95% confidence intervals

[Trond]

I just came up with this, after flipping through tables and books to do each stage of the confidence intervals. What do you think?

HP 67 Program: 95% Confidence intervals, one variable:
Assumes normal distribution. I have obtained the coefficients from classic statistics literature on t-distributions and normal distributions.

Collect data (one variable) with ∑+.
Run A for 95% confidence interval of a sample size n=5; B n=10; C n=15; D n=20; E n=30 or above
(E is exact at 60 but a good estimate; for other samples you can also try the next lower and higher value for an idea)

The program puts the +/- value around the mean in in the X register and in variable Reg.1, the lower limit of the interval in R2,
the upper limit of the interval in R3, number of data n in RA, the mean in RB (and Y register), the standard deviation in RC, and Student t value in RD



LBL A
2
.
7
7
6
STO D
GSB 1
RTN

LBL B
2
.
2
6
2
STO D
GSB 1
RTN

LBL C
2
.
1
4
5
STO D
GSB 1
RTN

LBL D
2
.
0
9
3
STO D
GSB 1
RTN

LBL E
2
STO D
GSB 1
RTN

LBL 1
P<>S
RCL 9
P<>S
STO A
Mean X
STO B
Stdv
STO C
RCL A
SQRT
/
RCL D
x
STO 1
RCL B
+
STO 3
RCL B
RCL 1
-
STO 2
RCL B
RCL 1
RTN


Example:
What is the mean weight of adult blue whales in the wild, with a 95% confidence interval, assuming normal distribution?
Say that we have the following sample of weights in metric tons from old whaling records:
90, 176.5, 95, 60, 92, 80, 100, 120, 62, 98
Collect data with ∑+
Run B: the +/- value is in the x register and the mean in the Y register : 97.35 tons +/- 23.63, or check by RCL 2 and RCL 3 (73.72-120.98 tons)