PSYC6060: Stats Exam
A distribution has µ=100 and SD=30. One researcher samples 100 individuals and reports M=70, with the 95% confidence interval [100, 40]. What is a likely confidence interval that researcher 2 can report? a) M=130, 95% CI [100, 160] b) M=110, 95% CI [80, 140] c) M= 60, 95% CI [90, 30] d) M=100, 95% CI [130, 70]
d
A graduate student researcher is looking to examine whether daydreaming increases student productivity. SONA has only given the researcher 50 participants for this semester. The study employs random sampling - all participants are undergraduate students from the University of Guelph. What design (paired vs. Independent samples t-test) would be the most appropriate given the information provided? a. Independent because there is a small sample size b. Independent because there is random sampling c. Paired because there is random sampling d. Paired because there is a small sample size
d
A paired design with __________ can often justify a causal conclusion. a. Random assignment to conditions b. Small confidence intervals c. A large sample size d. Random assignment to counterbalanced orders of testing
d
What is true about the central limit theorem? a) It means that the mean of a number of independent variables has approximately a normal distribution b) The larger N is, the closer the distribution is to normal c) The central limit theorem means that the sampling distribution of sample means is often close to normally distributed, even for non-normal populations d) All of the above e) Only a and b
d
When calculating a CI for the difference between two group means we usually assume that: A. both groups represent random samples. B. in each group the scores being compared come from normally distributed populations. C. the population variance is the same for both groups. D. All of the above.
d
The regression line is chosen to... a. minimize estimation error b. minimize the standard deviation of the residuals c. minimize the slope d. Both A and B
d
Using the Overlap Rule for CIs on independent means, which graph shows the smallest amount of evidence of a population difference? A) Graph A B) Graph B C) Graph C D) Graph D
d
What is the main reason that a researcher should make their data and the details of the data analysis openly available? a) The researcher doesn't have to worry about saving it on their own computer and potentially losing it b) The researcher can get their article published in an open access journal c) Other researchers won't "scoop" their ideas d) Other researchers can conduct additional analysis as a check to see if they get the same results
d
What is the relationship between alpha and beta? a) alpha and beta are always equivalent b) alpha and beta are never equivalent c) the larger alpha is, the larger beta is d) the larger alpha is, the smaller beta is e) a and c
d
What 3 components are needed to calculate t for a single sample? A) Variability component, Effect Size, Sample Size B) p-value component, Effect Size, Sample Size C) z-score, variability component, p-value D) Variability component, Sample Size, z-score
a
What are the 4 aspects to think about when looking at a distribution of data? a) Measure, Location, Spread, Shape b) Measure, Location, Baseline, Midpoint c) Baseline, Midpoint, Shape, Spread d) Midpoint, Location, Shape, Spread
a
1. You are conducting a study that analyses differences in extraversion across sexes: male, female, and intersex. Which design will best suit your study? a) A one-way independent groups designs b) Multiple regression c) An independent groups t-test d) A complete randomized factorial design
a
When conducting a meta-analysis, what does the diamond in a forest plot represent? a) CI of the meta-analysis results b) Margin of error c) Point estimate d) Overall mean
a
The point estimate is the same as the sample mean. a) True b) False
b
2. Cummings presents the following graph depicting mean scores on an LSAT test, before and after taking a critical thinking course. How can the overlap of the CI's be interpreted given the design? a. The overlap of the confidence intervals cannot be meaningfully interpreted b. A large overlap between the two pre/post scores indicates that there was not a significant difference in performance on the LSAT after taking the critical thinking course c. The width of the confidence interval indicates a small sample size d. A large overlap between the two pre/post scores indicates there was a significant difference in performance on the LSAT after taking the critical thinking course
a
A poll of graduate students indicated that 75% supported learning R, with a 5% margin of error. Dr. Stanley wants to redo the poll with a new sample of students. He will not be satisfied with the results unless the poll shows that support is higher than 80%. He instructs you to repeat the poll 1000 more times using the same procedure but a different sample each time. Based on the results of the first poll, approximately how many polls will find support greater than 80%? a. 25 b. 50 c. 100 d. We cannot make this estimation. The margin of error is too large to make accurate predictions.
a
A researcher was interested in investigating the relationship between weight and personality traits. Previous research has indicated there was a relationship between weight and agreeableness, and the researcher wanted to see if this was also the case for a relationship between weight and extraversion. Using R, the researcher assessed the relationship between weight (in lbs) and extraversion scores and looked at the correlation table. To their surprise, the researcher found the correlation to be non-significant, and concluded there was no relationship between weight and extraversion. What did the researcher fail to do, which may have led to an incorrect conclusion? a. The researcher did not examine a scatterplot to assess the possibility of a non-linear relationship b. The researcher forgot to convert the data to the same format (you cannot correlate scores on a scale to weight in points) c. The researcher failed to consider the difference between correlation and causation d. The researcher did not remove outliers
a
Considering a Two-Way Independent Groups Design, with two independent variables (A and B) with two levels each, it is correct to say that the main effect of Variable "A" is: a) the overall effect of A on the DV, averaged over the two levels of B b) its effect at the main level of independent variable B. c) the difference between the effect of A at the two levels of B. d) the mean of its main level
a
How long is the margin of error, and what does it represent? A) Half the length of the CI, and the likely greatest error in the point estimate B) Half the length of the CI, and the expected error in the point estimate C) The length of the CI, and the expected error in the point estimate D) The length of the CI, and the likely greatest error in the point estimate
a
If r = 0, we use MY as the estimate for any X. a. True b. False
a
In a Two-Way Independent Groups Design, the homogeneity of variance assumption is probably reasonable if a) considering the group SDs, the largest SD is no more than about twice the smallest. b) considering all pairs of groups, all the correlations are heterogeneous. c) the group means are all related in a linear fashion. d) the DV has interval scaling.
a
Psychology, and most related human disciplines do not include this type of scale because: A. Ratio, since we often lack a true zero in the variables we measure B. Nominal, since we try to include at least one continuous variable in each study C. Interval, since we often use Likert scales which are ordinal scales D. Ordinal, since we don't typically rank order participant scores on a given variable
a
Sample means tend to be normally distributed around the true population mean. The standard deviation of the distribution of sample means is: a. the SE. b. z. c. X. d. the MoE
a
Samuel designed a study to measure whether people learned more about statistics in an online course, an in-person course, or a self-directed course (just the textbook). He is using a standardized test of university-level statistics to measure knowledge of statistics at the end of the course. He is wondering whether he should use an additional measure of self-reported level of understanding of statistics in his study. What would you tell him to do? a) Include the additional measure to provide converging evidence and draw more confident conclusions. b) Do not include the extra measure because it will complicate his study. c) Include the additional measure but only report it if he finds the same effect as he did with the standard test.
a
The SE gets smaller as ____ increases or as ____ decreases. a) N, σ; b) MoE, N c) SD, MoE d) None of the above
a
The statistic r only measures the ______ between two variables. a. Linearity b. Causality c. Distance d. Positive/negative relation
a
The value on the x axis that is strongly influenced by very low and very high points is also: a) the balance point of the data set b) always the value below and above which half of the data points lie c) is a measure of location that requires only ordinal scaling d) all of the above e) B and C
a
There are two requirements of random sampling, they are: A. Every member of population has equal chance of being chosen; all members of sample are chosen independently B. Every member of population has equal chance of being chosen; members of sample are chosen by snowballing C. Every member of population has equal chance of being chosen; members of sample are chosen based on representation in total population D. Every member of population is sampled; members of sample are volunteers
a
To calculate the p value in a single-sample study: a) use the z value when 𝜎 is known b) use the t value when 𝜎 is known c) always use the z value d) always use the t value
a
To compare two independent groups, the estimation approach focuses on the CI for the mean difference between groups. Considering NHST to compare two independent groups: a. NHST uses the same basic information as the CI: an effect size, a variability component, and a sample size component. b. NHST typically tests the null hypothesis that the difference between the groups is zero (H 0 : μ diff = 0). c. Unlike the estimation approach, the NHST approach yields a p value which is then used to make a yes/ no decision about the null hypothesis. d. All of the above.
a
Which of the following best describes the process of the regression to the mean? a) A hockey club has many teams. Each week they give a special mention to the player scoring the most points that week. The club noticed that only rarely does a player receive a special mention 2 weeks in a row. b) A restaurant notices that every Friday, they experience an increase in sales. c) An advertisement encourages people to enroll for further training by declaring that every extra year of education increases annual income by an average of $7,000. d) You have developed a computer game to help teach spelling. A group of children who scored more than one standard deviation below their age norm on a spelling test played your game for two sessions. They took the test again a week later and each obtained an even higher score. e) None of the above.
a
Which of the following situations gives the lowest Type II error rate? a. Large sample size, large alpha, and a within-participant design. b. Small sample size, small alpha, and a within-participant design. c. Large sample size, large alpha, and a between-participant design. d. Large sample size, large alpha, and a between-participant design.
a
The use of confidence intervals for estimating population parameters from sample statistics depends on what assumptions (choose 2). a) Normal distribution b) Random sampling c) Homogeneity of variance d) Large N
a + b
A researcher compares that statistics marks of students at Guelph and Waterloo and finds a non-significant difference (p = .99). What can the researcher conclude? a. There is no difference between the statistics marks of students at Guelph and Waterloo b. We failed to find evidence of a difference between the statistics marks of students at Guelph and Waterloo c. There is a small difference between the statistics marks of students at Guelph and Waterloo d. There is a large difference between the statistics marks of students at Guelph and Waterloo
b
A researcher examined the "good mood effect" by measuring the mood of shoppers in a mall after giving them a chocolate bar and found a significant relationship between giving a chocolate bar and positive mood. Jane wants to re-examine the "good mood effect" by complimenting people in her class before measuring their mood. Which type of replication is Jane proposing? A. Direct replication B. Conceptual replication C. Close replication D. This is not a replication
b
A researcher gathers a sample and discovers the range of age is 8-30, this is an example of a(n): A. Inferential statistic B. Descriptive statistic C. Parameter statistic D. None of the above
b
To find the median, order the data, then count halfway through the data points. If N is odd, average the two middle values. a) True b) False
b
What is a "pooled" standard deviation? a) The arithmetic mean of two group standard deviations b) A best estimate of the population standard deviation c) It is used when group variability is less than would be expected due to sampling error d) Used to calculate effect sizes, but only in repeated measures designs e) Used when groups are heteroscedastic
b
A researcher is examining the impact of taking critical thinking course on the LSAT scores of participants. Using a within-subjects design, the researcher tested each student on the LSAT. Then, the student completed a critical thinking course. Finally, the researcher re-tested the students on the LSAT. The researcher examines the confidence intervals on the pre-course LSAT score mean, the post-course LSAT score mean, and the mean difference between the two. They found the confidence interval for the mean difference to be smaller than the other two confidence intervals. Why is this the case? a. The sample size is large enough to detect a small effect b. Each participant is their own control, thus there is a limited difference between their pre/post scores c. The critical thinking course had no effect on LSAT scores d. The control group also improved in LSAT scores, despite not taking the critical thinking course
b
A researcher runs an experiment to determine if students who drink any amount of coffee are happier than students who never drink coffee. The difference between the two groups generated a p value of p = .32. This means: a) The null hypothesis is accepted b) The null hypothesis is not rejected c) There is a difference between the happiness of people who do and don't drink coffee d) There is no difference between the happiness of people who do and don't drink coffee
b
A top-tier law firm is recruiting law students for their summer internship program. The program is quite competitive, and the firm only recruits students who are in the top 20% of their class. When the law firm observes the correlation between school GPA and LSAT scores, they find that the correlation between the variables - specifically among the top students - is only r = .13. Why could this be? a. There is a curvilinear relation between school GPA and LSAT scores. b. The restriction on one variable causes an underestimate of the correlation in the whole population c. Because it is a standardized test, the LSAT cannot be correlated with school GPA d. There is a ceiling effect observed with school GPA
b
Cohen's widely-used values for effect size (d) are the following for small, medium, and large: A. 0.1, 0.3, 0.5 B. 0.2, 0.5, 0.8 C. 0.2, 0.4, 0.6 D. 0.1, 0.5, 0.8
b
Craig is using regression to use GRE scores to predict graduate school success. Specifically, he wants to describe the typical level of success expected for students who score 160 on the GRE. Which interval should he use? a) CI on b b) CI for mean of Y at X c) Prediction interval for individual values of Y at a specific X d) None of the above
b
Dr. Stanley David was interested in investigating the relationship between the percentage of students in a psychology statistics class that have downloaded his R package "APA Coffee Tables" and average grade amongst students in those classes. He has evidence to suspect that the relationship between these variables is r = .5. Dr. David wants to ensure a CI of approximately [.4, .6]. What would be an appropriate N (the # of psychology statistics classes examined) to investigate this relationship? a. N= 40 for each cell b. N = 160 c. N = 320 d. n = 260 for each condition
b
Gabriela and Sylvia are working as a team for their university's residential life program. They are both tasked with surveying students about their satisfaction with the dormitories. Today, Gabriela has managed to survey 25 students; Sylvia has managed to survey 36 students. The satisfaction scale they are using has a range from 1 to 20 and is known from previous surveys to have σ = 5. Which sample will have the smaller SE: the one collected by Gabriela or the one collected by Sylvia? a) The one collected by Gabriela, because it has lower N b) The one collected by Sylvia, because it has higher N c) They will be the same, since we know the population standard deviation d) We cannot tell based on the information given.
b
If we already know the population mean (µ), is it still meaningful to calculate a confidence interval? a. Yes; the confidence interval tells us what range of population values could have produced our sample. b. No; knowing µ renders a confidence interval meaningless. c. Yes; a confidence interval gives an indirect indication of the effect size. d. No; a confidence interval gives a meaningless indication of the effect size.
b
In a 2 × 2 Independent Groups design, the interaction is: a. the difference between the main effects of the two independent variables. b. the extent to which the effect of independent variable A on the dependent variable is different at the two levels of independent variable B. c. the average of the two main effects. d. none of the above
b
On an exam, Sam's teacher tells him that his standardizes score was z = 3. Should Sam be happy? a) Maybe - this means he did about average. b) Yes, this means he did much better than the average. c) No, this means he did much worse than average. d) Maybe - this means he did about average. e) None of the above.
b
On average, a confidence interval has a(n) __ % chance of capturing the mean of a replication experiment. a. 50 b. 83 c. 95 d. The confidence interval only tells us about the population mean
b
One of the main assumptions of calculating a confidence interval (CI) on r is that the data is drawn from a random sample, and the population R follows a normal distribution. Usually, accepting this assumption is appropriate for the practical applications of psychological research. However, there are three main scenarios when this is not an appropriate assumption. In which of the following scenarios would it be appropriate to assume random sampling and a normal distribution, and proceed with calculating a CI on r? a. The sample data is not representative of a larger population b. Investigation of the scatterplot indicates a linear relationship between the variables c. Prior research indicates a non-linear relationship between the variables d. The scatterplot indicates a floor effect
b
Read the sentences below and choose the correct alternative; I - The data for each group are a random sample from a population that is normally distributed II - The samples are independent III - Variance is homogeneous across groups. Which sentences include basic assumptions for the Two-Way Independent Groups Design a) I b) I, II and III c) II and III d) I and III
b
The dance of the r values is related to what concept? a. Researchers failing to report the correct r-values b. Variability in sample r-values due to random sampling c. R-values in meta-analyses are often "dancing around" and indicate bad research practices d. Variability in population r-values due to random sampling
b
The focus of correlational data analysis is on spread within the means, not the samples. A) True B) False
b
What is a good rule of thumb for the number of effects that should be specified for planned analysis within one-way and factorial designs? a) No more than 3 b) No more than the study's degrees of freedom c) No more than the levels of the independent variable(s) d) Either a or b, whichever is larger in the given study
b
What is necessary to know whether a confidence interval contains the population mean (µ)? a. Inferring µ and the population SD from the sample. b. Knowing the value of µ. c. Knowing the value of µ and the population SD. d. A very small p value (p < .001).
b
What should you do to make a "good" graph (i.e., one that is not misleading)? a) Use pictographs that expand in size to show changes in amount b) Use markings of equal intervals throughout your scale c) Limit your scale to the range of numbers in which your scores fall between d) Both a & b
b
When analysing your results, you notice an interesting pattern in the data that indicates a group difference which you didn't expect. What conclusions can you draw about this difference? a) Conclusions that will be the primary focus of your Results section b) Tentative, speculative conclusions c) Strong conclusions, provided you include a 95% CI d) Conclusions that support your research question e) None of the above
b
When are long CIs NOT necessarily indicative of a lack of precise measured effect? a) When they reflective of independent means b) When they are within the context of a repeated measure c) When there are more than 3 dependent variables d) When there are more than 3 independent variables
b
When choosing from nominal, ordinal, interval, and ratio scales, people frequently get ordinal and interval scales confused. What question can you ask yourself to differentiate between these two types of scales? A. Do you have a scale that shows DISCRETE CATEGORIES? B. Do you have EQUAL DISTANCES between scale points? C. Do you have a ZERO in the scale? D. Do you have RANKINGS that place some scale points above and below others?
b
When considering a confidence interval, where does the most plausible estimate of the true population value lie? a) Each value within the CI is equally likely b) Near the centre c) At the tail ends d) None of the above
b
Which of the following describes the distribution shown below ............ooOOOoo a) Positive skew and Unimodal distribution b) Negative skew and Unimodal distribution c) Negative skew and Bimodal distribution d) Symmetric Distribution e) C and D
b
Why do repeated measure designs give relatively precise estimates of contrasts? A) There is no variability between the participants B) Variability is reduced within participants C) Correlations exist in the results D) Because there is only one group, the sample size is higher
b
With fixed N and r = .0, the CI will be ____; but with fixed N and r = .90 the CI will be _____ A) Shorter/asymmetrical; long/symmetrical B) Long; short C) Long/asymmetrical; short/symmetrical D) The CI of r isn't influenced by N (fixed or not fixed) E) Can't determine because CIs in correlation depend on more than just N and r
b
A researcher has begun to develop a study for her undergraduate honors thesis, which she wants to be impactful and interesting. To do so, she makes sure to brush up on all relevant material that she had learned in her previous courses on research methods and statistics. She knows there are many important differences between the many experimental designs. Starting at square one, she comes across hypothesis and hypothesis testing (beginning with the Null and Alternative hypotheses). Using NHST, what will her H0 be when using a paired design? A) That the sample mean difference will be less than zero B) Paired design experiments don't have a null hypothesis because participants are only tested once C) That the population mean difference will be zero D) That the sample mean difference will be zero
c
A researcher has successfully finished his experiment. He is chuffed with himself and is ready to analyze the data. Among other things, he will interpret the d values of his results. Each participant in his study went through both conditions, testing the impact of chewing gum during a simulated driving test. How should he interpret his d values when it comes time to do so? A) The researcher can analyze either the d or dunbiased values, it doesn't matter B) Analyze the data as if the participants did not go through both conditions (better safe than sorry!). C) Among other things, consider how big a practical difference the effect makes. D) He should expect the CIs to be long, thereby indicating uncertainty.
c
A researcher runs an experiment to determine if students who eat any amount of chocolate are happier than students who never eat chocolate. The difference between the two groups generated a p value of p = < .05. This means: a) The null hypothesis is accepted b) There is a 95% chance the effect we observed is accurate c) The null hypothesis value lies outside the 95% CI d) There is a significantly low probability that the null hypothesis is true
c
Before using regression, it is important to see the scatterplot. What are you inspecting for? a. Is the relationship linear? b. Are there outliers? c. Both A and B d. Only A
c
Consider the four panels below, representing different patterns of means of enjoyment of different types of music (Rock and Classical) in two different settings (Party and Church). Which panels suggest the presence of an interactions between the two IVs? a) All four panels b) Panels A and C c) Panels A, B and D d) Panel C
c
Cumming identifies three main reasons Cohen's d may be a useful tool for interpreting effect size. Which of the following is NOT a reason Cohen's may be useful? a. To compare effect sizes from measures that assess a wide range of characteristics b. To understand an original unit's effect size c. To assess the effect of the measure on the population d. To compare effect sizes from several measures that use different units, but measure the same characteristics
c
For what test does the statistic represent t he ratio of effect variability to reference variability? A) T-test B) Correlation C) ANOVA D) Chi-square
c
How does degrees of freedom (DF) relate to bias in a study? A. The larger the DF, the larger the bias B. The smaller the DF, the smaller the bias C. The larger the DF, the smaller the bias D. DF and bias are unrelated
c
If the margin of error is small... a) CI will be long and have high precision b) CI will be long and have moderate precision c) CI will be short and have high precision d) CI will be short and have low precision e) CI will not contain the true population value
c
If two 95% CIs just touch but do not really overlap, p is approx. ____, ________ amount of evidence of a population difference. a. .03, small b. .05, small c. .01, moderate d. < .01, large
c
In an independent groups design, Margin of Error is based on what three components? A) p-value component, effect size component, t component B) variability component, sample size component, p-value component, C) t component , variability component, and sample size component D) t component, effect size component, sample size component
c
Should you use Cohen's reference values for d-values to assign to small, medium, and large effect sizes across disciplines such as psychology, medicine, education, and pharmacology? Why or why not? A. Yes, because Cohen's reference values were compiled from a survey of major articles published in ten disciplines over three decades, so they represent reasonable estimates for effect sizes across disciplines B. Yes, because all these fields deal with effects on the general human population, humans are expected universally to show the same effects to interventions/treatments C. No, because based on literature reviews, psychology, medicine, and education often have lower reference values for effect size than Cohen originally proposed D. No, because Cohen based his effect size estimates off a single journal from a single discipline many decades ago and this is not trustworthy as a benchmark
c
The 99% confidence interval is about ____ longer than the 95% confidence. a. One-fifth b. One-quarter c. One-third d. One-half
c
The University of Guelph's coordinates are 43.5329° N, 80.2262° W. This is an example of what type of measurement? A. Ratio B. Nominal C. Interval D. Ordinal
c
The standard deviation is not: a) A measure that tells us how spread or dispersed data points are b) The square root of the variance. c) Influenced by very low and very high points d) All of the above e) A and B
c
There are two important slogans for Open Science brought up in Chapter 1. Which questions come up in the following situation: A poll is conducted to gauge students' opinions towards their final exam format. The preferences are presented to a group of faculty without the specific questions the students were asked. A) Were the results selected in a way that might mislead? B) Do we have full information? C) A and B D) Neither A or B
c
What is recommended for an IV with more than two levels? a) Only use exploratory post-hoc analyses to examine possible effects b) Make sure to keep each level separate to ensure analysis is not confounded c) Consider combining some levels to reduce degrees of freedom d) Consider eliminating some levels to focus on effects of interest
c
What is the takeaway from the Holzel study regarding use of repeated measures? a) Repeated measures greatly increase the risk of Type 1 error if the correlation of the levels of the repeated measure is high b) Repeated measures greatly increase the risk of Type 1 error if the correlation of the levels of the repeated measure is low c) Repeated measures greatly increase precision if the correlation of the levels of the repeated measure is high d) Repeated measures greatly increase precision if the correlation of the levels of the repeated measure is low
c
What is true of the distribution shown below? OOOoo...o...o a) The mean is less that the median b) The mean and median are the same c) The mean is greater than the median d) None of the above
c
What term most accurately describes the pattern within this data set? a. Curvilinear relation b. Parallel relation c. Ceiling effect d. Floor effect
c
Which of the following situations is NOT an example of a Type I error? a. A study found a drug had an effect, but did not control for the placebo effect. Nonetheless, the researcher concludes the drug had an effect. b. A study found that a cognitive intervention decreased the number of errors made on a reaction time task, despite not controlling for practice effects on the task. Nonetheless, the researcher concluded that the drug had an effect. c. A researcher concluded that a cognitive intervention did not improve mental rotation performance over time and the study did not control for age. d. A researcher thinks p values indicate the probability that results are due to chance, when in reality, this is not the case.
c
Why are the CI lines of a regression line curved towards the ends? a. Because there are more outliers towards the ends. b. Curved lines indicate a weak correlation. c. Due to less data at the extremes. d. They aren't curved; only prediction intervals show curving.
c
Why is it important to limit the number of comparisons you conduct in one-way independent designs? a) There aren't any good reasons to limit your comparisons; the more tests you conduct, the more likely you are to find something significant. b) To keep p-values low c) To limit the Type I error rate d) a and b e) b and c
c
A researcher at the University of Ottawa collected data from 45 participants in his laboratory. The experiment involved having participants take a verbal test (the pre-test). After a short delay, the same participants were given an energy drink then asked to take another verbal test (post-test). The researcher found that the difference between the pre- and post-test was minimal. What can he conclude from this information alone? A) Carry over effects may have been the reason why there was only a slight difference between the pre- and post-test. B) That some additional effect must have interfered, thereby influencing the relationship between verbal scores and energy drink intake. C) Because the difference between pre-test and post-test was small, but because there was a large t-value, the researcher can still reject the null hypothesis. D) Mdiff was small (0.3), therefore the researcher failed to reject the null hypothesis.
d
A researcher examined the effect of compliments on self-esteem. To do so, she measured participants' self-esteem level, subsequently exposed them to a compliment, and then measured self-esteem again. The researcher calculated the d-value for her study, and it was suspiciously large. What error could the researcher have made when calculating the d-value? a. She added an extra 0 to her standardizer b. She incorrectly used sav as the standardizer in the calculation, as opposed to correctly using sdiff c. She forgot to standardize the scores in her calculation d. She incorrectly used sdiff as the standardizer in the calculation, as opposed to correctly using sav
d
A researcher interested in studying the behavior of sea sponges and starfish collected data by observing a total of 220 starfish and 198 sea sponges. When the sea sponges and starfish were around one another they acted uncharacteristically. Therefore, because the sea sponges and starfish acted differently around one another the researcher concluded that it was the sea creatures' friendship that must have caused them to act in an uncharacteristic and immature manner. The correlation (r) obtained was 1.02. Was the researcher correct in their conclusion? A) Yes. While correlation doesn't necessarily imply causation, with such a high correlation obtained, the researcher can be confident in concluding that the friendship between the sea creatures caused the immature behavior. B) No. Although obtaining r = 1.02 is astounding, correlation can never be used to imply causation. C) Maybe. The researcher did not account for other variables that could have influenced the sea creatures' behavior, so the results should be viewed through a tentative lens. D) No. Something must have gone wrong for the researcher and therefore the results cannot (and should not) be used for conclusive purposes.
d
A researcher, Jane, studies the 5K race finish times of runners. She found that runners came in 1st to 168th place. Based on their finish times, she wanted to calculate a mean race finish time for all 5K race runners. Is this possible, why or why not? A. Yes, because you can look at the score sheet to sum up all the finish times and divide this sum by 168. B. Yes, because this is an example of an interval scale, which has equal time distances between rankings so you can calculate a mean. C. No, because this is an example of a ratio scale, which does not have equal time distances between rankings so you cannot calculate a mean. D. No, because this is an example of an ordinal scale, which does not have equal time distances between rankings so you cannot calculate a mean.
d
According to the formula below for one-way independent groups design, what is the null hypothesis? H0: μA1 = μA2 = μA3 = ... a) That all the group population means are the same. b) That the group means are all zero. c) That the main effect of Factor A is zero. d) A and C e) None of the above.
d
All other things held constant, which of the following changes will decrease the size of the confidence interval (ie., the range between its lower and upper limit)? a) Increase in the sample size b) Decrease in the sample standard deviation c) Increase in the sample mean d) Both A and B
d
Alpha is: a) the probability of rejecting the null hypothesis when it's true b) Type I error c) Type II error d) a and b e) a and c
d
An error bar represents: a. The 95% confidence interval b. Standard deviation c. Standard error d. It could represent any of these statistics
d
An ordinal scale can be used to find which descriptive statistics? a) The mean b) The mode c) The median d) Both b & c
d
For a linear regression, the residual is: a) An estimation error b) The difference between the observed and expected dependent variable scores c) Another word for confidence interval d) Only a and b e) Only a and c
d
For a standardized regression line, what does β stand for? a. Standardized regression weight b. Population parameter c. Beta-weight d. Both A and C
d
How can we use degrees of freedom? a. To specify the t distribution. b. To estimate the quality of a study's estimates. c. As an indicator of the number of independent pieces of information. d. All of the above.
d
If you were to adopt "meta-analytic thinking" when developing your thesis proposal, what might you be likely to do? a) Conduct a meta-analysis of an original study and one other replication study. b) Conduct a replication study that uses the same methodology and analysis but a new sample. c) Conduct a study that could be easily replicated because all your data and analysis are openly available d) All of the above
d
Joe measures ice cream sales and murder rates over the course of one year. He discovers that these variables are positively related, as they both increase in the summer and decrease in the winter. He concludes that eating ice cream makes you more murderous. Why is this an incorrect conclusion? A. There is no random assignment of ice cream sales. B. There is no random assignment of murder rates. C. There is no manipulation of ice cream sales. D. A and C E. All of the above
d
Publication bias: A. Refers to journals' preference for publishing novel studies with positive and significant results B. Results in smaller effect sizes than reflected in the true population C. Leads to a lack of published replication studies D. A and C E. All of the above
d
Sound interpretation of a study's conclusions... a) Requires consideration of the context and implications of findings b) Can be done by looking only at statistical results c) Considers both point estimates and interval estimates d) a and c e) b and c
d
When choosing a set of planned contrast, you should make sure that ... a) Your planned contrasts match your research question b) You set a limited number of contrasts in advance c) Examine a large number of contrasts d) A & B
d
When conducting a t-test, the homogeneity of variance is likely to be violated if: A. Both standard deviations differ considerably B. Both sample sizes differ considerably C. Both means differ considerably D. A and B E. A and C
d
When is it appropriate to use post hoc tests? a) Anytime you are using a one-way design b) Only when they are interpreted as exploratory c) As a follow-up after planned analyses are complete d) Both b & c
d
When researchers refer to cherrypicking, they are talking about: A. Presenting all exploratory analyses as planned B. Presenting all planned analyses as exploratory C. Not creating planned analyses, instead only doing exploratory analyses D. Selectively reporting some, but not all, results of analyses
d
Which of the following could be a null hypothesis? a) µ = 0 b) 3 < µ < 7 c) µ = 50 d) A and C
d
Which of the following is a way to reduce the margin of error? a) Decrease the sample size b) Increase the sample size c) Reduce the standard deviation d) B & C
d
Which of the following is true about r2? a. r2 is calculated simply by multiplying r by itself (r × r). b. r2 reflects the strength of the correlation and can vary from 0 to 1. c. r2 represents the proportion of variance of Y accounted for by knowing X. d. All of the above
d
Which of the following statements about p values is true? a. P values correlate with effect size. b. P values indicate the probability of H0. c. P values indicate the probability that values as or more extreme than obtained occurred due to chance. d. Type I errors become less likely as p values fall below increasingly conservative alpha values.
d
Which statements about normal distributions are true? a. The standard normal distribution has a mean of 1 and a SD of 0. b. About 95% of values lie within 1 SDs of the mean. c. A normal distribution can have more than one mode. d. None of the above.
d
Which statements regarding the standard error are false? a. The standard error is the standard deviation of the sampling distribution of the sample mean. b. The standard error is a measure of sampling variability. c. The standard error tells us how far from the population mean the sample means are likely to fall. d. The standard error gets smaller as sigma increases.
d
Why does the sampling distribution of the mean have a normal shape? a. Because the larger N is, the closer the distribution to normal. b. Due to the central limit theorem. c. Because the mean of a number of independent variables has, approximately, a normal distribution, whatever the distributions of those variables. d. All of the above.
d
You are conducting a study to examine how students majoring in different subjects (Psychology, Biology, English) differ in ratings of student satisfaction with their program. You hypothesize that Psychology students will rate their program satisfaction higher than students from both of the other programs. How will you test this hypothesis? a) Using the Tukey post-hoc procedure b) With a planned comparison c) With a subset contrast d) B and c
d
You decide to answer all the questions on your graduate statistics final exam completely at random. There are 100 multiple choice questions, and each have four response options. You use a random number generator on your calculator to decide your answer to each question. What is the highest possible grade you could receive with this method? a. 0% correct b. 25% correct c. 95% correct d. 100% correct
d
A stacked dot plot: a) Can vary in shape depending on bin boundaries b) Gives information about every point in the data set c) Can vary in shape depending on the number of bins d) Shows the overall shape of the data set e) B and D
e
In the textbook, a scenario is presented in which a researcher is examining the impact of taking a critical thinking course on the LSAT scores of participants. Using a within-subjects design, the researcher tested each student on the LSAT. Then, the student completed a critical thinking course. Finally, the researcher re-tested the students on the LSAT. Which method is most appropriate to reduce carry-over effects? a. Counterbalancing b. Two parallel forms c. Control group d. A and C e. B and C
e
The normal distribution a) Is never assumed to be true in a population of interest b) Has tails that extend infinitely in each direction c) Is symmetric around it's midpoint d) A and B e) B and C
e
What is true about the standard error? a) It is the standard deviation of the independent variable. b) It is the standard deviation of the dependent variable. c) It is the standard deviation of the sampling distribution of the sample means. d) To halve the standard error, we would need N four times as large e) Both c and d
e
