Deck 9: Non-Experimental Design I: Survey Methods

Question 1

Which of the following is true about the history of survey research?

A)surveys were not used until after World War I
B)the first survey research was Kinsey's pioneering sex surveys
C)Darwin and his cousin, Francis Galton, created the first surveys
D)the method of the "questionary" was vigorously promoted by William James

Answer 1

surveys were not used until after World War I

Question 2

What was the lesson derived from the example of the marketing campaigns for the Whopper and the Big Mac?

A)some surveys have self selection problems
B)be careful of writing survey items that have double-barreled questions
C)beware of leading questions
D)some items use words that are linguistically ambiguous

Answer 2

beware of leading questions

Question 3

When survey items are all worded favorably , the result can be a response bias called

A)the Hawthorne effect
B)a social desirability bias
C)response acquiescence
D)nonresponse bias

Answer 3

response acquiescence

Question 4

A major problem with phone surveying is

A)sampling; too many people have unlisted numbers
B)cost; face-to-face interviews are much cheaper
C)logistics; it is difficult to collect a lot of data per unit of time
D)selling disguised as surveying (sugging)

Question 5

All of the following are closed questions except

A)Is this a closed question?
B)Is this an open question?
C)How is a closed question different from an open question?
D)Is this a poor multiple choice question?

Question 6

Which of the following is true about return rates for mailed surveys?

A)anything less than 100% creates problems for interpretation
B)researchers are happy with return rates between 70% and 85%
C)as long as the total number of replies is large (>100), the rate is return is irrelevant
D)if less than 80%, representativeness is a serious problem

Question 7

What is the major strength of the interview format for conducting survey research?

A)it has fewer logistical problems than other methods
B)more in-depth information can be acquired than with other methods
C)unlike other methods, interviewer bias is seldom a problem
D)it is cheaper than other methods

Question 8

Recommendations about survey wording include all of the following except

A)negatively phrased questions are useful because they force respondent's to pay attention
B)always opt for simplicity over complexity when wording survey items
C)avoid the use of jargon that might not be understood by respondents (e.g., "summative")
D)when asking for opinion, use "do you support or oppose…" rather than "do you oppose…"

Question 9

What was wrong with the survey done by Literary Digest in 1936?

A)their sample did not accurately reflect the proportions of Republicans and Democrats
B)they only had 10% of their surveys returned
C)they used simple random sampling instead of the more sophisticated cluster sampling
D)they used cluster sampling, when stratified sampling was required

Question 10

Compared to written surveys, the interview format

A)reduces the problem of question ambiguity
B)is a lot cheaper
C)is immune from interviewer bias problems
D)has more of a problem with return rate

Question 11

Which of the following is a problem for written surveys?

A)question ambiguity
B)survey return rate
C)both alternatives a. and b.
D)none of the above

Question 12

Compared to experimental methods found previous chapters, survey research

A)does not have a problem with reactivity
B)is more likely to use convenience sampling
C)has a greater need for the sample to be representative
D)will tend to have a smaller sample size

Question 13

The best return rate will be for written surveys that are

A)surveying attitudes about sex rather than about politics
B)clearly unrelated to a sales pitch
C)filled with more open questions than closed questions
D)printed on colored paper than on white paper

Question 14

All of the following characterize interview surveys except

A)cost can be a problem
B)the interviewer can clarify ambiguous questions
C)getting a representative sample is easier than with the other methods
D)they might be limited to a small geographical area

Question 15

Hall developed the "questionary" in order to

A)discover the "contents of children's minds"
B)disprove Galton's ideas about the inheritance of intelligence
C)study facial expressions of emotion in different cultures
D)prove that growing up in a city was vastly superior to a country life

Question 16

When should DK ("don't know")alternatives be used?

A)when collecting demographic information
B)never-they will be overused by respondents
C)whenever you are surveying people with moderate to low IQs
D)when there is good reason to believe that some respondents will genuinely not know an answer

Question 17

When magazines report the results of reader surveys, what is the major weakness?

A)self selection
B)low number of returns
C)use of nonprobability rather than probability sampling techniques
D)they fail to survey the entire population

Question 18

When using a written survey,

A)return rates will be higher if the survey uses more closed questions than open questions
B)a return rate of 5% is adequate if the survey is well planned
C)return rates will be higher if the survey uses more open questions than closed questions
D)experimenter bias is never a problem

Question 19

Demographic information includes

A)how knowledgeable a person is about some topic
B)information about the income level of the respondent
C)someone's opinion about democracy (or related issues)
D)all of the above

Question 20

The Kinsey Report of sexual behavior is a classic example of

A)case study research
B)archival research
C)survey research using a face-to-face interview format
D)electronic surveying

Question 21

In order for a Pearson's r to reflect a meaningful relationship, which of the following must be true?

A)the relationship must be nonlinear
B)the variables must be measured on at least an ordinal scale
C)the range of scores must be restricted on at least one of the variables
D)the variables must be measured on either an interval or a ratio scale

Question 22

In order for Pearson's r to be calculated,

A)both variables must be measured on a ratio scale
B)the relationship must be a linear one
C)at least one of the variables must have a restricted range
D)the variables must be measured on at least an ordinal scale

Question 23

Which of the following would be most likely to produce a negative correlation?

A)the relationship between college grades and parents' IQ
B)the relationship between college grades and total TV watching time
C)the relationship between college grades and high school grades
D)the relationship between college grades and time spent in library

Question 24

Which of the following correlations shows the weakest relationship?

A)+)23
B)+)07
C)-)09
D)-)99

Question 25

Picture a regression line as a diagonal line running from the lower left to the top right on a scatterplot. For which of the following correlations would the points be closest to this regression line?

A)-)90
B)-)30
C)+)40
D)+)70

Question 26

In a negative correlation,

A)high scores on one variable are accompanied by high scores on the second variable
B)low scores on one variable are accompanied by low scores on the second variable
C)high scores on one variable are accompanied by low scores on the second variable
D)both alternatives a. and b.

Question 27

Which of the following correlations shows the strongest relationship?

A)-)72
B)+1.21
C)+)70
D)-)09

Question 28

On surveys, participants can try to present a positive image of themselves, a problem referred to as

A)experimenter bias
B)social desirability bias
C)question ambiguity
D)good subject bias

Question 29

In order for a Pearson's r to reflect a meaningful relationship, which of the following must be true?

A)the relationship must be linear
B)the relationship must be nonlinear
C)the relationship must be inverse
D)the relationship must be a positive one

Question 30

A regression line for a correlation of -1.00 would

A)be parallel to the X-axis
B)be parallel to the Y-axis
C)higher for low values on the X-axis than for high values on the X-axis
D)higher for high values on the X-axis than for low values on the X-axis

Question 31

Which of the following was true about the survey by Hartwig and Dunlosky (2012)on student study strategies?

A)it contained all open-ended questions
B)it was an interview format
C)it had a small sample size
D)it contained all closed questions

Question 32

Zezima (2010)reported that while nearly 100% of people reported that they wash their hands after using public restrooms, however actual observations of public hand washing behavior demonstrated much lower percentages. This illustrates

A)invasion of privacy
B)question ambiguity
C)biased sampling
D)a social desirability bias

Question 33

What were the results of the Hartwig and Dunlosky (2012)survey of student study strategies?

A)students who tested themselves also had higher GPAs
B)use of flashcards was strongly related to GPA
C)cramming led to better test performance than study that was distributed over time
D)the most often reported study strategy was rereading the material

Question 34

In a positive correlation,

A)high scores on one variable are accompanied by high scores on the second variable
B)low scores on one variable are accompanied by low scores on the second variable
C)high scores on one variable are accompanied by low scores on the second variable
D)both alternatives a. and b.

Question 35

Phone surveys

A)usually need to be brief and to include few if any open questions
B)usually have very high response rates
C)avoid problems with experimenter bias
D)are limited to cluster samples only

Question 36

In the box describing the case of the female journalist who was fired on the basis of a survey, which of the following occurred?

A)the survey was not actually completed; the data were manufactured
B)even though an elaborate (and appropriate)statistical analysis was done, the survey itself was hopelessly biased
C)there was an inappropriate statistical analysis
D)it a well-designed and fair survey, but used a biased sample

Question 37

A regression line for a correlation of +1.00 would

A)be parallel to the X-axis
B)be parallel to the Y-axis
C)higher for low values on the X-axis than for high values on the X-axis
D)higher for high values on the X-axis than for low values on the X-axis

Question 38

How much information is in a correlation coefficient?

A)the strength of a relationship
B)the direction of a relationship
C)both alternatives a. and b.
D)none of the above

Question 39

Nonresponse bias occurs when

A)the population from which the sample is taken is too small
B)response rate is lower than anticipated
C)those returning a survey differ in some important way from the initial sample
D)those responding to a survey fail to answer a significant number of questions on the survey

Question 40

All of the following are problems with online surveying except

A)cost
B)debriefing
C)self selection bias
D)some sampling bias (middle and upper incomes)

Question 41

If a researcher is trying to predict college grades, knowing high school grades, then

A)the predictor variable will be college grades
B)the criterion variable will be high school grades
C)the prediction will be poorer if only the top 20% of high school students are used than if they are all used
D)multiple regression rather than bivariate regression must be used

Question 42

Which of the following is true about a correlation of +.70 between variables A and B?

A)about half of the variability in A can be accounted for by variability in B
B)the scores for A and B will be about the same 70% of the time
C)there would be a greater ability to predict than for a correlation of -.70
D)A causes B about 70% of the time; B causes A about 30% of the time

Question 43

To predict college success from a combination of high school grades, weighted letters of recommendation, and SAT scores, an admissions department would probably use what procedure?

A)bivariate analysis
B)partial correlation
C)factor analysis
D)multiple regression

Question 44

Multiple regression involves

A)one bivariate procedure and two multivariate procedures
B)one criterion variable and two or more predictor variables
C)one predictor variable and two or more criterion variables
D)performing a bivariate regression procedure several times

Question 45

A study using a ________ helps to solve the third variable problem.

A)cross-lagged panel correlation
B)partial correlation
C)multiple regression
D)factor analysis

Question 46

Suppose you investigate the relationship between college grades and time spent on Facebook and discover that the correlation is -.71. What does this tell you?

A)longer times spent on Facebook are related to higher grades
B)longer times spent on Facebook are related to lower grades
C)to improve your grades, get off of Facebook
D)time spent on Facebook really doesn't have anything to do with grades

Question 47

Eron and his colleagues (1972)were able to show that TV viewing preferences were related to aggressiveness ten years later. What procedure did they use?

A)cross-lagged panel correlation
B)partial correlation
C)multiple regression
D)factor analysis

Question 48

A study using a ________ helps to solve the directionality problem.

A)cross-lagged panel correlation
B)partial correlation
C)multiple regression
D)factor analysis

Question 49

A regression equation is used for

A)making predictions about a second variable, given the value of the first variable
B)solving the directionality problem
C)solving the third variable problem
D)calculating the size of the correlation

Question 50

If there is a strong relationship between variable A and variable B, it could be that A is causing B to occur, but it could also be that B is causing A to occur. This is known as the

A)directionality problem
B)third variable problem
C)nonlinearity problem
D)range restriction problem

Question 51

Research using regression analyses is useful for

A)establishing causal relationships
B)making predictions
C)manipulating independent variables
D)factorial-type designs

Question 52

A study using a partial correlation helps to solve the _______ problem.

A)third variable
B)restriction of range
C)directionality
D)multiple regression

Question 53

A study using a cross-lagged panel correlation helps to solve the _______ problem.

A)third variable
B)restriction of range
C)directionality
D)multiple regression

Question 54

The coefficient of determination

A)will never be a negative number
B)solves the directionality problem
C)enables conclusions about cause and effect in correlational studies
D)is always closer to 1.00 than the value for Pearson's r

Question 55

A researcher believes that anxiety disorders in children can be predicted by a combination of several variables: shyness in preschool, mother dominance, and father absence. This study will use which procedure?

A)multiple regression
B)partial correlation
C)factor analysis
D)bivariate correlations

Question 56

Ed took a Civil Service Test and was subsequently hired by the Post Office, where he performed well. The Civil Service Test was used because

A)it had been shown to be an effective predictor variable
B)it had been useful as a criterion variable
C)scores on the test correlated perfectly with IQ
D)scores on it regressed to the mean

Question 57

A researcher discovers a correlation of +.87 between study time and GPA. What can be concluded?

A)if you study a lot, your GPA will go up
B)if you have a high GPA, you will be forced to study more to keep up
C)students who study a lot tend to have higher GPAs than those who don't study much
D)the underlying cause of the correlation is native intelligence

Question 58

A researcher discovers a strong correlation between the number of hours spent watching TV and the level of aggression shown at school. What can be concluded?

A)watching a lot of TV causes children to become aggressive
B)naturally aggressive children are led to watch a lot of TV
C)knowing TV-watching patterns enables a prediction of aggressiveness
D)because of the distraction of TV, those who watch a lot of it are not very aggressive

Question 59

In a study examining the relationship between A and B, the coefficient of determination

A)indicates how much variability in A can be accounted for by B
B)is found by squaring the value of Pearson's r
C)both alternatives a. and b.
D)none of the above

Question 60

If there is a strong correlation between variables A and B, what is the most likely explanation?

A)A is causing B
B)B is causing B
C)A and B are caused by C
D)could be any of the above - the correlation by itself does not allow one to decide

Question 61

A ______ variable is one that helps explain under what conditions a correlation between two variables exists.

A)predictor
B)criterion
C)mediating
D)moderating

Question 62

By using a __________ procedure, Eron and his colleagues (1972)were able to rule out several potential third variables that might have affected the relationship they found between preference for TV violence and aggressiveness.

A)cross-lagged panel correlation
B)partial correlation
C)multiple regression
D)factor analysis

Question 63

A ______ variable is one that helps explain why a correlation between two variables exists.

A)predictor
B)criterion
C)mediating
D)moderating

Question 64

In the textbook example, impulse control helps explain why a correlation between alcohol use and engaging in risky sex. In this case, impulse control is a ______ variable.

A)predictor
B)criterion
C)mediating
D)moderating

Question 65

Suppose there is a high positive correlation between reading speed and reading comprehension. You suspect that IQ is a potential third variable. You complete a partial correlation procedure and, with IQ controlled statistically, the high correlation between speed and comprehension remains. What do you conclude?

A)IQ is an important third variable affecting the correlation
B)IQ can be ruled out as a third variable underlying the correlation
C)IQ must be the cause of the correlation
D)not enough information to decide

Question 66

In the textbook example, the type of living conditions helps explain under what conditions a correlation between alcohol use and engaging in risky sex. In this case, living condtion is a ______ variable.

A)predictor
B)criterion
C)mediating
D)moderating

Question 67

Suppose there is a high positive correlation between reading speed and reading comprehension. You suspect that IQ is a potential third variable. You complete a partial correlation procedure and, with IQ controlled statistically, the high correlation between speed and comprehension virtually disappears. What do you conclude?

A)IQ is an important third variable affecting the correlation
B)IQ can be ruled out as a third variable underlying the correlation
C)IQ must be the cause of the correlation
D)not enough information to decide