Final Exam

Question 1

A psychologist is studying the effect of eating ice cream on memory performance. She assigns participants into groups that eat 1, 2, or 3 scoops one hour before they all take a memory test. What term best describes the role of “scoops of ice cream” in this study?

Answer 1

independent variable

Question 2

Sum of Squares (SS)

Answer 2

Σ(X –M)^2

Question 3

Variance (SD^2)

Answer 3

SD^2 = SS/N

Question 4

Standard Deviation (SD)

Answer 4

square root of variance

Question 5

To be scientific, a theory must be testable. This means the theory must be

Answer 5

falsifiable

Question 6

A researcher expects that baboon mothers are generally more protective of their female offspring than their male offspring. When the researcher observes baboons in the wild, he focuses more of his attention on the mother-with-female offspring interactions, and takes more detailed notes, than when witnessing mother-with-male offspring interactions. Follow-up analysis demonstrated that the experimenter’s increased attention to mother-with-female offspring interactions actually caused the mothers to behave in a more protective way. Which of the following problems is evident?

Answer 6

Experimenter reactivity, experimenter bias

Question 7

Where does a Cohen’s d effect size of -1.39 fall in terms of Jacob Cohen’s suggested conventions?

Answer 7

Large

Question 8

What’s the point of Geoff Cumming’s slogan about a confidence interval, “It might be red!”?

Answer 8

It might not capture the true population mean from which the sample was drawn

Question 9

In psychological research, a fact is something that can be precisely and objectively measured, like the diameter of the circles drawn in a sociogram. A ____ is the term for an inferred state/characteristic, something that can’t be directly measured, such as intelligence, anxiety, implicit self-inflation

Answer 9

construct

Question 10

Which of the following is necessary to establish causation?

Answer 10

Alternative explanations have been ruled out, Covariation of the events, A time-order relationship

Question 11

A particular standardized test has a population mean score of 75 and standard deviation of 7.5, and the scores are normally distributed. If a person’s score places them at the 5th percentile on this test (from the bottom), what is this person’s score?

Answer 11

z*σ + x

Score = (-1.64*7.5) + 75 = -12.3 + 75 = 62.7

Question 12

The median is typically a better measure of central tendency than the mean in which of the following cases?

Answer 12

the distribution is highly skewed

Question 13

Which of these describes the shape of a frequency distribution that has clearly been affected by a ceiling effect?

Answer 13

negatively skewed, skewed left

Question 14

The results of a study measuring reaction time in response to a stimulus show a mean of 3.50 seconds and a standard deviation of 1.30 seconds. This allows us to calculate that a participant with a Z-score of +1.5 took ___ seconds to respond to the stimulus

Answer 14

score = z*σ + x

1.5(1.3) + 3.5 = 5.45

Question 15

Which of these is a Z-score you would use if conducting a two-tailed Z test with an alpha level of 0.01?

Answer 15

2.58

Question 16

A Z-score of 2.53 means that the score is ________

Answer 16

2.53 standard deviations above the mean

Question 17

When there is no actual difference between the populations being studied, but the researcher rejects the null hypothesis, what kind of statistical decision error has occurred?

Answer 17

Type 1, false positive

Question 18

If a measurement instrument gives consistent scores, the technical term we use to describe this is

Answer 18

reliability

Question 19

The accuracy or goodness of an instrument’s functioning is referred to by the technical term

Answer 19

validity

Question 20

When interpreting a boxplot, what does the top of the box tell you?

Answer 20

the 75th percentile score

Question 21

By holding constant the content of an email, but randomly assigning professors to see one of 10 different types of names for the sender (suggesting different gender/race identities), what type of validity did Katherine Milkman strengthen?

Answer 21

internal & construct

Question 22

A researcher is interested in children’s performance on a novel task after the 30 children from one classroom are randomly assigned to one of two groups: Group 1 children pretend to do the task (n = 15), while Group 2 children actually do the task (n = 15). Which of the following tests should the researcher use?

Answer 22

t-test for independent means

Question 23

What does a very small p value, say p < .0001, tell you about the effect size of a study?

Answer 23

It tells you nothing about the effect size

Question 24

power is the probability that…

Answer 24

if the research hypothesis is true, the experiment will support it.

Question 25

When figuring the pooled population variance estimate in a t test for independent means,

Answer 25

the variance estimates based on each of the samples are averaged in such as way as to give more influence to the estimate based on more participants

Question 26

A sample of children who scored highly on a measure of depression were given an energy drink once a day for one week. Three months later, their scores on the depression measure averaged significantly lower. The researcher concluded that the energy drink intervention was promising for reducing depression. Which major type of confound (from among the nine we've studied) is most threatening to this researcher's budding belief that the energy drink intervention is reducing depression?

Answer 26

regression to the mean

Question 27

How do we best ensure high interrater reliability?

Answer 27

Randomly check 20% of their ratings

Question 28

Which random assignment approach is best to use when sample sizes available for experimental conditions are small? The

Answer 28

matched

Question 29

Professor Smyth once taught Intro to Social Psychology at 6pm on Tuesdays and Thursdays. He noticed that his students were far more likely to miss on Thursdays and he wondered whether his Thursday attendance rate was significantly different from the University's overall attendance rate for Thursday classes. Fortunately, he was able to obtain the mean UVA Thursday attendance rate from University administrators. Which of the following tests would be most appropriate for his question?

Answer 29

Single sample t test

Question 30

A t distribution for df = 20 will have ____ tails compared to the Z distribution, so the cut off scores for that t distribution will be ____ .

Answer 30

fatter; more extreme

Question 31

A medical researcher hypothesizes an effect of half a standard deviation reduction (d = -0.5) in cancer cell growth from a new drug. He conducts a power analysis which indicates that a control group with N=100 and a treatment group with N=100 will yield 100% power. Using these sample sizes, the study was then executed perfectly with a double-blind design, but a statistically significant effect of the drug was not found. Which of the following is the most likely explanation for the inconclusive result?

Answer 31

The drug's effect is probably smaller than hypothesized

Question 32

Random sampling improves a study's ___ validity.

Answer 32

external

Question 33

When estimating the population variance (S2) of difference scores in a t test for dependent means, we _____

Answer 33

divide the SS by N-1

Question 34

What are the total degrees of freedom, df, for an independent means t test with total N = 467?

Answer 34

465

Question 35

The variance of a sample is a "biased" estimate of the population variance because...

Answer 35

the sample is generally less variable than the population

Question 36

Which of the following is true for both the t test for independent means AND the t test for dependent means?

Answer 36

the population means are unknown, the sample scores (in some form) are eventually compared to a t distribution, population variances are estimated from the information in the sample of scores actually studied.

Question 37

In actual practice, the usual reason for determining power before conducting a study is

Answer 37

to determine the number of participants needed to have a reasonable level of power.

Question 38

Which of the following is a consequence of increasing the sample size of a study?

Answer 38

Reduced false negative error rate, Reduced standard error

Question 39

Which one of the following correctly completes this statement: "The larger the sample size, _____

Answer 39

the higher the power."

Question 40

When conducting a planned contrast of two groups after finding a significant omnibus ANOVA result with three or more groups, which of these statistics must be recomputed?

Answer 40

The grand mean, SSbtwn, DFbtwn, the observed F value

Question 41

This graph shows the effects of three variables on a math test: gender (participants self-identified as female or male), strength of personal gender identification (relatively low or relatively high), and, by random assignment, whether or not gender identity was relevant to math performance (that is, half of participants were told that examining the relationship between gender and math performance was a focal point of the study, and half were not told this). The highest level of research design constraint used in this study is _____. As seen in the graph, there is not a simple answer to the question, "What is the effect of gender identity relevance on the math test scores of men and women?" Rather, the answer depends on whether participants' gender identification is low or high. In this case, strength of gender identification, low or high, would be referred to as a ____ variable

Answer 41

experimental, moderator

Question 42

You have conducted an Analysis of Variance of six groups that revealed a significant F value. Now you would like to test the significance of each of the possible pairwise comparisons. With one sentence, describe the purpose of applying the Bonferroni procedure as you make the comparisons.

Answer 42

The Bonferroni procedure allows setting more conservative (lower) alpha levels for each test so that the family-wise false positive rate can be maintained at a standard level, e.g., .05 or .01.

Question 43

In what way(s) are t-tests and ANOVAs similar?

Answer 43

We do not know the true population variance, We assume all populations have the same variance, We can estimate the variance of each population from the sample scores.

Question 44

Which of these is the recommended first step of data analysis?

Answer 44

Look at the data using graphical techniques

Question 45

Which of the following is true for F-tests?

Answer 45

F-tests are used to determine if three or more groups come from the same underlying population, The critical-F for a given alpha varies according to numerator and denominator degrees of freedom, F-tests cannot tell you the direction of a significant effect

Question 46

In regression, the baseline error (SSTotal) in the prediction of your criterion variable, Y, is derived from deviations of Y scores from

Answer 46

the mean of Y

Question 47

What does a statistically significant p-value indicate for a correlation coefficient?

Answer 47

The correlation is probably not zero.

Question 48

This table gives the multiple regression results for the prediction of Age of Menarche (first menstrual period). Put into words the meaning (specific interpretation) of the standardized multiple regression estimate for the effect of "harshness of early environment" in the context of this analysis.

Answer 48

With other predictors equal (or "accounted for" or "held constant"), a 1-standard deviation increase in the harshness of early environment predicts a 0.15-standard deviation decrease in age of menarche onset.

Question 49

For one variable, which value results in the smallest sum of squares?

Answer 49

mean

Question 50

The F-distribution is a heap of what sort of numbers?

Answer 50

RATIOS

Question 51

If a variable's range is restricted in your study, the strength of its correlation with another variable of interest is likely to be _______ than is true in the population.

Answer 51

lower

Question 52

if there are a few extreme outliers in an equal-interval, continuous variable, which is usually the best measure of central tendency?

Answer 52

median

Question 53

i would expect the correlation between students' scores on exam 3 and scores on the final exam in this class to be closest to...

Answer 53

0.5

Question 54

which of these was NOT among the "troubling trio" of study characteristics cited in 2015 in an editorial by the editor of Psychological Science?

Answer 54

a null result

Question 55

what does it mean to say that something is significant at the p<.05 level?

Answer 55

this result would occur less than 5% of the time if the null is true

Question 56

which Z score is legitimate and has the greatest magnitude?

Answer 56

-5.3

Question 57

what percentile ranking would a z-score of -2.0 correspond to (roughly)?

Answer 57

2%

Question 58

in an ANOVA, if the null hypo is true, then...

Answer 58

there is less variance among means of samples than if the null hypo were not true

Question 59

"Horse NUMBER 5 came in 2ND PLACE in the race, with a time of 57 SECONDS" what is the type of data corresponding to each of these 3 numbers?

Answer 59

nominal, ordinal, ratio

Question 60

if the score for your sample was -3.8, and the critical cutoff scores were -2.0 and +2.0 when testing with alpha = .05, then you would:

Answer 60

reject the null

Question 61

what is the definition of the comparison distribution?

Answer 61

the sampling distribution you would get if the null hypothesis is TRUE

Question 62

type 1 error is defined as...

Answer 62

rejecting the null when you should have failed to reject the null

Question 63

holding all else constant, as the DF decrease, the critical cutoff score...

Answer 63

gets more extreme (further from middle)

Question 64

if a planned study has estimated power of .05, would this be considered "good power" and proceeding with the study would be advised?

Answer 64

probably no

Question 65

the minimum number of studies necessary for a meta-analysis...

Answer 65

is 2

Question 66

in a study comparing the means of 2 samples, what measure of effect size would you typically use?

Answer 66

d

Question 67

you make up a new scale, and you are pretty sure that it measures what it's supposed to be measuring. another way to say this is that your scale has good

Answer 67

construct validity

Question 68

what is NOT true about the standard regression equation when you have 1 predictor?

Answer 68

Beta B is always the same as b

Question 69

if your unstandardized regression equation for sweet tooth (X) predicting candy liking (Y) is Yhat = 6.7 + .25X. this means:

Answer 69

with every one unit increase in sweet tooth we expect a .25 unit increase in candy liking

Question 70

if you increase the alpha (p-value) of your significance test (from .05 to .10):

Answer 70

type 1 error (false positive) rate gets bigger

Question 71

increasing the effect size is one way to increase power because:

Answer 71

it makes the score in your sample likely to be more extreme (step 4)

Question 72

according to "Scientific Utopia," which of these is an ineffective strategy for preventing false results?

Answer 72

conceptual replication & raising the barrier for publication

Question 73

when you do the 5 steps of hypothesis testing to test the significance of a correlation:

Answer 73

you use a t-table distribution

Question 74

if you calculated PRE and it was -.32, this means:

Answer 74

you did something wrong in your calculation

Question 75

I think the dogs on my block are unusually long and I want to test this hypothesis. I know the population mean of dog lengths from wiki, but not the SD. What do I do?

Answer 75

single sample t test

Question 76

which of these is NOT one of Nosek & Bar-anan's "Utopia I" recommendations for changing scientific communication?

Answer 76

anonymous peer review