Solve TI84 Regression Issue: 2 Calculators, 1 Result

[barryj]

I have two TI84 calculators. I entered the following data into L1 and L2 and tried to do a linear regression. My friend also had a TI84 and entered the same data into his calculator. We checked, double checked, and triple checked to see that the data was the same in all three calculators. Here is the data.

1980 1745
1982 1711
1984 1688
1986 1657
1988 1642
1990 1611

On my two calculators the regression was about 0.5 On my friends calculator the regression was about .99.
I was extremely puzzled as to why my two calculators gave vastly differences. So... we tried the RESET thing. Then all three calculators read the same, r = .99.

I am concerned as to what made my calculators give the erroneous answers until RESET was performed. Does anybody know of a setting that could have made a difference.

At first my friend thought his calculator was broken because my two calculators gave the same result but after doing the resets, my two calculators began to give the correct answer. I would hate to take a test knowing that there is some mysterious setting that will make the regression not work properly.

 

[jedishrfu]

Could you have had data from a prior linear regression still lurking inside?

In this article, they mention some holdover register r^2 and maybe it still had results still in it.

http://www.calcblog.com/performing-a-linear-regression-on-the-ti-83-or-ti-84/

Of course, there may be other STAT registers holding stuff too.

Here's some directions on doing a linear regression that you can compare to the way you do it now:

https://epsstore.ti.com/OA_HTML/csk...nge=null&fStartRow=0&fSortBy=2&fSortByOrder=1

 

[barryj]

I think I am doing the regression as per the articles above.
There must be some lurking variable that I cannot find or understand.
Now, I don't trust my TI84 anymore as this could happen again as I don't know how this condition was setup.