Answers for Topic 22, AP Stats 1998-1999

Activity 22-1

a) P(N)= 7/10 = .7 and P(O) = 5/10 = .5
According to this treatment N is better due to a higher percent of recovery's

b-f) answers will vary depending on your data, but recall in class that our curve tended to be fairly 'normal' in shape
g-h) I ran the simulation and received the following output
Recovery    0    1    2    3     4     5     6      7      8      9     10 
Deaths 10 9 8 7 6 5 4 3 2 1 0
Number 0 0 0 7 70 234 367 231 74 15 2
i) In this experiment the result was 322 of 1000 which is a proportion of .322 or basically a 1 in 3 chance

j) Yes, for this to occur it would be extremely unlikely due to pure chance.

Activity 22-2

f) According to these results the chances of this result occurring are essentially zero.

g) According to our experimental data there is extremely good evidence that AZT is more effective than a placebo in preventing transmission of HIV to newborn babies.

Activity 22-3

a) To make any judgements of any sort we need to know the sample sizes involved.

b) For these results a sample size that is very small would not influence the validity of the claims.

c) If the sample sizes were large in size, it would be more likely to reinforce the validity of the claims.

d) Table results are below
Old New p-value sig = .1 sig = .05 sig = .01
50 50 .1473 No No No
100 100 .0691 Yes No No
200 200 .0180 Yes Yes No
500 500 .00046 Yes Yes Yes

e) The larger the sample size, the more convincing the results of 60% and 70% of recovery rates become.

Activity 22-4

a) If that both treatments are "as favorable" the there should be no difference in the two groups. Then you would expect there to be only 6 or less on the histogram, so in this case the total is 10 or proportion of 10/1000 or p= .01

c) The two values are quite close .0094 vs .01.

d) There is considerable evidence at alpha = .01 level that the literature program made in a difference in the whether of not the parolee would commit another crime after being released. (Personal note: This kind of study shows how totally unrelated things can seemingly seem to make a difference. Recall in class, a survey I mentioned about people who like chocolate cookies are more likely to commit crimes than those who do not!)

Activity 22-5

In this case we would reject the null, and state that the 'wording' of a statement can make a difference in the results of a survey.

Activty 22-6

In this case again, we would reject the null, due to random chance this result would rarely happen, actually practically a zero chance. Thus, there is a difference in how a question is asked.

c) Remember that this problem deals with the proportion of the population who think X-rated moves should not be shown. thus

This again indicates that this is significant at the 0.01 level, so we reject null which stated there is no difference in how the question is asked, and state there is evidence that there is a difference in how the question has been asked.

d) The questions states to use the same ideas as parts a through c, so once again, we use the 'opposite' case, in that we are trying to determine if the goverment should not allow cigarette advertising on TV. At this point, I hope you are getting comfortable with the 2 proportion z-stat. formula, so I will instead use the calculator instead and just state the values needed.

Again we would reject the null, and accept the alternative hypothesis.

e) In two of the cases there is extremely strong evidence, but keep in mind that we also have other factors which may not be part of what we are studying. Examples could be anti-communist feelings, anti-pornography feeling, etc.

Activity 22-7 Remember that this survey is about forbidding smoking in on-campus building.

Please send any questions to me on the form below, I will be checking my mail daily.

Your email address:
© 1999 by DeeAnne Doseman-Flaws