Information on Using the Regression Feature
of the TI-83 calculator
DIAGNOSTICS ON - this should only be done
ONCE, before you do your first regression
2nd
– 0 (for CATALOG)
Type
D – and arrow down to DiagnosticsOn – and press ENTER
Entering Data
STAT
– 1:Edit
If
the lists L1 and L2 have values, these must be cleared out. To empty them, arrow up to the VERY top of
list L1, i.e.highlight L1, and press CLEAR, similarly L2, etc.
Then enter data x values into L1 and y values in L2
Viewing a Scatter Plot
2nd
– STAT PLOT (the Y= key or above the
2nd key)
1: ENTER,
On ENTER ¯ENTER ¯L1 ¯L2
- (this defines Plot 1 as a scatter plot L1 & L2)
Make sure that in Y=, all the functions are cleared out (or turned off)
ZOOM-9:ZoomStat will show a
Scatter plot of the data
Getting the Regression Equation
STAT ® CALC ¯¯¯ to 0:ExpReg ENTER ?ENTER
Record
the equation indicated for the values of a and b, and record r
Graphing the Regression Equation
In
Y=, enter the regression equation. You
may type it in by hand and graph. It should graph
the function over the scatterplot points.
**Note: A shortcut to typing the function into Y1 is in the Y= screen, at
the Y1= location,
paste in the function by pressing the following
sequence of keys:
VARS 5:Statistics ®®EQ 1: RegEq ?ENTER
***To turn off ALL Plots *** - 2nd – STAT
PLOT - 4:PlotsOff -
ENTER
Types of Regression Equations on the TI-83
|
QuadReg - |
ExpReg - |
|
CubicReg - |
PwrReg - |
|
QuartReg - |
Logistic - |
|
LnReg - |
SinReg - |
Example: Online Shopping
The commercialization of the Internet is marked by
massive increases in advertising and online
shopping.
The accompanying table illustrates this trend. Sales are in millions of dollars.
|
Year |
1996 |
1997 |
1998 |
1999 |
|
Online Sales of Recorded
Music Enrollment |
18 |
47 |
110 |
240 |
a) Find the exponential function to model this data.
Let t = 0 represent
1995. Enter the years into list L1, and
the sales into list L2. Define,
using STAT PLOT, Plot 1, as
a scatter plot on L1 and L2. Press
Zoom:9 to see the scatter plot.
STAT-CALC-0 – ExpReg
ENTER, gives f(x) = 7.966*2.368x. Enter this function into Y1,
and graph. Neat!! Huh!!
Also, let’s rename this
function into the y = Aekx form. To do this, realize
that 2.368 = ek, so
k = ln 2.368 » 0.862.
Thus the function could be renamed as f(x) = 7.966e.0.862x
b) Use the function to predict the online sales in 2002 … so find f(7) = 7.966e.0.862(7)
=3,326.
Another Example
**Remember to clear out the
functions in Y= between problems**
Example: U.S.
Population Model
The following data obtained from the U.S. Census
Bureau represent the population of the United
States. An
ecologist is interested in finding a function that describes the population of
the U.S.
|
Year |
1900 |
1910 |
1920 |
1930 |
1940 |
|
Population |
76,212,168 |
92,228,496 |
106,021,537 |
123,202,624 |
132,164,569 |
|
1950 |
1960 |
1970 |
1980 |
1990 |
|
151,325,798 |
179,323,175 |
203,302,031 |
226,542,203 |
248,709,873 |
a) Let t = 0 represent 1900.
Enter the years into L1 and population into L2. Define, using STAT
PLOT, Plot 1, as a scatter
plot on L1 and L2. Press Zoom:9 to see
the scatter plot.
b) STAT ®CALC ¯¯ -B:Logistic ENTER, (wait
patiently …, it’s VERY slow!!) gives
f(x) = 695129657/(1+7.9105 e-0.0166223kx) Enter this function into Y1, and graph over the points
on the scatter plot.
b) Use the function to predict the population of the U.S. in 2001,
this years census. (I got 280,744,859).
c) If this model continues, when will the population of the
U.S. reach 300 million?
Let Y2=300000000. Extend the WINDOW so that Ymax is over
300,000,000, and Xmas is 120.
Graph … and find the
intersection point – t = 107.9 … so approximately in 2007.