Psych 214- Stats

Question 1

If, in a perfect world, you were able to test every single individual in the world of interest to your study, you would be testing:

a. The population

b. The sample

c. The subsample

d. The group

e. The random sample

Answer 1

a. the population

Question 2

Descriptive statistics…

a. Describe characteristics and tendencies of your sample

b. Allow you to draw conclusions based on extrapolations and inferences

c. Use data from the sample of participants in the experiment to compare the treatment groups and make generalizations about the larger population of participants

d. Provide a quantitative method to decide if the null hypothesis (H0) should be rejected

e. Can only be qualitative

Answer 2

a. describe characteristics and tendencies of your sample .

Question 3

Which is NOT a measure of variability

a. Mean

b. Range

c. Standard Deviation

d. Variance

e. Sum of Squares

Answer 3

a. mean

Question 4

The range is…

a. One standard deviation away from the mean

b. The difference between the lowest and highest values, squared

c. The difference between the lowest and highest values divided by the number of participants

d. The mean of our data, plus one standard deviation

e. The difference between the lowest and highest values

Answer 4

e. the difference between the lowest and highest values

Question 5

Which are NOT sources of variability in data
a. Population means

b. Random (residual) errors

c. Treatment effects

d. Experimenter effects

e. Individual differences

Answer 5

a. Population means

Question 6

Matilda has three independent groups who each receive a different treatment to help them quit smoking – either hypnotherapy, psychotherapy or nicotine patches. To assess statistical differences between groups, Matilda should run:
a. Three independent group t-tests

b.A one-factor within-participant ANOVA

c. A two-factor between-participant ANOVA

d. A one-factor, between-participant ANOVA

e. A set of descriptive statistics

Answer 6

d. a one factor between participant ANOVA

Question 7

The grand mean

a. Takes all of the group means and divides them by the number of groups

b. Takes all of the group means and divides them by the number of participants

c. Takes all of the group means and divides them by the number of participants, minus one

d. Is another name for the mean of a subgroup

e. Will always be below 10

Answer 7

a. Takes all of the group means and divides them by the number of groups

Question 8

The larger the F ratio

a. The larger the noise in a study

b. The smaller the signal in comparison to the noise

c. The larger the signal in comparison to the noise

d. Generally, the less likely your difference between groups will be significant

e. The harder your equation was to calculate

Answer 8

c. the larger the signal in comparison to the noise

Question 9

The F ratio is the:

a. Another word for the T statistic

b. The within group variability divided by the between group variability

c. All the error within groups added together

d. The error term divided by the within group variability

e. The between group variability divided by the within group variability

Answer 9

e. The between group variability divided by the within group variability

Question 10

The between group degrees of freedom and the within group degrees of freedom

a. Are exactly the same thing, they are synonymous

b. Different. The between group degrees of freedom is the number of levels minus 1, while the within group degrees of freedom is the number of participants for each level minus 1.

c. Different. The between group degrees of freedom is the number of participants for each level minus 1, while the within group degrees of freedom is the number of levels minus 1.

d. Both tell you if you have skewed data

e. Should never both be included in an ANOVA calculation

Answer 10

b. Different. the between group degrees of freedom is the number of levels - 1
the within group degrees of freedom is the number of participants for each level minus 1

Question 11

Which statement about the ANOVA is NOT true
a. It is useful for examining mean differences between three or more groups

b. It provides a greater risk of producing type I errors than a series of T-tests

c. It requires at least one factor

d. It requires continuous outcome (DV) data

e. It is an Analysis of Variance

Answer 11

b. it provides a greater risk of producing type 1 errors than a series of T-Tests

Question 12

A p value of > 0.05 means
a. We have failed to reject the null hypothesis

b. We reject the null hypothesis

c. We have found significant differences between our groups

d. Our data are drawn from different populations

e. We have both an F ratio and a t value

Answer 12

a. we have failed to reject the null hypothesis

Question 13

NA1 means
a.The number of means across our groups

b.The number of factors in our study

c. The number of scores for group A1

d. The number of levels in our study

e. The mean of group A1

Answer 13

c. the number of scores for group A1

Question 14

A p > .05 means

a. There is insufficient evidence to conclude that any means significantly differs from any others.

b. At least one of the pairs of means is different. The question is, which pairs?

c. We reject the null hypothesis.

d. Our F ratio is identical to our t-test statistic.

Answer 14

a. there is insufficient evidence to conclude that any means significantly differs from any others

Question 15

What is the main problem of running multiple statistic comparisons:

a. This takes up a lot of CPU processing power.

b. This will change our F ratio statistic.

c. We inflate the probability of making a Type I Error.

d. We inflate the probability of making a Type II Error.

e. These will still not allow us to compare our different groups.

Answer 15

c. We inflate the probability of making a Type 1 Error

Question 16

Regarding planned comparisons, which of the following is NOT true:

a. Planned comparisons are a focussed approach to examine specific group differences.

b. Planned comparisons allow us to test differences between groups of interest without needing to examine group differences between all groups.

c. Planned comparisons should be pre-specified.

d. A researcher should keep the number of planned comparisons as low as possible to reduce the likelihood of producing a Type I error.

e. Planned comparisons will produce another F statistic

Answer 16

e. planned comparisons will produce another F-statistic

Question 17

Which is an assumption of ANOVA:

a. Homogeneity of variance.

b. Heterogeneity of variance.

c. Sum of squares variability.

d. Standard deviations of data.

e. Unequal ranges of data.

Answer 17

a. Homogeneity of Variance

Question 18

The assumption of normality assumes:
a. We have ‘normal’ participants taking part in our research.

b. All research is carried out using normal measures.

c. All research is carried out using normal methodologies.

d. We have normal distributions for each subgroup’s data.

e. We use normal parametric tests.

Answer 18

d. We have normal distributions for each subgroup’s data

Question 19

Which of the following statements is true:

a. Rogue data can never be transformed.

b. Rogue data is named after the motion picture ‘Rogue One: A Star Wars Story’.

c. There are no strategies which can improve rouge data.

d. If you solve your rogue data issue, apply a non-parametric test

e. There are solutions to rogue data, but none are perfect panaceas guaranteed to work.

Answer 19

e. there are solutions to rogue data, but none are perfect or guaranteed to work

Question 20

Which of the following statements is NOT true:

a. Outliers can bias or change our predictive models.

b. Outliers should always be removed before any analysis.

c. Outliers can lead to violations in our ANOVA assumptions.

d. Outliers can result in high error and exaggerated predictions.

e. Outliers are data points which significantly differ from other observations

Answer 20

b. Outliers should always be removed before any analysis

Question 21

A Bonferroni correction:

a. Is an adjusted p value, which divides our original p value by the number of study levels and results in a more conservative p value.

b. Is an adjusted p value, which divides our original p value by the number of participants and results in a more conservative p value.

c. Is an adjusted p value, which divides our original p value by the number of tests and results in a more conservative p value.

d. Is an adjusted p value, which divides our original p value by the number of study levels and results in a more liberal p value.

e. Is an adjusted p value, which divides our original p value by the number of tests and results in a more liberal p value.

Answer 21

c. Is an adjusted p value, which divides our original p value by the number of tests and results in a more conservative p value

Question 22

How do planned comparisons and post hoc tests differ:
a. Planned comparisons always happen before you run an ANOVA, whereas post hoc tests occur after.

b. Post hoc tests always happen before you run an ANOVA, whereas post planned comparisons occur after.

c. Planned comparisons produce F statistics, whereas post hoc tests produce Q statistics.

d. Planned comparisons test only the mean differences of interest, while post hoc tests compare all possible combinations of mean differences.

e. They are interested in different studies

Answer 22

d. planned comparisons test only the mean differences of interest, while post hoc tests compare all possible combinations of mean differences

Question 23

If running 4 statistical tests, what is the approximate probability of receiving a type I error?

a. 5%

b. 10%

c. 14%

d. 18.5%

e. 22.6%

Answer 23

d. 18.5%

Question 24

Transforming data involves:

a. Taking every score from each participant and applying a uniform mathematical function to each.

b. Replacing each individual’s score with the group mean.

c. Taking the mean value of the group and applying a uniform mathematical function to it.

d. Reporting a new data set and ignoring the older data.

e. Applying a mathematical function to normal data which does not violate assumptions.

Answer 24

a. taking every score from each participant and applying a uniform mathematical function to each

Question 25

The Kruskall-Wallace one-way Analysis of Variance by Ranks is:
a. A post-hoc test.

b. A planned comparison test.

c. A parametric alternative to the ANOVA.

d. A type of correlation.

e. A non-parametric alternative to the ANOVA

Answer 25

e. a non-parametric alternative to the ANOVA

Question 26

Which of the following is a type of data transformation:
a. The Tukey transformation.

b. The square root transformation.

c. The homogeneity transformation.

d. The outlier removal.

e. The post hoc transformation

Answer 26

b. the square root transformation

Question 27

Which of the following does NOT represent a within-participants design?
a. Each participant had four trials on a task and each score was recorded.

b. Each participant was given five difference dosage levels with testing after each dose.

c. Each participant was assigned to one of three groups, based on whether they were introvert, extrovert or neutral. A questionnaire then rated each participant’s political opinions.

d. Each participant was tested in the morning, afternoon and evening.

e. Each participant recorded their mood, had a therapy session and then recorded their mood again.

Answer 27

c. each participant was assigned 1 of 3 groups based on whether they were an introvert, extrovert or neutral

Question 28

Differences among participants in overall performance constitute a source of error for:
a. A participant

b. A within-participant design

c. A between-participant design

d. Independent groups

e. Random designs

Answer 28

c. a between participant design

Question 29

The assumption of sphericity is an assumption affecting:
a. The analysis of variance

b. The t-test

c. The linear regression

d. The between-participants ANOVA

e. The within-participants ANOVA

Answer 29

e. the within-participant ANOVA

Question 30

Which of the following is NOT a limitation of the within-participants design
a. Keeping individual differences constant

b. Demand characteristics

c. Practice effects

d. Fatigue effects

e. Order effects

Answer 30

a. keeping individual differences constant

Question 31

Participants of a within-participants design can differ in respect to their overall performance scores and also…:
a. in the overall number of treatments they receive

b. in the way they are manipulated during a study

c. in the total number of data points which are extracted

d. in the trends of their scores

e. in the conditions of the laboratory

Answer 31

d. in the trends of their scores

Question 32

If A1 = 6 and A2 = 10, what is ∆A
a. 0.4

b. 0.6

c. 60

d. 16

e. 4

Answer 32

e. 4

Question 33

The within-participant ANOVA allow us to omit the error that would be typically contributed by individual differences. This makes:
a. the chances of violating the assumptions of ANOVA more likely

b. it more likely we will find a statistically significant result, as there is less noise

c. it less likely we will have robust measures

d. it more likely we will need to control for a person’s personality

e. it less likely that the results are interpretable

Answer 33

b. its more likely we will find a statistically significant result, as there is less noise

Question 34

Low between-participant variability means:
a. our participants‘ means are similar

b. our participants‘ means are very different

c. the trends and direction of participant scores are consistent

d. the trends and direction of participant scores are inconsistent

e. the variability within a participant is volitile

Answer 34

a. our participants’ means are similar

Question 35

High residual variance means:
a. our participants‘ means are similar

b. our participants‘ means are very different

c. the trends and direction of participant scores are consistent

d. the trends and direction of participant scores are inconsistent

e. the variability within a participant is volitile

Answer 35

d. the trends and directions of participant scores are inconsistent

Question 36

In virtually all within-participants studies, we:
a. would be predominately interested in between-participant differences

b. would hypothesise the amount of change a participant would experience

c. would hypothesise that a score at one time would significantly differ from a score at another

d. would control for all extraneous variables

e. would expect that we would violate the majority of data assumptions

Answer 36

c. would hypothesise that a score at one time would significantly differ from a score at another

Question 37

A within-participants design is also known as:
a. A between-subjects design

b. An independent groups analysis

c. A between-participants design

d. A factorial repetition design

e. A repeated-measures design

Answer 37

e. a repeated measures design

Question 38

We calculate the F ratio for a within participants ANOVA the same as for the between participants ANOVA, with the exception that…:
a. We are not interested in how participants vary from one another

b. We are not interested in the error term

c. We are not interested in within-participant changes over time

d. We are not interested in the directional change of a participants score over time

e. We are no interested in the sum of squares

Answer 38

a. we are not interested in how participants vary from one another

Question 39

How can we mitigate the impact of order effects?
a. Recruit from a non-WEIRD population

b. Change our hypotheses in the opposite order

c. Run the experiment twice in fast succession

d. Change the time in which we run our experiment

e. Counterbalance our conditions for participants

Answer 39

e. counterbalance our conditions for participants

Question 40

What is removed from our F ratio when calculating a within-participants F statistic
a. The p value

b. The individual differences

c. The group differences

d. The error term

e. The signal to noise ratio

Answer 40

b. the individual differences

Question 41

Which measure of central tendency would be most appropriate when scrutinizing an asymmetrical distribution?
a. Mean

b. Median

c. Mode

d. Standard deviation

e. Range

Answer 41

b. Median

Question 42

Variance is…
a. The difference between the highest and lowest score

b. A subtraction of the lowest value in the distribution by the highest value

c. The average deviation around the mean of a distribution

Answer 42

c. The average deviation around the mean of a distribution

Question 43

Non-parametric tests…
a. Require that a dependent variable is ratio

b. Require normal data

c. Require homogeneity of variance

d. Are generally less powerful than parametric tests

e. Are less robust to violations of ANOVA

Answer 43

d. are generally less powerful than parametric tests

Question 44

A researcher wants to investigate whether those who have read Harry Potter are typically more empathetic than those who have never read Harry Potter. The researcher samples 200 avid readers, 100 of whom have read the Harry Potter series and 100 who have not. The researcher then asks participants to complete the Toronto Empathy Questionnaire (TEQ) and tests for group differences. Data are scrutinized and there are no data assumption violations. Which statistical test would be most appropriate in this context?

a. Paired samples T-test

b. One factor, between-participants ANOVA

c. One factor, within-participants ANOVA

d. Kruskall-Wallace one-way Analysis of Variance by Ranks

e. The Friedman repeated measures test

Answer 44

b. One factor, between participants ANOVA

Question 45

A researcher wants to know whether a new medication ‘R-roids’ would improve students’ statistical coding capabilities. The researcher randomly assigns students to one of three conditions – an ‘R-roids’ 50mg per day condition, an ‘SPSSirpin’ 50mg per day condition, or a placebo 50mg condition. The researcher then measures the number of coding bugs a participant can identify and solve. The data are scrutinized and the assumption of normality and homogeneity of variance are both grossly violated. Which statistical test would be most appropriate in this context?
a. Paired samples T-test

b. One factor, between-participants ANOVA

c. One factor, within-participants ANOVA

d. Kruskall-Wallace one-way Analysis of Variance by Ranks

e. The Friedman repeated measures test

Answer 45

d. Kruskall-Wallace one-way analysis of Variance by Ranks

Question 46

When running a between-groups ANOVA, a typical measure of effect size would be:
a. Eta-squared

b. Odds ratio

c. Cohen’s d

d. Pseudo r square

e. Cohen’s h

Answer 46

a. Eta squared

Question 47

When would we transform our data?
a. When data come from different populations

b. When data are normally distributed

c. When data come from normal populations

d. When data are skewed

e. When we fail the assumption of independence of data

Answer 47

d. when data are skewed

Question 48

How do we load a package in R?
a. With the library() function

b. With the install.packages() function

c. With the load() function

d. With the package() function

e. The packages a researcher needs are already loaded from the get-go.

Answer 48

a. with the library() function

Question 49

How many Idependent variables are there in a 2 x 3 x 2 factorial design

Answer 49

3

Question 50

For a two-factor research study with three levels for factor A and four levels for factor B, how many participants are needed to obtain ten scores in each treatment condition if both factors are between participants?
a. 10

b. 30

c. 60

d. 90

e. 120

Answer 50

e. 120

Question 51

For a two-factor research study with three levels for factor A and four levels for factor B, how many participants are needed to obtain ten scores in each treatment condition if both factors are within participants?
a. 10

b. 30

c. 60

d. 90

e. 120

Answer 51

a. 10

Question 52

For a two-factor research study with three levels for factor A and four levels for factor B, how many participants are needed to obtain ten scores in each treatment condition if factor A is between participants and factor B is within participants?
a. 10

b. 30

c. 60

d. 90

e. 120

Answer 52

b. 30

Question 53

A researcher who is examining the effects of chess expertise and the positions of chess pieces on memory for chess board positions uses a factorial study comparing three different groups of chess players (novice vs. intermediate vs. expert) and two types of chess positions (real positions vs. random positions). How many factors are there in the experiment?
a. 1

b. 2

c. 3

d. 4

e. 5

Answer 53

b. 2

Question 54

Main effects provide information about
a. The independent effect of each factor.

b. The effect of one factor at each level of the second factor.

c. Whether we should expect an interaction between factors.

d. How two factors combine to influence the dependent variable.

e. The chief effects of interest.

Answer 54

a. the independent effect of each factor

Question 55

A significant interaction implies that
a. The main effects of each factor cannot be significant.

b. Two factors are exerting independent effects.

c. The effect of one factor depends on the different levels of a second factor.

d. The effect of one factor is insensitive to the different levels of a second factor.

e. The main effect of one factor must be significant, whereas the main effect of the second factor must be nonsignificant.

Answer 55

c. the effect of one factor depends on the different levels of a second factor

Question 56

A significant interaction is interpreted by
a. Examining the main effects of each factor.

b. Multiplying the main effects of each factor.

c. Comparing the level means of one one factor against the level means of the second factor.

d. Performing a simple effects analysis.

e. Conducting an interaction decomposition analysis.

Answer 56

d. performing a simple effects analysis

Question 57

When inspecting an interaction (or line) plot of the results of a factorial study, there is an interaction if the lines
a. are parallel.

b. are nonparallel.

c. form a cross, and only if they form a cross.

d. have the same angle.

e. do not touch one another.

Answer 57

b. are nonparallel

Question 58

Which of the following is not a possible outcome from a two-factor study?
a. One significant main effect and a significant interaction.

b. Two significant main effects and a significant interaction.

c. Two significant main effects and no significant interaction.

d. No significant main effect for either factor but a significant interaction.

e. All of the above are all possible outcomes.

Answer 58

e. all of the above are all possible outcomes

Question 59

Why might the main effects in a factorial study not provide an accurate description of the results?
a. If there is a significant interaction, then the factors are combining to influence the dependent measure, meaning the results cannot be understood in terms of the unique effects of each factor.

b. There may be other factors that influence the results.

c. The main effects may not be valid.

d. The main effects may not be reliable.

e. The main effects always provide an accurate description of the results.

Answer 59

a. if there is a significant interaction, then the factors are combining to influence the dependent measure, meaning the results cannot be understood in terms of the unique effects of each factor

Question 60

How many error terms are used to calculate the F ratios for a between-participants two-factor ANOVA?
a. 1

b. 2

c. 3

d. 4

e. 5

Answer 60

a. 1

Question 61

A two-factor ANOVA has dfA = 1, dfB = 2, dfAXB = 2, and dfTOTAL = 59. What is the value of dfS/AB?
a. 1

b. 5

c. 20

d. 54

e. 76

Answer 61

d. 54

Question 62

A two-factor ANOVA has SSBETWEEN = 52, SSA = 18, and SSB = 22. What is the value of SSAXB?
a. 12

b. 30

c. 34

d. 40

e. 92

Answer 62

a. 12

Question 63

Which of the following are possible ways in which a pair of simple main effects might differ in their trend?
a. One of a pair has a significant difference but not the other.

b. Both simple main effects are significant, but in the opposite direction.

c. The interaction and all four simple main effects are significant, but the difference for both members of each pair of simple main effects is in the same direction.

d. a and b.

e. a, b, and c.

Answer 63

e. a, b, and c

Question 64

A mixed factorial design contains
a. at least one between-participants factor and at least one within-participants factor.

b. only between-participants factors.

c. only within-participants factors.

d. a mixture of dependent variables.

e. none of the above

Answer 64

a. at least one between-participant factor and at least one within-participant factor

Question 65

A fully within-participants factorial design contains
a. at least one between-participants factor plus at least one within-participants factor.

b. only between-participants factors.

c. only within-participants factors.

d. a mixture of dependent variables.

e. none of the above

Answer 65

c. only within participant factors

Question 66

In a two-factor mixed design
a. a single error term is used to test the significance of the main effect of each factor and the interaction.

b. the significance of the main effects and interaction are tested using their own unique error terms.

c. one error term is used to test the significance of the main effect of the between-participants factor, and another error term is used to test the significance of the main effect of the within-participants factor and interaction.

d. one error term is used to test the significance of the main effect of each factor and another error term is used to test significance of the interaction.

e. None of the above.

Answer 66

c. one error term is used to test the significance of the main effect of the between-participants factor and another error term is used to test the significance of the main effect of the within-participant factor and interaction

Question 67

When calculating the simple main effects for a two-factor mixed design using the pooled error terms approach
a. the error term for testing the significance of the simple main effects of both the between-participants and within-participants factors is the overall within-group variance.

b. the error term for testing the significance of the simple main effects of both the between-participants and within-participants factors is the within-participants factor error term from the initial ANOVA.

c. the error term for testing the simple main effects of both the between-participants and within-participants factors is the between-participants factor error term from the initial ANOVA.

d. the error term for testing the significance of the simple main effects of the between-participants factor is the overall within-group variance, whereas the error term for testing the significance of the simple main effects of the within-participants factor is the within-participants factor error term from the initial ANOVA.

Answer 67

d. the error term for testing the significance of the simple main effects of the between-participants factor is the overall within-group variance, whereas the error term for testing the significance of the simple main effects of the within-participants factor is the within-participants factor error term from the initial ANOVA.

Question 68

An alternative to performing a simple main effects analysis using F ratios would be to
a. just analyse the main effects of each factor

b. calculate a separate t-test for each pair of means being compared

c. calculate a separate effect size estimate for each pair of means being compared

d. calculate a separate ANOVA for each pair of means being compared

e. none of the above

Answer 68

b. calculate a separate t-test for each pair of means being compared

Question 69

A 2 (factor A: level A1 vs. level A2) x 3 (factor B: level B1 vs. level B2 vs. level B3) fully within-participants ANOVA reveals no significant main effect of factor A, a significant main effect of factor B, and no significant interaction. To interpret the significant main effect of factor B
a. no further action is necessary.

b. planned comparisons or post-hoc tests are necessary to identify which levels of the factor differ significantly from one another.

c. we must calculate the simple main effects of factor B.

d. we must plot the level means of the factor to see how they differ.

e. none of the above.

Answer 69

b. planned comparisons or post-hoc tests are necessary to identify which levels of the factor differ significantly from one another

Question 70

The sphericity assumption
a. is not something worth worrying about.

b. is just another name for the homogeneity of variance assumption.

c. applies to single-factor within-participants designs with three or more levels.

d. applies to single-factor and factorial within-participants designs containing factors with three or more levels.

e. applies to single-factor and factorial within-participants designs containing factors with three or more levels, as well as mixed designs containing within-participants factors with three or more levels.

Answer 70

e. applies to single factor and factorial within-participant designs containing factors with three or more levels, as well as mixed designs containing within-participant factors with three or more levels