Final Exam Review Flashcards
(142 cards)
1
Q
A
Quantitative
2
Q
A
Categorical
3
Q
A
Individual Houses
4
Q
A
States
5
Q
A
Variable = Percent of state residents with a college degree. This is a quantitative variable
6
Q
A
Residents of Connecticut
7
Q
A
Variable = whether or not they have a college degree. This is a categorical variable.
8
Q
A
30 leaves selected from this oak tree
9
Q
A
All oak tree leaves
10
Q
A
Not biased
11
Q
A
Not biased
12
Q
A
362, 17630, 1891, 3570, 1269, 1590;
13
Q
A
Matched Pairs experiment;
14
Q
A
Randomized comparative experiment
15
Q
A
Observational
16
Q
A
College fairs
17
Q
A
Observational study
18
Q
A
No, we cannot make that ocnclusion. This wa snot a randomized experiment ( which is the only way to conclude causation), and thus there could be another lurking variable that could explain why one representative tended to be more effective
19
Q
A
Media usage in the bedroom = Categorical
“Tiredness” rated on a scale = Quantitative
20
Q
A
Media usage in the bedroom = Explanatory
“Tiredness” = Response
21
Q
A
0.5307; calculated by dividing (college graduate - believe) by (total college graduates)
22
Q
A
23
Q
You need to draw out the table - it’s already filled out here - just look at labels
A
24
Q
A
Graph (b); It shows more parochial
25
If a distribution is heavily skewed to the left, which relationship between the mea and median is most likely?
mean < median
mean = median
mean > median
Mean < median; remember that outliers will heavily alter the mean so it's likely smaller or larger depending on how the graph is skewed
26
Skewed to the right
27
- The mean is likely to be larger than the median
- The mean is likely to be smaller than the median
- The mean and median are likely to be approximately equal
- the mean and median are likely to be approximately equal
28
- The mean is likely to be larger than the median
- The mean is likely to be smaller than the median
- The mean and median are likely to be approximately equal
The mean is likely to be smaller than the median
29
- 0.5
- 10
- 5
- 1
5; The scores are moderately spread around the mean, and estimating from the graph, the standard deviation of about 5 is the closest and most reasonable choice
30
Which statistic is more resistant to outliers (or extreme data values)?
- interquartile range
- standard deviation
Interquartile range; SD measures how much data points deviate from the mean; outliers have a huge impact but that is not the case with the IQR
31
--Answers entered, Careful--
mean = 68 (based on highest point/center in balanced distribution)
SD = 4.5; This gives doundaries from 59 to 76 which is about 8 or 9 units above and below the mean; we estimate the SD to be about 4 (in the range from 3 to 5)
32
--Caution; Answers Inserted--
3.17 standard deviations ABOVE the mean
33
The interval is 130 to 230;
lower: 180-2(25) = 130
upper: 180 + 2(25) = 230
34
- Roughly symmetric
- slightly skewed to the left
- slightly skewed to the right
Slightly skewed to the right; mean is bigger than the median
35
Yes - it is well out of the IQR (Q1-Q3)
36
Women; the IQR is larger
37
Kansas
38
Kansas
39
Iowa
40
South Dakota
41
- 0.56
- 0.63
- 0.70
- 0.91
0.63; population proportion: center of the distribution; the center (the mean of all sample proportions) estimates the population proportion (p)
42
Higher n value would have made the sample statistics have less variability
43
bar-x = (sum of all data points)/n; Extremely unlikely to ever occur
44
The sample means would be centered at the same value
45
p = proportion of German adults who exercised for 30 minutes three or more times per week
46
0.75; 5840/7786 = 0.7501
47
Sample a = 250
48
US adults (age 18 and over)
49
1012 randomly selected us workers
50
p = proportion for population/population parameter
51
p-hat is statistic = proportion of the sample of 1012 randomly selected us adults who are dissatisfied = 0.53
52
Correct/Incorrect
Incorrect
53
Correct/Incorrect
Incorrect
54
8.7 - 10.1
55
Possible
56
mhu = mean number of hours per week all customers use the internet
57
x-bar = 7.2 hours
58
6.28 - 8.12 hrs
59
6.322 hours to 8.082 hours
60
6.438 - 7.947 hours
61
the probability, when the null hypothesis is TRUE, of obtaining a sample as extreme as (or more extreme than) the observed sample
62
Sample D
63
Option e
64
B
65
0.8
66
98 because that is the size of the original sample
67
p-hat
68
x-bar1 = 18.3, b-bar2 = 15
69
rejecting the null hypothesis when the null hypothesis is true
70
Stay the same
71
Reject null hypothesis
72
73
Option D
74
This sample provides strong evidence that male students watch more television each week than female students at this university
75
Option A
76
This sample provides strong evidence to reject the null hypothesis and conclude that more than 1/3 of households in this city own pets
77
Option D
78
0.231
79
No, significance level of 10%
80
Test A
81
Discrete random variable
82
Continuous random variable
83
95%
84
0.064
85
0.520
86
-0.842
87
1.150
88
0.309
89
77 inches
90
52 beds
91
57.2 inches
92
177.5 seconds
93
155.0 seconds and 359.0 seconds
94
2
95
-0.833
96
1.875
97
0.031
98
Yes
99
6.51 to 7.19 hours
100
$3.86 - $4.09
101
n = 200
102
0.737 to 0.763
103
0.6625
104
Yes
105
6.3 - 8.1
106
Not appropriate
107
0.052 to 1.92
108
0.07
109
Yes
110
It has no effect on the center of the distribution
111
It decreases the spread of the distribution
112
49
113
Yes; even though the data are skewed for each sample, the samples sizes are rather large
114
-0.40 to 3.76; n = 50 - this is aa reasonably large sample size (an no real reason to suspect that heights are drastically skewed) so we can use a t-distribution to construct the confidence interval
n=5-0, so df = 49 and this t* = 2.68
115
1.688;
116
0.052
117
1.685
118
This statement is not correct, because the Central Limit Theorem only applies to sample statistics not individual samples
119
Matched pairs experiment; since the "treatment" students are each matched with a similar "control" student. We use paired data analysis.
120
Difference in means with two separate groups; Since there are two separate groups of men doing the rating after being randomly assigned to one of the conditions, this is a difference in means using two separate groups
121
C is the correct answer because r = 0.09 is positive even if it is weak. A isn't acceptable because it has a slight negative trend
122
B; the correlation shows a strong negative association, so the matching scatterplot
123
D/Last one
124
Both variables in the study are categorical, not quantitative. Correlation is a statistical measure of the linear relationship between two quantitative variables
125
The last option is correct because the least squares regression line formula uses the computed slope (0.208) and intercept(89), which best fit the data by minimizing the sum of squared residuals between actual and predicted enrollment numbers; y = a +bx
y = predicted value of the dependent variable
x = the independent variable
a = the y-intercept
b = the slode of the line (the change in y for a one unit change in x)
126
Explain least squared regression line formula
The line that best fits the data by minimizing the sum of the squared differences between the actual data points and the predicted points on the line
127
9 - it's way down there and below the ling; residual=actual−predicted
128
4; highest value above the line of regression
129
5; closest to the line
130
Positive; this means that people longer forearms rend to have larger feet
131
132
It would make the correlation weaker; reinforces the positive linear relationship between mass and length
133
If amount of fat consumed is 1 gram more, predicted number of calories goes up by 18
134
Parameter, p; remember that p-hat is the sample proportion -> p-hat = x/n (x is the # of successes/people with a certain characteristic and n = the total number of observations in the sample
p=hat = 30/100 = .3 <--like if 30/100 people like a certain product
135
r
136
No; the bootstrap distribution is extremely skewed - bootstrap does not need to be bell-shaped, but not extremely skewed either
137
Sample A
138
parameter p = correlation between point spread and steal differential for this team. Hypothesis H0 = p = 0; Ha p > 0; p represents parameter
139
Count number of dots above r (6) and divide by total sample size (100) = 0.06
140
141
There is very strong evidence that the population slope differs from zero, and thus is an effective model for predicting this response variable.
142
Yes;
- There appears to be a curved pattern in the data more than a linear pattern
- The variability seems to be much greater for low levels of alkalinity than for higher levels
- There appear to be two outliers in the middle of the plot that are quite far above the trend in the rest of the data
