·         SAMPLE MULTIPLE CHOICE - EXAM 2

• This practice quiz is designed to give you an idea of what type of multiple choice questions may be asked on the exam. 

        It is NOT intended to act as a complete review of the material.  Answers appear at the end of the questions.

·         The exam will also have short answer items similar to Interpreting Basic Statistics exercises at the end of course handout packets.

·         There will NOT be calculation problems on the exam.

Answers appear at the end

1.  Which of the following scatterplots or scatter diagrams represent a perfect, positive relationship?

       a.                                                 b.

      c.                                                 d.

2.  If there is a strong positive correlation between two variables this means that:

a.  a person’s raw scores on the variables will be very similar

b.  a person’s z-scores on the variable will be very similar

c.  a person’s z-scores on the variables will be the opposite (e.g., +2.34 and – 2.34)

d.  there’ll be no relationship between the z-scores on the two variables.

3  Which of the following might explain a low Pearson correlation?

       a.  Little or no relationship exists between the two variables.

       b.  The variables may be related in a nonlinear fashion.

       c.  The range of values on one or both of the variables may be restricted.

       d.  All of the above.

4.  Professor Hope has participants fill-out two questionnaires.  One  questionnaire measures
     how much people trust others and higher scores reflect more trust. The second questionnaire
     assesses anger expression.  Higher scores indicate individuals  who often express their anger
     (i.e., lose their tempers). There is a negative correlation between trust and anger expression. 
    This means that

      a.  people with high trust scores tend to have high anger expression scores.

      b.  there is no relationship between trust and anger because the correlation is negative.

      c.  people with low trust tend to lose their temper more than people with high trust.

      d.  there is no relationship between trust and anger because everyone loses their

           temper at least once in a while.

5.  To investigate the relationship between aggressiveness and alcohol consumption

       Dr. Bud Weiser recruits subjects from the violent ward of a large drug rehabilitation

       center.  An low Pearson correlation between the subjects' record of aggressive behavior

       and past alcohol consumption was found.  Dr. Weiser concludes that there is no relationship

       between the two variables.  If you were reviewing this research, which of the following

       responses would be appropriate?

       a.  "Dr. Weiser, these variables are obviously measured on a ordinal scale.

             You should have used a Spearman r.
       b.  "Dr. Weiser, have you considered the possibility that the variables might

             be negatively correlated?"

       c.  "Dr. Weiser,  because of your sample, you may only have subjects on the high end of

             both aggressiveness and drinking.  Thus, the range of the two variables might be

             truncated. This restriction of range might explain the low correlation."

       d.  "Dr. Weiser, I commend you for your excellent understanding of  correlation.  Please
             accept this million dollar grant from the Beer Union Research Board (BURB)."

6 .  The more minutes a student sleeps in class, the fewer notes the students tends to write.  This describes what sort of relationship?

a.  negative, because as one variable increases the other decreases.

b.  positive, because as one variable increases the other increases

c.  negative because it’s bad to miss notes

d.  postive because the student is catching up on his or her sleep

7. If a student sleeps more than average in class, what would you predict in terms of notes
     taken? (based on correlation described in number 6)

a.  the student would take above average number of notes

b.  the student would take below average number of notes

c.  the student would be taking the average number of notes

d.  one can’t predict based on a correlation. 

8.  To conduct Pearson correlation and regression which of the following has to be true:

a.  variables must be at least ordinal, the relationship has to be linear

b.  you need at least two variables and they must be psychological in nature

c.  variables must be internal or ratio, the relationship must be linear, and the
     distributions of both variables should be skewed.

d.  variables must be interval or ratio, the relationship must be linear, and the distributions
     of both variables should be approximately normal

9.  If the correlation between two variables is .50 than the percent of variation in one
     variable accounted for by the other is

a. 50%

b. 25%

c. 10%

d. 1%

10.  Another name for r² is

a.  coefficient of determination

b.  coefficient of explanation

c.  coefficient of regression

d. coefficient of correlation

11-15  These questions all refer to the correlation matrix presented below.

  You’ve calculated the correlations between measures of early childhood trauma, family
   history of depression and current depression scores.

11.  The correlation between family history and current depression is:

a.  .28 and it is significant

b.  .31 and it is significant

c. .45 and it is significant

d. .45 but it is not significant

12.  If you were going to conduct a regression analysis using family history and current depression which variable would be the predictor and which would be the criterion?

a.  predictor:  family history,           criterion:  current depression

b.  predictor:  current depression,  criterion: family history

c.  predictor: childhood trauma,     criterion: current depression

d.  predictor: depression                criterion: childhood trauma

13.  Which variable in your study is likely to be the best predictor of depression?

a.  Best predictor of depression would be current depression.

b.  Childhood trauma would be the best predictor

c.  Family history would be the best predictor

d.  None of the variables would be useful as predictors.

14.  If you used both childhood trauma and family history to predict depression you would be
       using

a.  coefficient of determination

b.  curvilinear regression

c.  bivariate correlation

d.  multiple correlation

15.  If you used both childhood trauma and family history to predict depression you would expect
       R to

a.  equal .31 + .45 the sum of the two correlations between the predictors and the criterion

b.  likely be less than sum of the correlations in a., because the predictors are correlated with
     each other, so there is some overlap

c.  likely be more than the sum of the correlations in a., because the point of using more than one
     predictor is to improve prediction

d.  a waste of time.

16.  In the above figure, what represents errors in prediction?

a.  a – the dot on the regression line

b.  b – the deviation of actual score from predicted score

c.  c. – the deviation of predicted score from the mean

d.  a b and c

17.  The standard error of estimate is

a.  the standardized regression coefficient

b.  the unstandardized regression coefficient

c.  the coefficient of determination

d.  the average errors in prediction

18 - 20.  Below are the results of a SPSS regression analysis using mechanical aptitude to predict work performance.

Use the information to answer the questions that follow.

18.  In the equation , b is

a.  the y-intercept or 54.245

b.  the significance level  or .002

c.  the slope of the regression line or .648

d.  the standardized regression coefficient or .436

19.  a from the equation is

 

a.  the y-intercept or 54.245

b.  the significance level  or .002

c.  the slope of the regression line or .648

d.  the standardized regression coefficient or .436

20.  The correlation, r, between mechanical aptitude and work performance is the same as:

a.  a.  the y-intercept or 54.245

b.  the significance level  or .002

c.  the slope of the regression line or .648

d.  the standardized regression coefficient or .436

21.  In a standard normal distribution what percentage of scores are above and below one standard deviation of the mean?

a. 100%

b. 68%

c. 50%

d. 33 %

22.If a score is at the mean of a distribution, the z-score will be

a.  0

b. +1

c.  -1

d.  10

23.  Many variables of interest in psychology form distributions that are _____________ in shape.

a.  rectangular

b.  curvilinear

c.  bimodal

d.  normal

24.  A percentile rank is

a.  the percent of items a person correctly answers on a test.

b.  another name for an ordinal variable

c.  the percent of scores that fall at or below a certain score

d. The percent of Dr. Boland’s salary you must use to bribe her to pass this class.

25.  A parameter is to a statistic as a ______________ is to a _________________.

a.  population is to a sample

b.  regression is to acorrelation

c.  meanis to a standard deviation

d. random sample is to a nonrandom sample