Table of Contents
Advertisement
T
&
T
T
&
T
I
P
S
R
I
C
K
I
P
S
R
I
C
K
New columns as functions of old
As with univariate statistics, you can use
functions of old columns as new sets of data.
See the Univariate version of this section for
two different ways of doing this.
For example, a set of data (C1,C2) that you suspect is exponential could be
straightened by setting up S2: as (C1,LN(C2)).
The effects of changes of scale and origin on data and summary statistics
can be investigated in this way by storing (say) -2*C2+3 into C2. You can
even combine columns in this way, such as storing C1+C2 into C3.
Using values from
It is often useful to be able to retrieve values such as the mean and standard
deviation for use in further calculations. With most simpler calculators these
values are found by pressing keys rather than reading from a
so doing a calculation like 'multiply the mean by 3.5' is not hard. The values
shown on the
relatively easily.
For example, the set of data below contains a suspected outlier (erroneous
value). In this case one might suspect a missing comma.
{2, 3, 5, 2, 1, 5, 3, 6, 7, -2, 3, 5, 5, 55}
A common test for outliers is to calculate the
mean and standard deviation without the
presence of the suspected outlier, and then to
check whether the suspect piece of data is
within three standard deviations of the mean.
If not, then it is discarded.
Enter the data without the suspected outlier into column C1 with the
calculator in
then press the
-
B
-
B
S
I
V
A
R
I
A
T
E
S
I
V
A
R
I
A
T
E
in calculations
screen can also be retrieved for use on the hp 39g+
mode. Ensure that the SYMB view is set up correctly and
key.
D
D
A
T
A
A
T
A
143
screen,
Hide quick links:
Advertisement
Table of Contents
