true parameter value lies in this interval.
NOTE: Because σ2 was very small and n was fairly large, we have a very narrow confidence
interval for µ (which is good).
11 /
Confidence Interval for µ - Normal parent population, known
σ2
The 95% CI is wider... i.e. All else being held constant, if you want to
be more confident you capture µ, you’ll have to make your net bigger.
12 /
Confidence Interval for µ - Normal parent population, known
σ2
Looking at the form of the confidence interval:
x¯ · √σ
z α/2 n
±
s˛¸x s˛¸x s˛¸x
Sampl multiplier valu
e mean based on ebased on
confidence σ and
level sample size
↑
changes
for different
% CI
↑
standard
error of the
sample
mean 13 /
Confidence Interval for µ - Normal parent population, known
σ2
For example, the 95% CI for box fill weights was: [12.030,
12.070]
15 /
Confidence Interval Interpretation
For a 95% CI, we are 95% confident that the true µ lies in the
interval. This statement of confidence reflects the following...
16 /
Confidence Interval Interpretation & Simulation
17 /
Sample Size Calculation for µ
E=|x¯ − µ|
18 /
Sample Size Calculation for µ
CI
. zα / 2 ·
Σ2
Pre-specified error: E = zα / 2 √σ
n
⇒ n= σ E
·
Sample size for estimating µ with 100(1-α)% confidence and error E:
. z
α/2 · σ
n=
Σ2 E
19 /
Sample Size Calculation for µ
NOTE: Read sample size problems closely to determine if they are giving precision as a
half-width of a CI which is E (the cushion up or down), or the full width of the CI which is
2E. 20 /
One-sided Confidence Bounds for
µ
√
µ ≤ x¯ + zασ/ n and this gives an interval (−∞, x¯+
√
zασ/ n).
(an upper bound on µ)
√ √
x¯ − zασ/ n ≤ µ and this gives an interval (x¯ − zασ/ n,
∞).
(a lower bound on µ)
21 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
What if we don’t know σ? Can I just plug-in my estimator for σ
(or s) and again have the same 95% CI?
Σ
2 2
(x i − 2
σˆ = s =
n−
x¯)
1
This was a 95% CI for µ when σ was known
σ σ
x¯− z0.025 √ ≤ µ ≤ x¯+ z0.025 √
n n
where...
23 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
Connection to the Z-distribution...
X¯ − µ
Z = σ/ √n ∼ N (0,
1)
A T random variable follows a t-distribution
X¯ − µ
T = S/√n ∼ n−1
t
where tn−1 is a t-distribution with n − 1 degrees of freedom.
24 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
What does the t-distribution look like?
There is only one Z-distribution, but there are many t-distributions
(distinguished by their degrees of freedom df as tdf ). They look a lot
like the N (0, 1), except they have heavier tails.
For estimating a single parameter µ, the degrees of freedom is n − 1.
The heaviness of the tails depends on the degrees of freedom
(the subscript on the t), so it depends on the sample size n.
Differing t-distributions are shown below with df = k.
25 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
For a large sample size n, df = n − 1 is very large, and the tn−1 looks
just like the N (0, 1).
26 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
s s
µ
x¯− t α/2,df √ ≤ ≤ x¯+ t α/2,df √
n n
27 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
How do I get the t α/ 2 ,n−1 value? Similar to getting a z-value.
When α = 0.05 (for 95% CI) and the sample size is n = 10,
t α/2,n−1 = t0.025,9
This is the t-value for a t9 distribution with 2.5% above and 97.5%
below. Looking at the table...
t0.025,9 = 2.262
28 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
Example (CI for µ using t-distribution)
Suppose a sample of size n = 10 is taken from a normal population and
x¯= 8.94 and s = 4.3. Construct a 95% CI for the population mean.
Upper end-point: .
x¯ + α/2,n−1
t √s
n
= x¯+ 0.025,9 ·√ n
s
= 8.94 + Σ √4.3
10
=
·
Lower end-point: t 2.262 10.02
.
x¯ −α/2,n−1
t √s
n
= 8.94 − Σ √4.3
10
=
· 2.262 5.86
We are 95% confident that the true mean µ is between 5.86 and 10.02.
29 /
Confidence Interval for µ - Normal parent pop’n, unknown
σ2
30 /
100(1 − α)% Confidence Interval for
µ
RULE OF THUMB
σ σ
When σ is known, use x¯− z √ ≤ µ ≤ x¯+ z √
α/2 α/2
n n
s s
When σ is unknown, use µ
x¯− t α/2,df √ ≤ ≤ x¯+ t α/2,df √
n n