A Performing a Specific Regression Calculation

Example: To use the statistical data shown below to perform logarithmic regression to obtain the correlation coefficient and the estimated value of y when x = 100

xy

291.6

5023.5

7438.0

10346.4

11848.9

1.Display the logarithmic regression calculation result screen.z6(RESULT)2(Reg)3(Log)2.Press J to return to the STAT editor screen.3.Press z1(/COMP) to display the COMP Mode calculation screen.4.Determine correlation coefficient r and the estimated value of y when x = 100.• Correlation Coefficientz7(STAT)2(VAR) ccc4(r)E• Estimated Value of y when x = 100100z7(STAT)2(VAR) ccc7(n)E

Note

The r and n values produced by this example are for logarithmic regression because we first displayed the logarithmic regression calculation screen in step 1 before using the commands to calculate the values were wanted. If you do not display a particular regression calculation result screen first, the values produced by the commands are for linear regression by default.

For details about statistical calculation commands, see “REG Mode Statistical Command Reference” (page 80).

A REG Mode Statistical Command ReferenceMean and Standard Deviation Commands

n

z7(STAT)2(VAR)1

Obtains the number of samples.

n = (number of xi-data items)

E-80