410 CHAPTER 6 Inferential Statistics
PROGRAM: CHISQ
: Input“ DF ”,N
: 0→S
:For(I, 1 ,N+ 1)
:S+ (L 1 (I)−L 2 (I))^2 /(L 2 (I))→S
:End
: 1−χ^2 cdf(0,S,N)→P
:Disp“CHISQ:”,S
:Disp“P-VALUE”,P
Running the above program (using the fact that there are 5 degrees
of freedom) results in the output:
CHISQ : 5. 98
P−VALUE :. 308
Example 3.^31 Suppose that before a documentary was aired on pub-
lic television, it had been determined that 7% of the viewing public
favored legalization of marijuana, 18% favored decriminalization (but
not legalization), 65% favored the existing laws, and 10% had no opin-
ion. After the documentary was aired, a random sample of 500 viewers
revealed the following opinions, summarized in the following one-way
table:
Distribution of Opinions About Marijuana Possession
Legalization Decriminalization Existing Laws No Opinion
39 99 336 26
Running the above TI code yielded the following output:
CHISQ:13.24945005
P-VALUE:.0041270649
This tells us that there is a significant departure from the pre-existing
proportions, suggesting that the documentary had a significant effect
(^31) This example comes fromStatistics, Ninth edition, James T. McClave and Terry Sinich,
Prentice Hall, 2003, page 710. (This is the text we use for our AP Statistics course.)