Deck 17: Correlation

Question 1

Which of the following statements is correct regarding the table below?

A)There is a zero correlation between group and outcome.
B)There is a positive correlation between group and outcome.
C)There is a negative correlation between group and outcome..
D)There is a curvilinear correlation between group and outcome.

Answer 1

There is a zero correlation between group and outcome.

Question 2

A strong correlation between an independent and dependent variable indicates that the independent variable is a cause of the dependent variable.

Answer 2

False

Question 3

In a study assessing whether experimental group participants spend fewer days incarcerated than do control group participants, the appropriate correlation coefficient to calculate is:

A)Pearson's r
B)Spearman's rho
C)phi
D)eta

Answer 3

eta

Question 4

Pearson's r is best used with variables at the nominal level of measurement.

Question 5

When there are many ties in the rank ordering of two ordinal variables, it is appropriate to calculate Kendall's tau-b.

Question 6

In a study assessing the relationship between the number of close friends who drink alcohol regularly and the number of times a person in treatment for alcohol abuse drank alcohol after completing treatment, the appropriate correlation coefficient to calculate is:

A)Pearson's r
B)Spearman's rho
C)phi
D)eta

Question 7

A correlation of +.60 between quality of therapist-client relationship and the client's level of self-esteem means that improving the therapist-client relationship is likely to cause the client's self-esteem to improve.

Question 8

When a scatterplot is clearly curvilinear:

A)The calculated correlation coefficient for all the data in the scatterplot may be low.
B)The calculated correlation coefficient for all the data in the scatterplot will be significant.
C)As one variable goes up, the other consistently goes down.
D)None of these.

Question 9

Which of the following statements is correct about keeping or eliminating outliers when calculating the correlation coefficient?

A)They should always be eliminated.
B)They should never be eliminated.
C)The decision should be based on the option that best supports the hypothesis.
D)None of these.

Question 10

When a correlation coefficient is near perfect:

A)One variable is a cause of the other.
B)The same variable may have been measured in two different ways.
C)The relationship is positive.
D)All of these.

Question 11

Which of the following statements is correct regarding obtaining a correlation coefficient of zero?

A)Because there is no correlation, there is no need to examine the scatterplot.
B)If the scatterplot is curvilinear, that would further support the notion that there is no meaningful relationship between the variables.
C)If the scatterplot is curvilinear, that might indicate that there is a relationship, but that its nature changes at different levels of the variables.
D)None of these.

Question 12

Suppose as the number of abusive incidents increases, the severity of traumatic symptoms increases.This would be:

A)A negative correlation.
B)A positive correlation.
C)An inverse correlation.
D)Both a negative and an inverse correlation.

Question 13

A correlation of +.95:

A)Is stronger than a correlation of -.95.
B)Signifies an important relationship between two different variables.
C)Both of the above.
D)None of these.

Question 14

In a study assessing the relationship between two variables at the ordinal level of measurement, an appropriate correlation coefficient to calculate is

A)Pearson's r
B)Spearman's rho
C)phi
D)eta

Question 15

In a study assessing whether gender is related to whether or not a student drops out of school after participating in a dropout prevention program, the appropriate correlation coefficient to calculate is:

A)Pearson's r
B)Spearman's rho
C)phi
D)eta

Question 16

A study finds a statistically significant correlation showing that the more treatment appointments missed by spouse abusers, the more abusive incidents.This means that:

A)Treatment sessions lower the number of abusive incidents.
B)Missing treatment sessions increases the risk of abusive incidents.
C)There is a negative correlation between the number of treatment appointments missed by spouse abusers and the number of abusive incidents.
D)None of these

Question 17

A curvilinear scatterplot will usually be associated with a strong overall correlation coefficient.

Question 18

A correlation of -.40 is stronger than a correlation of +.30.

Question 19

A correlation of -.60 between quality of therapist-client relationship and number of antisocial behaviors means that the better the therapist-client relationship, the fewer the antisocial behaviors.

Question 20

When two variables are at the ratio level of measurement, which of the following statements is true about calculating Pearson's r?

A)We can disregard whether the degrees of variance in the two variables are similar.
B)The degree of linearity in the scatterplot won't matter.
C)The presence of outliers can influence the calculation and interpretation of the correlation coefficient.
D)All of these.

Question 21

Just as in other tests of significance, sampling error when testing the statistical significance of correlation coefficients involves the use of theoretical sampling distributions with an assumed mean correlation of zero.

Question 22

When comparing the means of two or more categories of a nominal variable, we can use the eta statistic to depict the strength of the relationship.

Question 23

Phi should be used to calculate the correlation between variables at the ratio level of measurement.