Download Study Guide

PSY 211: Exam #4
Course Reference # 22032197
Mike Hoerger
Name: _____________________________
Part I. Multiple Choice (4 points each)
Choose the best available response.
1
This statistic is not an example of an effect size
a) p -- 93%, r = .27
b) r
c) d
2
What is a hypothesis?
a) Broad research question, aim, or goal motivating a particular study
b) Detailed rationale explaining the logical basis for an observed finding
c) Concise, testable statement describing the relationship between two
variables -- 100%, r = .00 (too easy)
3
Drug companies generally design drugs and then try to find a use for them at a
later date. In one study, they examine whether a new drug impacts appetite.
What type of hypothesis test should they use?
a) No-tail test
b) One-tail test
c) Two-tail test -- 73%, r = .16
4
What did William Kingdon Clifford say?
a) “The danger to society is not merely that it should believe wrong things,
though that is great enough; but that it should become credulous, and
lose the habit of testing things and inquiring into them; for then it must
sink back into savagery.” -- 84%, r = .36
b) “But if we are empiricists, if we believe that no bell in us tolls to let us know for
certain when truth is in our grasp, then it seems a piece of idle fantasticality to
preach so solemnly our duty of waiting for the bell.”
c) “It is better to be a human being dissatisfied than a pig satisfied, better to be
Socrates dissatisfied than a fool satisfied. And if the fool or the pig are a
different opinion, it is because they only know their own side of the question.”
5
A null hypothesis states describes that there are no differences expected to
occur at this level:
a) measurement
b) sample
c) population -- 73%, r = .18
6
Effect size for r2 = .26
a) small
b) medium
c) large -- 89%, r = .27
7
The null hypothesis indicates:
a) the effect will be negative, or opposite of expected
b) an obtained difference will exceed chance
c) no effect or no important difference -- 96%, r = .00 (too easy)
8
Term indicating that results from any two particular studies might differ slightly
due to chance:
a) sampling error -- 78%, r = .27
b) standard error of the mean
c) carryover effects
9
The results of a multiple regression indicate that R2 = .20, p = .50. Is this result
statistically significant?
a) non-significant -- 78%, r = .27
b) significant
c) not enough information
10
Using a more strict alpha level for a z-test will cause the critical z-value to
a) become more extreme -- 58%, r = .55
b) stay the same
c) become less extreme
11
What type of test might you use to determine whether female doctors reliably
differ from their romantic partners on IQ?
a) correlation
b) between-group t-test
c) within-subject t-test -- 69%, r = .73
12
What does a p-value describe?
a) the probability that the result is reliable
b) the probability that the result is due to chance -- 89%, r = .36
c) the probability that the result will replicate
13
Assume that the alternative hypothesis is accurate. If the researcher finds no
effect, what has occurred?
a) Type I error
b) Type II error -- 87%, r = .18
c) Iatrogenic error
14
Effect size for Cohen’s d = .36
a) small -- 82%, r = .36
b) medium
c) large
15
You have people rate how much they enjoy Coke and how much they enjoy
Pepsi. The mean enjoyment rating for Coke is a bit higher than for Pepsi, and
you find d = 1.25, t = 1.67, p = .63. How would you most likely get a result like
this?
a) large effect but a small sample -- 73%, r = .46
b) unusual sample
c) poor methodology or survey items
16
This was not an example of a repeated-measured t-test discussed in class:
a) Changes in tension before and after massage therapy
b) Changes in physiological arousal before and after crossing an old bridge
-- 85%, r = .16
c) Changes in sleepwalking before and after sleep deprivation
17
Which of the following distributions is the flattest?
a) z distribution
b) t distribution where n = 3 -- 80%, r = .64
c) t distribution where n = 37
18
Effect size for r = .42
a) small
b) medium -- 91%, r = .18
c) large
19
This type of test is not used to compare a treated sampled to an untreated
population.
a) z-test
b) single-sample t-test
c) between-group t-test -- 69%, r = .46
20
There is no “depression gene” yet every year some researcher claims to have
found it. The results are often reported on CNN and the other news networks,
who fail to realize they are making this error.
a) Type I error -- 84%, r = .27
b) Type II error
c) Hindsight bias
21
Carefully examine the Output below for a series of between-group t-tests. The
analyses compare “Early Birds” to “Night Owls” on five variables. Some
information is missing or covered up. How many of the five results show a
reliable group difference?
Group Statistics
68. Depress ion
79. Pro-Environment
104. Neuroticism
106. Agreeableness
108. Delay of Gratification
17. Sleep Schedule
Early Bird
Night Owl
Early Bird
Night Owl
Early Bird
Night Owl
Early Bird
Night Owl
Early Bird
Night Owl
N
95
184
95
184
95
184
95
184
95
184
Mean
3.55
4.01
6.83
6.83
4.68
5.08
6.99
6.66
6.18
5.52
Std. Deviation
2.123
1.994
2.272
2.024
2.335
2.231
1.533
1.591
1.896
1.740
Std. Error
Mean
.218
.147
.233
.149
.240
.164
.157
.117
.195
.128
Independent Samples Test
Levene's Test
for Equality of
Variances
F
68.
Depres sion
79.
ProEnvironment
104.
Neuroticism
106.
Agreeableness
108. Delay of
Gratification
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
a) 1 -- 62%, r = .91
b) 2
c) 3
.602
2.799
.598
.225
.986
t-test for Equality of Means
Sig.
(2-tailed)
Mean
Difference
Std. Error
Difference
277
.073
-.464
.258
-.971
.044
-1.764
180
.079
-.464
.263
-.982
.055
.000
277
1.000
.000
.267
-.525
.525
.000
172
1.000
.000
.277
-.546
.546
-1.368
277
.172
-.392
.286
-.956
.172
-1.349
183
.179
-.392
.291
-.965
.181
1.644
277
.101
.326
.199
-.064
.717
1.664
196
.098
.326
.196
-.061
.713
2.899
277
.004
.657
.227
.211
1.103
2.820
176
.005
.657
.233
.197
1.117
Sig.
t
df
.439
-1.799
.095
.440
.636
.322
95% Confidence
Interval of the
Difference
Lower
Upper
22
Using the SPSS Output on the previous page, calculate the t-value for the
group differences in Agreeableness, and choose the answer below closest to
that t-value.
a) 1.3
b) 1.7 -- 69%, r = .36
c) 2.1
23
Using the SPSS Output on the previous page, what would be the degrees of
freedom for the analysis examining group differences in Neuroticism
a) 277 -- 62%, r = .64
b) 278
c) 279
24
Study the following correlation table very carefully. The name of the first variable
is blacked out. Based on the pattern of correlations, what might this variable be?
Correlations
93. Happiness
45. Books
Read in Past
Year
Pearson Correlation
Sig. (2-tailed)
N
Pearson Correlation
Sig. (2-tailed)
N
59. Worrying
70. Parents
Fought
Growing Up
83.
Self-Es teem
Pearson Correlation
Sig. (2-tailed)
N
Pearson Correlation
Sig. (2-tailed)
N
Pearson Correlation
Sig. (2-tailed)
N
93.
Happiness
1
279
.036
45. Books
Read in
Past Year
.036
.553
279
1
.553
70. Parents
Fought
59.
Growing
Worrying
Up
-.284**
-.111
.000
.065
279
279
-.070
-.044
.243
279
279
279
-.284**
.000
279
-.111
.065
279
.587**
.000
279
-.070
.243
279
-.044
.462
279
.017
.775
279
1
???
279
.181**
.002
279
-.316**
.000
279
.462
83.
SelfEs teem
.587**
.000
279
.017
.775
279
279
.181**
.002
279
1
-.316**
.000
279
-.123*
.040
279
1
279
-.123*
.040
279
279
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
a) depression
b) ACT score
c) happiness -- 56%, r = .46
25
What do you suppose the value would be for the correlation in the “???” box?
a) r = -.11
b) r =.02 -- 80%, r = .23
c) r = .08
Part II. Calculations [10 points]
26. Z-test – 4 points
Calculate a z-score, given the following
information, and indicate whether it is
significant.
Sample mean = 18
Sample mode = 22
Sample standard deviation = 10.75
Population mean = 20
Population mode = 16
Population standard deviation = 12.45
Sample size = 256
Z = -2.57
Significant?
yes / no
[Only need information in bold]
SE = 12.45 / sqrt(256) = 12.45/16 = .778
Z = (18-20) / .0486 = -2.57
critical Z = + 1.96
27. t-test – 6 points
28 people enroll in a clinical trail for a new
schizophrenia medication. During the trial,
ten patients drop out, and three more commit
suicide. On a symptom rating scale, patients
see an average reduction of 2.2 points (SD =
4.3) for their symptoms. Calculate the withinsubject t-value. Indicate whether the result is
significant. In a sentence or two, describe
how you would interpret this finding.
n = 28 – 10 – 3 = 15
df = 15 – 1 =14
critical t = + 2.145
SE = 4.3 / sqrt(15) = 1.110
t = 2.2 / 1.110 = -1.9815
t = 1.98 (positive or negative is okay)
Significant?
yes / no
Interpretation:
Although patients had a slight
reduction in symptoms, the change
was not reliable. The high dropout
rate also calls into question the
results.
Part III. Short Answer [10 points]
28. Pick One of the following:
a) List 5 types of inferential statistics and note when you would use each.
Anything correctly identifying 5 statistics, such as r, Z, single-sample t, betweengroup t, within-subject t, ANOVA, chi-square, or any other statistic used to draw
inferences about the population received full credit. No credit for responses
related to mean, median, mode, standard deviation, variance, range, or degrees of
freedom.
OR
b) List and briefly describe 5 major differences between the between-group and withinsubject approaches.
Any 5 reasonable responses from the final page of the t3.doc lecture notes would
be excellent.
OR
c) In a short paragraph, describe how inferential statistics are used in the social
sciences. Make sure to mention the following terms: hypothesis testing, critical value,
p-value, and effect size.
In the social sciences, researchers test their hypotheses or predictions by
studying samples and then using inferential statistics to draw conclusions about
the population at large. Due to sampling error, small, unreliable findings are
expected to occur within any given sample. Thus, obtained results are often
compared to some critical value (e.g. z or t) to determine whether the results were
reliably better than chance. The p value indicates the probability that the result
would be obtained by chance. Measures of effect size indicate the magnitude of
the finding. Thus, statistics enable researchers to judge the reliability and size of
observed effects.
Study Guide
Exam #4
PSY 211
Disclaimer:
 This is a guide of the main points that I consider to be important. You are responsible for
all material covered in the book or in lecture, unless otherwise noted by me.
Format:
 The exam is 120 points. There will be 25 multiple choice questions (4 points each);
multiple choice includes concepts and quick calculations. The remaining 20 points will
come from additional calculations or short answer.
Formulas:
 You will be given a formula sheet (attached) and the t-table from the book.
Not on this Exam:
 Do not use the formulas or d values in the book. Particularly, for the t-test, we used some
simplified version of the formulas in class, which are easier, and still provide a close
estimate of the longer formulas
 There are no story problems based on the one-tail t-test, you just need to know how it
differs from a two-tail test.
Suggested Learning Strategies:
 Study in small groups
 Ask questions or e-mail Mike
 Review last semester’s exam: http://www.psychmike.com/psy211f07/guide4.doc
 Know how to do all homework and study guide problems
 Make up your own multiple choice questions
Conceptual Questions (may appear in multiple choice or short-answer):
 * Be able to interpret Output in order to write the results in APA style (mainly t-tests, but
correlation and regression Output are also fair game)
 * Be able to understand and define the differences between the z-test and the three types
of t-tests. Given a story problem or list of data, be able to correctly identify which test to
use.
 Know the four basic steps of hypothesis testing
 When writing out the null and alternative hypotheses, are we describing the expected
difference in the sample or population?
 Know the difference between t-obtained and t-critical
 Know how probability relates to hypothesis testing
 How does the shape of the t-distribution differ from the shape of the z-distribution?
 How does sample size impact the critical t-value?
 What are degrees of freedom (df)?








Why is it difficult to draw conclusions about group differences based on visual inspection
of the data (such as, histograms)?
What does a significant t-value mean?
What does p = .000 really mean in the SPSS Output?
Why is it possible for a result with a large effect size to be unreliable (not statistically
significant)?
Why is it possible for a small effect size to be reliable (statistically significant) in some
cases?
Know any real-life stats example mentioned in class (I will not repeat these to skippers,
but we can discuss it together as a class)
Know the main differences between the between-group and within-subject t-test
Skepticism
Calculations:
 * Given r, r2, or d, be able to indicate the effect size
 * On the multiple choice, I may give you a story problem and ask you which statistical
test you would use but not make you carry out the calculations. In the problem solving
section, if there are any story problems, I will tell you which formula to use.
o Be able to calculate Cohen’s d, z-values, and various t-values given formulas
 Range of possible values for r, r2, and d
 If an obtained t-value is blacked out from SPSS Output, know how to calculate it from
the remaining variables in the Output
Terms (may appear in multiple choice or short-answer):
 * p < .05
 Hypothesis testing
 Null hypothesis
 Alternative hypothesis
 Effect size
 Sampling error
 Iatrogenic
 Alpha (or significance) level
 Critical region
 Type I error
 Type II error
 Power
 One-tail vs. two-tail t-test (know the difference, but you won’t need to do any story
problems based on one-tail tests)
 Pre-post study
 Manipulation check
 Matched sample
 Carryover effect
Formulas
I will also give you the t-table from the book!
You do not need to print them in advance to bring to the exam.
Beware, that in story problems, I might use “SD” as a symbol for standard deviation,
or “SE” to represent standard error of the mean.
Z = (M – μ)
σM
Single sample
where σM = σ /
t = (M – μ)
sM
n
where sM = s /
n
(M 1  M 2 )
Between-group t =
s
s12
where sM =
M
Cohen’s d for between-group t-test =
n1

s22
n2
(M 1  M 2 )
s
To find the value of s, average the standard deviation from each of the two subsamples, if
available. If only the s for the combined sample is available, use that.
Within-subject
t = MD / sMD
where
sMD = ( s /
Cohen’s d for within-subject t-test d = MD / s
n)
Old Exam Questions
I have taken these questions from last semester’s exam. I think they will give you a good idea of
the types of questions to expect. You can check answers here:
http://www.psychmike.com/psy211f07/guide4.doc
3
When writing the alternative hypothesis, researchers are describing differences
that are expected to occur at this level
a) sample
b) construct
c) population
4
If SPSS shows that a correlation has a p-value of 0.12, the correlation is said to
be
a) statistically non-significant
b) statistically significant
c) not enough information
5
If the alternative hypothesis is really correct, and based on sample data you
reject the null hypothesis, you have made a
a) Type I error
b) Type II error
c) correct decision
6
A hypothesis that men and women do not differ on intelligence
a) Null hypothesis
b) Alternative Hypothesis
c) One-tail hypothesis
8
How will choosing a more strict alpha level impact the critical t-value?
a) Increase it
b) Decrease it
c) No impact
9
If Cohen’s d = - 4.39, the result will be significant
a) Always
b) Sometimes
c) Never
10
Type of t-test commonly used in twin studies
a) single-sample
b) between-group
c) repeated-measures
12
If the null hypothesis is really true, and your sample supports the alternative
hypothesis, you have made a
a) Type I error
b) Type II error
c) correct decision
13
Effect size when Cohen’s d = .78
a) small
b) medium
c) large
14
As sample size increases, the z value needed for significance will
a) increase
b) stay the same
c) decrease
17
Effect size when r2 = 0.26
a) small
b) medium
c) large
18
Which of the following distributions is the flattest?
a) z distribution
b) t distribution where n = 5
c) t distribution where n = 10
19
The p-value shown in SPSS indicates the probability that
a) the null hypothesis is true
b) the alternative hypothesis is true
c) the probability of a Type II error
Below is the Output from a recent homework problem, in which you were to run two
t-tests, comparing gender differences in both “beauty concerns” and “war support,”
but some of the information has been removed. Use these charts for #21-22.
Group Sta tisti cs
2. Gender
44. Beauty Concerns Female
Male
70. War Support
Female
Male
N
Mean
4.462
3.990
3.058
3.175
223
103
223
103
St d. Deviat ion
1.4603
1.5178
1.8432
1.8652
St d. Error
Mean
.0978
.1496
.1234
.1838
Independent Samples Test
Levene's Test for
Equality of Variances
F
44. Beauty Concerns
70. War Support
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
.068
.142
Sig.
.794
.707
t-test for Equality of Means
t
df
Sig. (2-tailed)
Mean
Difference
Std. Error
Difference
95% Confidence
Interval of the
Difference
Lower
Upper
2.677
324
.008
.4716
.1762
.1250
.8182
2.639
191.759
.009
.4716
.1787
.1191
.8240
-.528
324
.598
-.1165
.2204
-.5501
.3172
-.526
196.410
.599
-.1165
.2214
-.5531
.3201
21
What is the value for the degrees of freedom?
a) greater than 325
b) 325
c) less than 325
22
What is the t-value for the “beauty concerns” t-test?
a) greater than 2.37
b) 2.37
c) less than 2.37
Below is some new Output, from our classroom data file, but not from any particular
homework assignment. Instead of looking at gender differences, these t-tests examine
differences across people who described themselves as an athlete (“Yes”) or not an
athlete (“No”). I compared non-Athletes and Athletes along five variables: leadership,
beauty concerns, physical complaints, authority problems, and life stress. These
variables are very straightforward (no tricks), but I can answer questions if you want any
details. Use these charts for #23-25 (next page).
Group Statistics
34. Leaders hip
44. Beauty Concerns
61. Physical Complaints
62. Authority Problems
73. Life Stress
4. Athlete
No
Yes
No
Yes
No
Yes
No
Yes
No
Yes
N
Mean
4.622
5.034
4.284
4.337
3.865
3.247
1.750
1.770
4.797
4.573
148
178
148
178
148
178
148
178
148
178
Std. Deviation
1.3722
1.1832
1.5651
1.4336
1.8393
1.7994
1.2282
1.3096
1.4379
1.4449
Std. Error
Mean
.1128
.0887
.1287
.1075
.1512
.1349
.1010
.0982
.1182
.1083
Independent Samples Test
Levene's Test for
Equality of Variances
F
34. Leaders hip
44. Beauty
Concerns
61. Physical
Complaints
62. Authority
Problems
73. Life Stress
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
Equal variances
as sumed
Equal variances
not ass umed
7.359
2.226
.043
.011
.147
Sig.
.007
.137
.836
.915
.701
t-test for Equality of Means
t
df
Sig.
(2-tailed)
Mean
Difference
Std. Error
Difference
95% Confidence Interval
of the Difference
Lower
Upper
-2.911
324
.004
-.4121
.1416
-.6906
-.1336
-2.872
292.184
.004
-.4121
.1435
-.6945
-.1297
-.321
324
.749
-.0533
.1663
-.3804
.2738
-.318
301.699
.751
-.0533
.1676
-.3832
.2766
3.055
324
.002
.6177
.2022
.2199
1.0155
3.049
310.666
.002
.6177
.2026
.2190
1.0163
-.139
324
.890
-.0197
.1416
-.2983
.2590
-.140
319.306
.889
-.0197
.1408
-.2967
.2574
1.398
324
.163
.2243
.1604
-.0913
.5398
1.399
313.774
.163
.2243
.1603
-.0912
.5397
23
How many of the five results shown are statistically significant?
a) 1
b) 2
c) 3
24
Based on these findings, if you met a random athlete, how might you expect
them to differ from a non-athlete?
a) More life stress
b) More authority problems
c) Fewer physical complaints
25
What percentage of the sample self-reported that they were an athlete?
a) less than 52.3%
b) 52.3%
c) more than 52.3%
26
Calculate a z-score, given the following information,
and indicate whether it is significant.
Sample mean = 200
Sample mode = 210
Sample standard deviation = 160
Population mean = 250
Population mode = 248
Population standard deviation = 111
Sample size = 16
z=
Significant? yes/no
Ignore numbers not in bold above.
28
Psychologists have been fascinated by lottery
winners. Using standard 9-point happiness rating
scales, they have tracked how their level of
happiness changes after winning. In fact, 1 year
after winning the lottery, most winners (n = 9) note
that they are about 0.75 points happier (SD = 2.5).
Calculate the within-subject t statistic, Cohen’s d,
and note whether the result is statistically
significant.
t=
d=
Significant? yes/no