Eliminate some Type I error, but only by increasing Type II
error
All of the above
We conduct scientific experiments to measure the heat produced by
a chemical reaction. The average reading was 10 calories of heat
produced. If the standard deviation was 2 calories, and the
measurements are normally distributed...
What percentage of measurements were less than 10 calories?
50%
10%
20%
100%
What percentage of measurements were greater than 10 calories?
50%
10%
20%
100%
What percentage of measurements were between 8 calories and 12
calories?
16%
95%
34%
68%
What percentage of measurements were less than 6 calories?
2%
8%
16%
47%
What percentage of measurements were greater than 16 calories?
Less than 10%
Over 50%
Less than 0%
Less than 1%
What percentage of measurements were between 6 and 14 calories?
Less than 1%
About 95%
About 5%
About 50%
If a scientist got a measurement of 2 calories, how should you
regard that measurement?
Very suspicious: it is massively unlikely.
Nothing to see here; move along!
It definitely confirms the scientists was faking their
results.
Not enough information.
To a mathematician, the definition of probability is:
a personal belief
a real number between 0 and 1
an integer greater than one
a question Tornike will never get right
There are four suits in a deck of cards: hearts, diamonds, spades,
and clubs. Each suit has 13 cards. So if you draw 24 cards
randomly, you should expect 6 of each suit.
Here is your actual outcome when Kenneth deals you the cards:
Suit
fo
fe
Hearts
11
6
Diamonds
2
6
Spades
3
6
Clubs
8
6
In this situation, the null hypothesis should be:
The deal is fair
Kenneth looks like a card shark, so the deal is most likely
unfair
Why is the professor picking on Kenneth?
Can't say
Using the above data, we get a chi-square score of:
90
18
9
1
If we had set α at .05, we should:
reject the null hypothesis
fail to reject the null hypothesis
conclude with certainty that Kenneth was cheating
not enough information
If we had set α at .01, we should:
reject the null hypothesis
fail to reject the null hypothesis
conclude with certainty that Kenneth was cheating
not enough information
If we look at the above data for what might be interesting, we
should conclude:
why are there only four categories?
all of those sixes look fishy
too many clubs
too many hearts; too few diamonds
We run the same experiment with Davia dealing, and we get:
Suit
fo
fe
Hearts
6
6
Diamonds
5
6
Spades
4
6
Clubs
9
6
We get a Χ2 of:
3.8
2.3
.7
12.8
If we had set α at .05, we should:
reject the null hypothesis
fail to reject the null hypothesis
conclude with certainty that Davia was cheating
not enough information
If we had set α at .01, we should:
reject the null hypothesis
fail to reject the null hypothesis
conclude with certainty that Davia was cheating
not enough information
Chi-square is best for
nominal data
ordinal data
ratio data
interval data
In the chi-square test we have been doing (one-sample), the degrees
of freedom equals
7
the number of categories
the number of categories plus one
the number of categories minus one
If an experiment reports p > .05, and α = .05,
we should