Deck 11: The Hypothesis of Association: Correlation

Question 1

When two variables, A and B, are strongly correlated, it means that

A) A has caused B
B) B has caused A
C) some third variable has caused both A and B to occur together
D) none of these

Answer 1

D

Question 2

Of the following correlation coefficients, which expresses the strongest association?

A) .10
B) 0
C) -.50
D) -1.00

Answer 2

D

Question 3

When two sets of measures have been found to be totally independent of each other, then the

A) correlation between them must be negative
B) correlation between them must be positive
C) correlation between them must be zero
D) no attempt to calculate the correlation should be made

Answer 3

C

Question 4

The correlation coefficient, in and of itself,

A) never proves causation
B) never rules out causation
C) allows for better-than-chance predictions
D) all of these

Question 5

The originator of the product-moment correlation technique was

A) Karl Gauss
B) Blaise Pascal
C) Karl Pearson
D) Alexandra McKenna

Question 6

Among elementary school children, the correlation between age and height would most probably be

A) positive
B) negative
C) zero
D) non-linear

Question 7

The correlation between parental income and the amount of financial aid awarded by the college to students should most probably be

A) positive
B) negative
C) zero
D) none of these, since income is a nominal measure

Question 8

When the array of points on a scatter plot slopes from lower left to upper right, then

A) the correlation is positive
B) the correlation is negative
C) there is no correlation
D) no predictions can be made

Question 9

A scatter plot on which the array of points goes from upper left to lower right, indicates

A) a positive correlation
B) a negative correlation
C) a zero correlation
D) all of these, depending on the strength of the correlation

Question 10

Each data point on a scatter plot represents

A) the frequency of occurrence
B) a pair of scores
C) a score on one measurement
D) none of these

Question 11

Negative correlations

A) may never be significant
B) may never be used for prediction
C) may never generalize to the population
D) none of these

Question 12

On a scatter plot, the ordinate is used to represent

A) the scores on the Y measure
B) the scores on the X measure
C) the frequency of occurrence
D) the product moment

Question 13

The most common error in interpreting correlational research studies is in

A) assuming that the data have been incorrectly reported
B) assuming that the appropriate statistical analyses have been used
C) assuming that a causal factor has been isolated
D) all of these

Question 14

The more a correlation deviates from zero,

A) the more it is certain to be positive in sign
B) the more it is certain to be negative in sign
C) the more it is certain that the measured variables are independent of each other
D) the better its predictive accuracy

Question 15

The major reason for using the correlational techniques is to

A) make cause-and-effect statements
B) make better-than-chance predictions
C) identify a specific percentile
D) establish a significant difference

Question 16

Of the following correlation coefficients, which expresses the weakest association?

A) .10
B) 0
C) -.50
D) -1.00

Question 17

A "significant" correlation is one which

A) is especially meaningful
B) does not deviate from zero
C) indicates a valid difference
D) can be generalized to the population

Question 18

When variable A correlates highly with B, then

A) A has caused B
B) B has caused A
C) variable B correlates highly with A
D) none of these

Question 19

A maximum linear correlation is expressed as

A) 100%
B) 99
C) +10.00
D) +1.00

Question 20

In order to establish the significance of a correlation, one must know the value of the correlation coefficient and also

A) the number of paired scores
B) whether it relates to other measures
C) Whether it specifies the direction of the association
D) the sign of the correlation

Question 21

If the researcher uses these data to predict GPA, then the scatter plot's ordinate would represent

A) TV viewing time
B) GPA
C) the correlation
D) frequency of occurrence

Question 22

The reason for the typical shape assumed by the array of points in a scatter plot, is due to

A) the central tendency expected on the two distributions
B) the wide variability expected on the two distributions
C) the significant difference between the means of the two distributions
D) the significant difference between the variabilities of the two

Question 23

When high scores on the first variable associate with high scores on the second, and low scores on the first associate with low scores on the second, then the correlation is

A) non-existent
B) positive
C) negative
D) zero

Question 24

The correlation between the two variables is generally

A) positive
B) negative
C) zero
D) positive, negative and sometimes even zero

Question 25

If these data were to be placed on a scatter plot, the general slope of the array of data points would be

A) straight across
B) upper left to lower right
C) lower left to upper right
D) U shaped

Question 26

If the correlation between TV viewing and GPA were significant, then the only valid conclusion one can draw from these data, is that

A) TV viewing lowers GPA
B) TV viewing increases GPA
C) TV viewing is independent of GPA
D) A knowledge of TV viewing time allows for the prediction of GPA

Question 27

When a correlation is not significant, it means that

A) There is a failure to reject Ho
B) the correlation can't be generalized to the population
C) the correlation is negative
D) a and b, but not c

Question 28

The more any correlation value deviates from zero

A) the more likely it is to be significant
B) the stronger the association between the two variables
C) the greater its predictive efficiency
D) all of these

Question 29

Regardless of the degrees of freedom, any Pearson r of .90 or higher must be

A) significant
B) non-significant
C) more than significant
D) none of these, since without the degrees of freedom significance cannot be

Question 30

For the Pearson r, the alternative hypothesis states that

A) there is a significant correlation in the population
B) there is no correlation in the sample
C) there is no correlation in the population
D) the sample no longer represents the population

Question 31

If the researcher uses these data to predict GPA, then the scatter plot's abscissa would represent

A) TV viewing time
B) GPA
C) the correlation
D) frequency of occurrence

Question 32

For the Pearson r, when Ho is rejected, it means that

A) the correlation is positive
B) the correlation is negative
C) the correlation is significant
D) none of these

Question 33

In doing a Pearson r on 20 pairs of scores, degrees of freedom would equal

A) 20
B) 18
C) 40
D) 38

Question 34

When the knowledge of one event allows for the prediction of another event, then the two events are

A) significantly different
B) causally related
C) correlated
D) due to chance

Question 35

Each point on the scatter plot portraying these data would indicate

A) performance on GPA only
B) hours of TV viewing only
C) both performance on GPA and hours of TV viewing
D) none of these

Question 36

When a relationship is inverse it means that the correlation must be

A) non-significant
B) non-linear
C) significant
D) negative

Question 37

The Pearson r may be used to test the hypothesis of

A) difference
B) association
C) central limit
D) all of these

Question 38

In general, the array of points on a scatter plot tends to form the shape of

A) a circle
B) a rectangle
C) a square
D) an oval

Question 39

For the Pearson r, the null hypothesis states that

A) there is a significant correlation in the population
B) there is no correlation in the sample
C) there is no correlation in the population
D) the sample no longer represents the population

Question 40

When a correlation is discovered between two variables, A and B, it is possible, though not certain that

A) A has caused B
B) B has caused A
C) X has caused A plus B
D) all of these

Question 41

When the underlying distribution of interval scores is violently skewed, the appropriate correlational technique should be

A) the Pearson r
B) the t test
C) the Spearman r
D) none of these, since correlations cannot be computed for skewed

Question 42

The Pearson r may only be used when the data form is at least

A) nominal
B) ordinal
C) interval
D) ratio

Question 43

Any Pearson r which is significant at an alpha of .01, must also

A) be significant at .001
B) be significant at .05
C) be significant at any alpha level
D) be positive

Question 44

For the Spearman r, in order to establish a significant correlation in the population

A) There is a failure to reject Ho
B) Ho must be rejected
C) Ho must be equal to alpha
D) none of these, since the Spearman r cannot be used to infer population

Question 45

When Ho is rejected for the Pearson r, then

A) the correlation is assumed not to be significant
B) the correlation is assumed to be significant
C) alpha error is eliminated
D) none of these, since when Ho is rejected no statement of significance is

Question 46

A Pearson r value of -.20

A) can never be significant
B) must always be significant
C) is significant only at p of .01
D) none of these, since the number of pairs of scores was not given

Question 47

To calculate the degrees of freedom for the Spearman r, one must

A) subtract 2 from the number scores
B) subtract 2 from the number of paired scores
C) add the number of scores to the constant 2
D) none of these

Question 48

The value of the Pearson r indicates

A) the strength of the linear relationship between two sets of interval measures
B) the strength of the non-linear relationship between two sets of interval measures
C) the strength of the linear relationship between two sets of nominal measures
D) all of these

Question 49

With 5 degrees of freedom, the critical tabled value of the Pearson r is .75 for an .05 alpha level. If one were to obtain a Pearson r value of -.76, then

A) There is a failure to reject Ho
B) Ho would be rejected
C) the alpha level would become negative
D) none of these, since a negative Pearson r cannot be evaluated

Question 50

In doing a Pearson r with 10 degrees of freedom, there must have been

A) 10 separate scores
B) 22 separate scores
C) 10 pairs of scores
D) 12 pairs of scores

Question 51

The proportion of the variability in the Y distribution which is accounted for by the variability in the X distribution, is determined by

A) the Spearman r
B) the null hypothesis
C) the coefficient of contingency
D) the coefficient of determination

Question 52

A Pearson r of .70 yields a coefficient-of-determination accuracy estimate of

A) 70%
B) 35%
C) 49%
D) .05%

Question 53

The Pearson r assumes that

A) the samples are randomly selected
B) the association between the variables is linear
C) the measures do not depart from normality
D) all of these

Question 54

When one set of measures is in interval form, and the other in ordinal, then the appropriate correlational technique should be

A) the Pearson r
B) the t test
C) the Spearman r
D) the coefficient of determination

Question 55

When both sets of measures are in ordinal form, the appropriate correlational technique should be

A) the Pearson r
B) the t test
C) the Spearman r
D) the coefficient of determination

Question 56

For the Spearman r, the null hypothesis states that

A) there is no correlation in the sample
B) there is no correlation in the population
C) the correlation is significant
D) the alpha level is zero

Question 57

Homoscedasticity refers to the

A) central tendency within one of the two distributions
B) variability within the two distributions
C) number of degrees of freedom
D) all of these

Question 58

For the Pearson r, as the sample size increases, so also will

A) the degrees of freedom
B) the alpha level
C) the number of measures taken on each subject
D) none of these, since sample size is not relevant to the Pearson r

Question 59

For the Pearson r, rejecting Ho with an alpha of .05, means that

A) the correlation is not significant
B) there is no possibility that Ho should have been accepted
C) the probability is .95 that Ho should have been accepted
D) the probability is .05 that Ho should have been accepted

Question 60

For a given number of degrees of freedom, when the obtained r is equal to the critical tabled value of r, then

A) There is a failure to reject Ho
B) Ho is rejected
C) Ho is re-written as Ha
D) Ho cannot be evaluated

Question 61

A significant Pearson r

A) may be generalized to any population
B) may only be generalized to the population from which the sample measures were selected
C) may only be generalized to finite populations
D) may never be generalized to a population

Question 62

In order to rank-order the following IQ scores, 120, 110, 110, 105, the appropriate ranks would be

A) 1, 2, 3, 4
B) 1, 2, 4, 5
C) 1, 1.5, 1.5, 2, 3
D) 1, 2.5, 2.5, 4

Question 63

A Pearson r value of .50 or higher must be significant, regardless of the sample size.

Question 64

When portraying a correlation on a scatter plot, the array of data points usually forms an elliptical shape.

Question 65

The Spearman r may be appropriately used whenever

A) both sets of measure are in ordinal form
B) one set of measures is in interval form, and the other in ordinal form
C) both sets of measures are in interval form, but the distributions are badly skewed
D) all of these

Question 66

In order to make better-than-chance predictions, the Pearson r must be

A) positive
B) greater than .70
C) greater than .90
D) significant

Question 67

Whenever a correlation coefficient deviates significantly from zero, then

A) it may or may not be used for predictions, depending on its size
B) it may be used to establish totally accurate predictions
C) it may not be used for predictions
D) it may be used for establishing better-than-chance predictions, regardless

Question 68

On a scatter plot, the ordinate should always displays the measures from the Y distribution.

Question 69

A significant correlation of -.40

A) cannot be used for prediction
B) predicts with an accuracy of 40%
C) predicts more accurately than a significant correlation of +.35
D) predicts with less accuracy than a positive correlation of any size

Question 70

Whenever a correlation is significant, then a cause-and-effect relationship must be present.

Question 71

Whenever ordinal data are converted into interval data,

A) the tied ordinal scores must be averaged, and assigned the same interval value
B) the tied ordinal scores must be eliminated from the analysis
C) the tied ordinal scores must be assigned different interval values
D) none of these, since ordinal data should not be converted into interval

Question 72

Whenever interval data are converted into ordinal data,

A) the tied interval scores must be averaged, and assigned the same rank value
B) the tied interval scores must be eliminated from the analysis
C) the tied interval scores must be assigned different rank values
D) none of these, since interval data should not be converted into ordinal

Question 73

Significant correlations always predict with an accuracy which is better- than-chance.

Question 74

Whenever measures are in interval form, regardless of the shape of the distribution, the Pearson r may be used.

Question 75

For the Spearman r, the degrees of freedom are established in the same way as for the Pearson r.

Question 76

If the correlation between X and Y is zero, and the correlation between X and Z is zero, then the correlation between Y and Z should be zero.

Question 77

The Spearman r is to ordinal data as the Pearson r is to interval data.

Question 78

For the Pearson r, a non-significant correlation

A) should not be used for predictions
B) cannot be generalized to the population
C) is simply due to chance
D) all of these

Question 79

The higher the value of the Pearson r, the lower the value of the coefficient of determination.

Question 80

For the Pearson r, the sample size, not the value of r, determines the number of degrees of freedom.