Transcript Conf Int on TI

Computing Confidence Intervals using the TI-83
The TI-83 can compute an ENTIRE
confidence interval from either
summary statistics or data.
These functions can be accesed
by pressing STATTEST
1
Estimating  from a Large
Sample
If n > 30 the sample is considered to be
large, regardless of its distribution’s
shape.
Press STAT, choose TESTS
and choose 7:ZInterval
2
Example
• A study of 50 iris flowers revealed a mean petal length of 2.03
cm, sample standard deviation of 0.27 cm. Compute a 90%
confidence interval for the population mean petal length.
• x-bar = 2.03, s = 0.27, n = 50 are our summary statistics.
• Since n>30 okay to use Zinterval as well as s in place of 
• Step 1: Press STAT, selects TESTS, select Zinterval
• Step 2: For method of input (Inpt:) select STATS, since we
have summary statistics.
• Step 3: Enter 0.27 for , 2.03 for x-bar and 50 for n.
• Step 4.Set C-Level( Confidence Level) to .90
• Step 5: Select Calculate. (Double Check Entries First)
3
Screen Shots
4
Estimating  from a Small
Sample
If n ≤ 30 the sample is considered to be
small, the population must be normal
and  unknown.
Press STAT, choose TESTS
and choose 8:TInterval
5
Example
• Consumer Reports gave the following
information about the life(hours) of AA
batteries in toys.
2.3
2.5
4.2
6.1
5.7
5.5
1.3
1.5
5.3
1.8
1.9
5.2
1.8
5.1
1.6
5.4
• Assume the population is normally distributed,
compute a 95% confidence interval to estimate
the true mean life of AA batteries in toys.
6
Example
• Since n = 16 which is less than 30,  is unknown and
the population is normal we use a TInterval.
• Since we have data we must enter it into a list.
• Step 1: Press STAT, choose Edit, enter the values into
L1.
• Step 2: Press STAT, select TESTS, select 8:Tinterval.
• Step 3: Select Data for method Input.
• Step 4: Enter L1 for List, 1 for Freq
• Step 5.Set C-Level( Confidence Level) to .95
• Step 6: Select Calculate. (Double Check Entries First)
7
Screen Shots
8
Estimating p
If the sample is a SRS, binomial and
np  5 and nq  5 are both satisfied.
ˆ
ˆ
Press STAT, choose TESTS
and choose A:1-PropZInt
9
Example
• Example: In a survey of 2503 men and women aged 18 to 75
years and representative of the nation as a whole, 1927 people
said the homeless are not adequately assisted by the
government. Find a point estimate and a 90% confidence
interval for the proportion p of adults in the general population
who agree that the homeless are not adequately assisted by the
government.
• n = 2503, x = 1927, since we have more than 5 success and 5 failures it is
okay to use a 1-PropZInt.
• Step 1: Press STAT, select TESTS, select A:1-PropZInt.
• Step 2: Enter the number of success for x and sample size for n.
• Step 3.Set C-Level( Confidence Level) to .90
• Step 4: Select Calculate. (Double Check Entries First
10
Screen Shots
11