HP 48g Graphing Calculator - Manual (Page 282)

for that model. It also computes and returns the covariance (sample
or population) and the correlation coefficient for the regression.
To perform a regression
for
two variables
in
the current data;
1. Press (i^fsW) rri m to open the L'HTfl form.
FIT
EDAT:
K-CDL: 1 V-CDL: 2
MODEL: Linear Fit
ENTEIS STATISTICAL DATA_____________
The FIT DATA Screen
21
2. Enter or choose the data matrix containing the data you wish to fit.
3. Enter the independent variable in K-CuL; and the dependent
variable in Y-CuL:.
4. Choose one of the four regression models (or !3est Fit, which
automatically selects the model with the correlation coefficient with
the largest absolute value).
5. Press O.K . You will see the computed regression model on
level 3, the correlation coefficient on level 2, and the covariance on
level
1 .
"1 use His -i;. islet Led
/ariabis;
1. Press (i^fswimm OK to open the FIT DFiTi-i form.
2. Enter or choose the data matrix containing the data you wish to fit.
3. Enter the independent variable in Y-CuL.s and the dependent
variable in Y-COL:.
4. Choose one of the four regression models (or Best Fit, which
automatically selects the model with the correlation coefficient with
the largest absolute value.
5. Press PF;ED to display the PREDICT VALUES form.
6 . Enter the assumed value in either the tb or the Yi: field.
7. Move the highlight, if necessary, to the variable field whose value
you wish to predict, and press PRED . The computed value is now
displayed in the field. Press EDIT to view the complete number.
21-10 Statistics and Data Analysis