QUESTION 1
Which of the following is false regarding the sampling distribution in null hypothesis testing?
a. Multiple theoretical samples of the same size as your sample are drawn from the population.
b. p calculated is obtained using your sample's distribution as a reference point.
c. Your sample will be one of the theoretical samples drawn.
d. The sampling distribution is a distribution of statistics rather than scores.
QUESTION 2
If you commit a Type II error, you are also doing what? (Select the best answer)
a. You are finding something that is not there in reality.
b. You are not finding something when in reality something is there.
c. You are rejecting the null.
d. You are accepting the null.
QUESTION 3
For a statistically significant result for a t-test, we want:
a. pcalc to be less than pcrit
b. pcalc to be greater than pcrit
c. tcalc to be less than tcrit (in absolute value)
d. none of the above
QUESTION 4
Use the following scenario to answer the next 3 questions.
Because you think highly of the University of North Texas, you are considering a study in which you would examine if UNT graduate students are smarter (in terms of IQ) than people in general.
Which type of statistical test would be most appropriate to answer your research question?
a. dependent samples t-test
b. independent samples t-test
c. single sample t-test
d. Pearson r correlation
QUESTION 5
What sample size would be best for your study?
a. 10
b. 20
c. 30
d. 40
e. 50
QUESTION 6
Your null hypothesis (H0) would be tested with a:
a. one-tailed test
b. two-tailed test
c. none of the above
QUESTION 7
What sampling distribution is appropriate for an independent t-test?
a. sampling distribution of the mean
b. sampling distribution of the difference between means
c. sampling distribution of the standard error
d. sampling distribution of the standard error of the mean
e. none of the above
QUESTION 8
As sample size decreases, the sampling distribution of the mean:
a. decreases in variability
b. increases in variability
c. comes closer to resembling a normal distribution
d. develops a lower range
QUESTION 9
Based on a historical pre-election poll of likely U.S. voters, 47% of those polled said that they intended to vote for George W. Bush and 41% said they intended to vote for John Kerry. There was 95% margin of error of +/- 3% points for each estimate. Based on your knowledge of confidence intervals, you can determine that:
a. John Kerry has no chance of winning the election.
b. The election is too close to call.
c. The probable winner will be John Kerry.
d. The probable winner will be George Bush.
QUESTION 10
Use the following scenario to answer the next 4 questions.
You conduct a study to examine if teacher education majors at UNT score above the expected 3.0 GPA. You collect GPA data on 25 undergraduate teacher education majors and find that their average GPA is 3.25 with a standard deviation of .2. You then conduct the appropriate t-test.
What is the tcalc for this example?
a. 3.25
b. 6.25
c. .04
d. .25
e. none of the above
QUESTION 11
What is the tcrit for this example (assume alpha = .01)?
a. 1.711
b. 2.485
c. 2.492
d. 2.788
e. 2.797
QUESTION 12
Is your result statistically significant?
a. Yes
b. No
c. I wish I could tell, but I just haven't figured this statistical signficance thing out yet.
QUESTION 13
What is the 99% confidence interval around the sample mean for this example?
a. 3.15 to 3.35
b. 3.17 to 3.33
c. 3.18 to 3.32
d. none of the above
QUESTION 14
Use the following scenario and data to answer the next 5 questions. Assume you are a researcher interested in distance learning for statistics courses. You want to examine whether the scores for two different types of online statistics courses are different from a traditional face-to-face course. You are able to randomly assign masters students to one of three conditions: (a) traditional course, (b) online course based only on the textbook and its practice problems for instruction, and (c) online course with enhanced instruction from webpages, powerpoint lessons, and discussions topics. For the purposes of this scenario, assume that you have controlled for other important variables that might result in group differences, such as content variation. All three groups get the same statistics exam at the end of the class. Below are the final scores for these three groups. Round all calculations, if needed, to two decimal places at each step. ******** What is the appropriate statistical test to conduct for this research situation and data?
a. single sample t-test
b. independent sample t-test
c. dependent sample t-test
d. Z test
e. ANOVA
QUESTION 15
Is there a statistically significant difference between the groups (assume alpha=.05)?
a. Yes
b. No
c. I can't tell.
d. Maybe, but you'll have to ask someone who knows what statistical significance means!
QUESTION 16
How much variance in the statistic scores is explainable by what type of instruction the students received?
a. 90%
b. 7%
c. 36%
d. 5%
e. None of the above
QUESTION 17
Which of the following best describes the results of this study?
a. The traditional group performed the worst on the dependent variable.
b. The online-enhanced group outperformed the other two, but the difference was not statistically significant.
c. Both online groups were approximately equal in their outcomes.
d. There was no difference between the means of the three groups.
QUESTION 18
What is the primary problem with this research study?
a. low sample size
b. lack of distinction between the instructional groups
c. failure to randomly assign students to groups
d. the alpha level is too low
e. None of the above is a problem
QUESTION 19
Which of the following differences between two means is least likely to be statistically significant (assuming the variances of all groups are equal and each difference is tested using the same alpha)?
(1) M1 - M2 = 2.47 and n = 10 for both groups.
(2) M1 - M2 = 2.47 and n = 20 for both groups.
(3) M1 - M2 = 2.47 and n = 30 for both groups.
(4) M1 - M2 = 2.47 and n = 40 for both groups.
a. (1)
b. (2)
c. (3)
d. (4)
e. not enough information to determine
QUESTION 20
Which of the following studies would likely yield the most statistically precise single sample t-test results?
a. Study 1: n=100 s=25
b. Study 2: n=150 s=15
c. Study 3: n=150 s=25
d. None of the above are likely to be more precise than the rest.
QUESTION 21
In a single sample t-test, if the standard error of the mean becomes larger, the risk of committing Type I error:
a. increases
b. decreases
c. is always 95%
d. there is no risk of Type I error
e. not enough information to determine
QUESTION 22
Use the following scenario to answer the next 2 questions.
After getting back grades for the final, Candace discovered that she earned a raw score of 92. Candace also knows that the class mean was 82 and standard deviation of 9. Because Candace loves this stuff and just can't get enough, she decides to determine the probability of obtaining a score at least as high as hers.
Assuming the class scores are normally distributed, her probability is:
a. .3665
b. .8665
c. .1335
d. 1.111
e. none of the above
QUESTION 23
If Candace also wants to know if her 4-person study group's mean of 89 was statistically significantly larger than the 40-person class mean of 82 (assume that the SD of her study group was 7), and she was willing to risk a Type I error in her decision 5 times out of 100, then which of the following would be her conclusion?
a. Her group did better than the class average and results were statistically significant.
b. Her group did better than the class average and results were not statistically significant.
c. Her group did not do better than the class average and results were statistically significant.
d. Her group did not do better than the class average and results were not statistically significant.
e. not enough information to say
QUESTION 24
Use the following scenario to answer the next 3 questions.
Assume you conducted a study to compare whether males scored lower than females on a test of extroversion. (Your theory is that females tend to be more extroverted than males.) You conduct an appropriate statistical significance test with a total sample size of 28 and equal group sizes. Assume an alpha level of .01.
What is the appropriate analysis for this example?
a. dependent samples t-test
b. independent samples t-test
c. single sample t-test
d. Z-test
e. none of the above
QUESTION 25
Assuming tcalc was 2.437, what was your decision concerning the null hypothesis and statistical significance in this case?
a. Fail to reject the null because your results WERE statistically significant.
b. Fail to reject the null because your results WERE NOT statistically significant.
c. Reject the null because your results WERE statistically significant.
d. Reject the null because your results WERE NOT statistically significant.
e. Not enough information to determine
QUESTION 26
As a kinesiology professor, you conduct a study to examine the body mass index (BMI) of 4 groups of young adults after having gone through 3 different types of exercise training versus a control group (n=30 in each group). The means of these 4 groups were 23, 23, 23, and 23. The standard deviations around these means were not all the same, but they were all above 0. Given this, which of the following must be true?
a. The ANOVA will be statistically significant.
b. The post hoc tests will show that exercise benefits BMI.
c. The sum of squares between will be negative.
d. The eta-squared effect size will be 0.
e. None of the above is true.
QUESTION 27
Post hoc tests differ from a series of independent samples t-test because post hocs:
a. examine different data situations than t-tests.
b. build in corrections for inflation of family-wise Type I error.
c. analyze relationships between variables rather than differences between means.
d. reduce the risk of Type II error.
e. None of the above.
QUESTION 28
The purpose of a one-way ANOVA is to compare:
a. variances
b. means
c. standard deviations
d. standard errors
e. None of the above
QUESTION 29
Professor Brubaker conducted a study regarding a possible relationship between GPA and participation in college athletics. His study included 40 college students and he generalized to all possible athletes in NCAA Division I schools. Dr. McCall later attempted to replicate Professor Brubaker's study using 65 college students (and also hoped to generalize to all possible athletes in NCAA Division I schools). Which of the following is true regarding these two studies?
a. The standard error in both studies is the same.
b. The standard error in Professor Brubaker's study is lower than the standard error in Dr. McCall's study.
c. Dr. McCall's study will yield larger mean differences than Professor Brubaker's study.
d. None of the above are true.
QUESTION 30
I am more informed of basic statistical practices now than when I began this class, and I furthermore will forever be grateful for the opportunity to study statistics - an experience which will no doubt revolutionize my life.
a. True
b. False
c. You've got to be kidding.
d. Ask me after the grades come out.
e. I still think this is a trick question. Or maybe it's not! Maybe the question is just designed to make me think it is a trick question when really it is not a trick question! Ahhhhh!