Data Analysis in Psychological Research Exam 3

(1-way ANOVA)
What is the difference between within groups variation (SS_within) and between groups variation (SS_between)? What contributes to within and between variation?

Between measures how much group means differ from grand mean. Within measures how much individual scores differe from their own group mean. Between caused by differences between treatment conditions and is influenced by IV, within is influenced by individ differences between participants random variability and measurement error.

(1-way ANOVA)

What is the F-statistic made up of?

MSbetween divided by MSwithin.

(1-way ANOVA)

What are the properties of the F-distribution? (e.g., shape, df, etc.)

Values are never negative. Distribution is positively skewed. Shape depends on dfbetween and dfwithin. As df increases, distribution becomes more normal.

(1-way ANOVA)

When do we do Post Hoc tests?

Only when F-test is significant.

(1-way ANOVA)

Why do we use HSD as a post hoc, rather than just doing several t-tests? (What does this have to do with experiment-wise alpha level?)

Multiple tests inflate type 1 error rate.
Tukey’s HSD controls this by keeping constant alpha level.

(1-way ANOVA)

What are the null and alternative hypotheses in ANOVA?

Hnull: all group means are equal
Halt: at least one group mean is different

(1-way ANOVA)

In a repeated measures ANOVA, how is the variance partitioned? How is this partitioning different from the independent groups 1-Way ANOVA?

Total variance is divided into between treatments (effect of IV), between subjects (individ differences) and error.
Removes between-subject variability, reducing error —> increases power

(2-way ANOVA)

What is a main effect?

the effect of one IV averaged across levels of the other variable.

(2-way ANOVA)

What’s the difference between additive and interactive effects?

Additive: Each IV affects the DV indepnedently and no interaction present.
Interactive effects: the effect of one IV depends on the level of the other IV

(2-way ANOVA)

Be able to interpret main effects, interactions, and interaction plots and how main effects and interactions do or do not relate to each other.

Lines in an interaction polot are not parallel.

(2-way ANOVA)

What is the difference between within groups variation (SSwithin) and between groups variation (SSbetween)? What contributes to within-subjects and between-subjects variation?

Sswithin: variability within each group (error, individ diff)
SSbetween: variability between group means (effects of IVs)
Within-subjects variation: changes within the same individuals

Between subjects variation: differences between different individuals

(2-way ANOVA)

Understand the breakdown of SS for 2-Way independent groups ANOVA

SStotal = all others of them combined.

(2-way ANOVA)

What are the null and alternative hypotheses in a 2-Way ANOVA?

The one with 3 diff H tests, one for each IV and another of the both together.

(2-way ANOVA)

What are a 2-Way Independent Groups ANOVA, a 2-Way RM ANOVA and a 2-Way Mixed ANOVA used for?

Independent: two between-subjects IVs

RM: two within-subjects IVs

Mixed: one between-subjects IV and one within subjects (IV)

(Correlation / Regression)

When do you use correlation?

Examingin strength & direction of a linear relationship between two continuous variables

(Correlation / Regression)

What can a regression analysis allow us to do?

allows us to predicgt values of the DV from the IV

(Correlation / Regression)

What is the range of possible values for r and what do they mean?

1 on other side is perfect pos/neg, 0 is no relationship.

(Correlation / Regression)

How does Pearson’s r indicate the strength/magnitude and direction of the relationship?

sign indicates direction, absolute value indicates strength

(Correlation / Regression)

What is the Coefficient of determination (r2) and what does it mean?

r² is the proportion of variance in Y explained by X

(Correlation / Regression)

What happens when there is a perfect positive or perfect negative correlation between X and Y?

all points are exactly on a stright line and there is no prediction error.

(Correlation / Regression)

What are the null and alternative hypotheses for a correlation?

Hnull: p=0
Halt: p does not equal 0

(Correlation / Regression)

What is restricted range (including the idea of floor or ceiling effects) and how is related to the size of the correlation coefficient?

occurs when values ar elimited, reduces size of correlation coefficient

(Correlation / Regression)

How can extreme values affect the size of the correlation coefficient?

can increase or decrease the correltion

can distort the relationship

(Correlation / Regression)

How can outliers affect the size of the correlation?

can artificially inflate or deflatre r

can change direction

(Correlation / Regression)

What are the slope and the intercept in a regression equation?

slope are thange in Y or a 1-unit increase in X

intercept are value of Y when X=0

(Correlation / Regression)

Know how to interpret a regression equation (i.e., what do the slope and intercept mean?)

Slope: how much Y changes with X

Intercept: baseline value of Y when X=0

(Chi-Square)

When do we use the chi-square test?

When analyzing categorical data

(Chi-Square)

What is the difference between the goodness of fit χ2 and the χ2 test for independence?

1. compares observed vs expected distr

2. relationship between two categorical variables

(Chi-Square)

What are the null and alternative hypotheses for the chi-square tests mentioned above?

Goodness: observed = expected

Independence: variables are unrelated

(Chi-Square)
What are the properties of the chi-square distribution?

values are greater than or equal to 0

right-skewed

shape depends on df

What test to use?
Is the number of HW assignments that students turn in on time associated with their scores on the last exam? ________

Correlation/Regression

What test to use?

Is there an equal number of sophomores, juniors, and seniors in this class? ________

Chi-Square

What test to use?

Is there a difference in exam scores between sophomores, juniors, and seniors? ________

One-Way Independent Groups ANOVA

What test to use?

Is there a gender difference in exam scores? ________

Independent Groups t

What test to use?

Did students in this class do better on exam 1 than they did on exam 2? _______

Related/Dependent Groups t

What test to use?

Among pet-owners, are women more likely than men to own cats rather than dogs? ____

Chi-Square

What test to use?

Who has the largest number of friends, cat owners, dog owners, owners of other pets (not cats and dogs) or non pet-owners? ________

One-Way Independent Groups ANOVA

What test to use?

When people are asked to rate, on a 7-point scale, how much they like a type of animal, do people report liking dogs more than they like cats? _______

Related/Dependent Groups t

What test to use?

A researcher tells participants that they either failed or succeeded at a task and then measures their self-esteem on a 10-point scale. Are men more likely than women to report lower self-esteem after a failure? ________

2-Way Independent Groups ANOVA

What test to use?

A researcher is interested in the effects of music choice on cognitive performance. Participants perform a cognitive task (which can be scored up to 100 pts) while listening to no music at all, listening to soothing music, or listening to loud rock music. ________

One-Way Independent Groups ANOVA

What test to use?

At a house party, people can choose to drink beer, vodka mixed drinks, gin mixed drinks, or non-alcoholic beverages. Are all 4 drink choices equally preferred? ________

Chi-Square

What test to use?

Do people who score highly on a self esteem questionnaire tend to score low on a depression questionnaire? ________

Correlation/Regression

What test to use?

A researcher interested in the effects of noise level and lighting on cognitive task performance has participants take an SAT math test while in a quiet or loud room that is either dark or brightly lit. ________

2-Way Indepndent Groups ANOVA

What test to use?

Are psych majors more likely than science majors to go to graduate school? ________

Chi-Square

What test to use?

Does adolescents’ agreement with their parents’ values (rated on a 50-point scale) increase or decrease from age 14 to age 17 (measured at both ages)? ________

Related/Dependent Groups t

What test to use?

A researcher studying the effectiveness of a new blood pressure medication recruits participants and gives them either the new medication or a placebo and then measures their blood pressure. __________

Independent Gruops t

What test to use?

Does the number of romantic relationships one has had prior to marriage predict one’s score on a marital satisfaction scale? ________

Correlation/Regression

What test to use?

To see how job satisfaction changes over time, a researcher measures the same individuals’ job satisfaction at three time points: just after being hired, after 1 year on the job, and after 5 years on the job. _______

One-way Repeated Measures ANOVA

What test to use?

A researcher wants to determine if exposure to violence causes individuals’ to remember more violent than nonviolent words. The researcher places participants into one of two conditions: viewing a violent film or viewing a nonviolent film. Then the participants listen to a list of words that contains both violent and nonviolent words and are asked to write down as many words as they can remember. ______

2-Way Mixed ANOVA

What test to use?

Participants are asked to evaluate the competence of workers who vary in gender (male or female) and agency (agentic personality or communal personality). All participants are given information about 4 different workers, representing each possible combination of gender and agency. ______

2-Way Repeated Measures ANOVA

In a one-way independent groups ANOVA, the value of SS_{within treatments} is determined by:

1. The size of the treatment effect

2. The size of the sample variances

3. The sample size

4. All of the above

The size of the sample variances

In the F-ratio for the independent groups ANOVA, the numerator of the F-ratio represents:

1. The treatment effect

2. The sample variance

3. The critical value

4. The effects of individual differences

The treatment effect

When conducting an independent groups ANOVA, a significant result is typically followed up by a post hoc test, such as Tukey’s HSD, rather than simply computing separate t-tests to compare each pair of groups. This is because:

1. Tukey’s HSD reduces the chance of a Type I error

2. Tukey’s HSD reduces the chance of a Type II error

3. Tukey’s HSD is more likely to yield a significant result

4. T-tests will result in computation errors

Tukey’s HSD reduces the chance of a Type I error

In a 1-way independent groups ANOVA, the null hypothesis states that:

A. All of the groups have the same mean.

B. At least one group mean is different from at least one other group mean.

C. All of the group means differ from one another.

D. The variances of all the groups are different from one another.

All of the groups have the same mean.

Which of the following is NOT true of the F-distribution?

1. F-values are always positive

2. The F distribution is normally distributed.

3. The F distribution has two parameters.

4. All of the above are true.

The F distribution is normally distributed.

The calculations for repeated measures (RM) ANOVA differ from the calculations for independent groups ANOVA because:

1. In RM ANOVA, the variability due to individual differences is not part of the F-ratio’s denominator.

2. In RM ANOVA, the variability due to individual differences is included in the F-ratio’s denominator.

3. In RM ANOVA, variability due to treatment effects is not part of the F-ratio calculation.

4. In independent ANOVA, variability due to treatment effects is not part of the F-ratio calculation.

In RM ANOVA, the variability due to individual differences is not part of the F-ratio’s denominator.

If a regression analysis predicting test scores from the number of hours students spend studying shows evidence that studying improves test scores, we would expect the analysis to give us:

1. a positive Y-intercept

2. a negative slope

3. a positive slope

4. a slope of zero

a positive slope

A researcher studying the correlation between household income and crime rates samples only people who live in relatively poor neighborhoods. This would likely result in ____ for the measure of income.
A.      directionality
B.      floor effects
C.      high variability
D.      ceiling effects

floor effects

10. A non-parametric test that assesses how well an observed distribution of frequencies corresponds to an expected distribution of frequencies is:

A. Linear regression

B. Correlational analysis

C. The χ2 goodness of fit test

D. The χ2 test for independence

The χ2 goodness of fit test

11. A non-parametric test that assesses the extent to which two nominal variables are related to one another is:

A. Linear regression

B. Correlational analysis

C. The χ2 goodness of fit test

D. The χ2 test for independence

The χ2 test for independence

12. If X and Y are positively correlated that means that as scores on X increase, scores on Y:

A. decrease

B. increase

C. remain the same

D. none of the above are necessarily true

increase

13. In the regression equation Y = 10 + 4X, 10 is the _____ and 4 is the _____

A. Intercept; slope

B. Slope; intercept

C. Predictor; outcome

D. Mean; variability

A. Intercept; slope

4. The correlation between two variables, X and Y = -1.00. This means that:

A. A calculation error was made (a correlation of -1.00 is impossible).

B. The two variables are perfectly negatively correlated.

C. The two variables are perfectly positively correlated.

D. The two variables have no significant association with each other.

B. The two variables are perfectly negatively correlated.

15. When two variables, X and Y, are perfectly correlated, this means:

A. The correlation coefficient will be zero.

B. The correlation coefficient will be 100.

C. The dots will form a straight line on the scatterplot of X and Y.

D. The correlation coefficient will be positive.

C. The dots will form a straight line on the scatterplot of X and Y.

16. Pearson’s r ranges from ____ to _____

A. -1.00 to 1.00

B. 0 to 1.00

C. -1.00 to 0.

D. 0 to 100.

A. -1.00 to 1.00

17. A politician is interested in how whether or not people who live in the city are more likely than people who live in the suburbs to support a proposal to increase subway fares. If these two variables are actually independent, what is the expected value for the number city dwellers or favor the policy?

Residence Opinion about proposal

Favor Oppose

City 60 110

Suburbs 150 80

A. 60

B. 89.3

C. 210

D. 80.8

B. 89.3

18. Which of the following correlation coefficients represents the strongest relationship between the two variables?

A. –0.7

B. +0.6

C. 0

D. –0.5

A. –0.7

19. Which of the following is most likely to increase the size of Pearson’s r?

A. There are outliers in the sample.

B. X has a restricted range.

C. Participants were sampled at the two extremes of X (both extremely high and extremely low values of X).

D. There are ceiling effects

C. Participants were sampled at the two extremes of X (both extremely high and extremely low values of X).

20. If there is an interaction between two variables, A and B, this means that:

A. A has an effect on B .

B. The effect of A is the same at all levels of B.

C. The effect of A depends on the level of B.

D. Both A and B affect the dependent variable.

C. The effect of A depends on the level of B.

21. A researcher is studying the effects of growth mindset induction (participants randomly assigned to a growth mindset condition or a control condition) and math anxiety (participants divided into high or low math anxiety groups) on math test performance. She predicts that students with low math anxiety will perform better on the math test than students with high math anxiety. She also expects that participants in the growth mindset condition will perform better than the control group. Which of the following best describes the researcher’s predictions?

A. significant main effect of math anxiety, significant main effect of mindset induction, no interaction

B. significant interaction only with no main effects

C. significant main effect of math anxiety, with no main effect of mindset induction and no interaction

D. significant main effect of math anxiety, significant interaction.

A. significant main effect of math anxiety, significant main effect of mindset induction, no interaction