User's Manual

56 Section 4: Statistics Functions
Linear Estimation. With the statistics accumulated, an estimated value for
y, denoted ŷ, can be calculated by keying in a proposed value for x and
pressing ´j.
An Estimated value for x (denoted
x

) can be calculated as follows:
1.
Press ´L.
2.
Key in the known y-value.
3.
Press ® - ® ÷.
Correlation Coefficient. Both linear regression and linear estimation
presume that the relationship between the x and y data values can be
approximated by a linear function. The correlation coefficient, r, is a
determination of how closely your data fit a straight line. The range is -1  r
 1, with -1 representing a perfectly negative correlation and +1
representing a perfectly positive correlation.
Note that if you do not key in a value for x before executing ´j, the
number previously in the X-register will be used (usually yielding a
meaningless value for ŷ).
Example: What if 70 kg of nitrogen fertilizer were applied to the rice field?
Predict the grain      
the correlation coefficient is automatically included in the calculation, you
can view how closely the data fit a straight line by pressing ® after the
y prediction appears in the display.