Deck 7: Introduction to Descriptive Methods

Question 1

A correlation between two events allows us to say

A) that one event caused another
B) that two events are related
C) no relationship exists between the two events
D) the events are highly likely to occur together

Answer 1

that two events are related

Question 2

Correlation does not imply

A) description
B) causation
C) prediction
D) relationship

Answer 2

causation

Question 3

_______ studies allow us to attribute causality to a relationship, whereas ______ studies allow us to determine the existence of a relationship.

A) experimental; correlational
B) statistical; experimental
C) correlational; statistical
D) correlational; experimental

Answer 3

experimental; correlational

Question 4

Which of the following goals of science is the first step to all of the remaining goals?

A) understanding
B) explanation
C) prediction
D) description

Question 5

Descriptive research can do which of the following?

A) stimulate causal hypotheses
B) increase our ability to predict behavior
C) establish causality
D) a and b

Question 6

Which of the following is NOT an important goal of science?

A) description
B) prediction
C) explanation
D) none of the above; all are important goals

Question 7

Which of the following are difficult to do without using scientific methods?

A) accurately measuring psychological variables
B) determining the extent to which variables are related
C) accurately generalizing from a sample to a population
D) all of the above

Question 8

If we find a positive correlation between going to therapy and mental health, we know that

A) therapy increases mental health.
B) being well-adjusted causes one to go to therapy.
C) more affluent people go to therapy
D) people in therapy are in better mental health than those who do not go to therapy.

Question 9

Illusory correlation refers to a

A) correlation that is not statistically significant
B) correlation that exceeds the value of +1.00
C) cluster of points on a scatterplot that suggests a correlation between two variables
D) perception of a relationship between two variables when none actually exists

Question 10

In looking at two variables, people may

A) see a relationship even though none exists
B) see a negative correlation when a positive correlation exists
C) see a positive correlation when a negative correlation exists
D) all of the above

Question 11

In generalizing from their experience to the world as a whole,

A) people have trouble because their experience is only a small sample of experience
B) people have trouble because their experience is not a random sample of experience
C) people have trouble because they can't discount the role of coincidence
D) all of the above

Question 12

If you do ex post facto research, your data will probably have which of the following advantages over collected and previously tabulated archival research?

A) higher internal validity
B) higher external validity
C) higher construct validity
D) a and b

Question 13

Ex post facto research is weakest in terms of:

A) internal validity
B) external validity
C) construct validity
D) b and c

Question 14

Archival research is considered to be a nonreactive method of data collection because

A) the compilers of the archives will be deceased at the time of the archival analysis
B) most often the records were kept for reasons unrelated to behavioral research
C) archival records may be assumed to be error free
D) studying public records never poses questions of ethics or values

Question 15

Which of the following is least likely to be a weakness of collected and tabulated archival data?

A) Data may be about groups (aggregated data) rather than data about individuals.
B) Data may lack construct validity.
C) Data may lack external validity.
D) Data may be tainted by the instrumentation bias.

Question 16

Trying to categorize a wide range of free responses in an objective manner, such as looking for the amount of sexism in writtendocuments is referred to as

A) the archival method
B) narrative analysis
C) ex post facto research
D) content analysis

Question 17

Which of the following is true of content analysis?

A) it can be done by computer
B) it is perfectly objective
C) when it is objective, it is valid
D) good content analysis helps a study's internal validity

Question 18

Which of the following is NOT good advice for doing a content analysis?

A) before developing a content analysis scheme, do a mini-study to determine the type of material you will be coding
B) review the research literature to see if there are pre-existing coding schemes that you could modify
C) provide raters with examples of behavior that would fall into each category
D) do not try to control raters by training them or restricting their freedom to make subjective judgments

Question 19

The construct validity of archival data that is already coded (e.g., government data) is LEAST likely to be weak because of

A) reactivity
B) instrumentation bias
C) poor match between the measure and the construct
D) both a and b

Question 20

Content analyzed archival data, relative to archival data coded by others, should be less vulnerable to

A) reactivity
B) instrumentation bias
C) poor match between the measure and the construct
D) both b and c

Question 21

With observation, you run into the problem of

A) the observer having an effect on the observed
B) objectively coding your data
C) both a and b
D) neither a nor b; these are not problems in observational research

Question 22

A researcher spends several weeks suspended underwater so that he can study shark behavior. Initially the sharks attack the cage, but later the sharks ignore it. This is an example of

A) a true experiment
B) a quasi-experiment
C) participant observation
D) naturalistic observation

Question 23

If you enter several bars and lounges over a period of weeks, and secretly observe and take notes on male-female interactions, you are using the ________ method of research.

A) archival
B) experimental
C) ex post facto
D) naturalistic observation

Question 24

With testing, ____ is usually less of a problem than with other descriptive methods.

A) construct validity
B) internal validity
C) external validity
D) a and b

Question 25

Reactivity is least likely to be a problem in _____ research.

A) ex post facto
B) archival
C) observational
D) testing

Question 26

Objectivity is most likely to be a problem in _____ research.

A) ex post facto
B) archival
C) observational
D) testing

Question 27

Scatterplots

A) depict relationships as a cluster of dots
B) provide a preview of the fully measured relationship
C) show any pronounced curvilinearity in the data
D) both a and c

Question 28

A dot cluster that tilts from upper left to lower right reflects

A) little or no correlation
B) a positive correlation
C) a negative correlation
D) a curvilinear correlation

Question 29

Which scatterplot suggests that there is no relationship between variables X and Y?

A)
B)
C)
D)

Question 30

Which scatterplot indicates a positive correlation between variables X and Y?

A)
B)
C)
D)

Question 31

A large and negative correlation coefficient means that

A) high scores on one variable are associated with high scores on the other
B) low scores on one variable are associated with low scores on the other
C) high scores on one variable are associated with low scores on the other
D) the variables are not related

Question 32

Without doing any calculations, describe the relationship for the following pairs of observations.
$\begin{array}{cc}\mathrm{X}&\mathrm{Y} \\\hline6&118\\12 & 109 \\3 & 144 \\10 & 111 \\1 & 142\end{array}$

A) positive
B) negative
C) little or no relationship
D) curvilinear

Question 33

Pearson's r can range from

A) .00 to +1.00
B) -2.00 to +2.00
C) -.50 to +1.00
D) -1.00 to +1.00

Question 34

If you measured the relationship between weight and intelligence, and found the correlation coefficient to be equal to +2.47, you would conclude that

A) the higher the weight, the greater the intelligence
B) the lower the weight, the greater the intelligence
C) you must have a bad sample; weight and intelligence cannot be correlated
D) you made an arithmetic error

Question 35

If there is no relationship between two variables, the value of the correlation coefficient should be closest to

A) -1.00
B) 0.00
C) .50
D) +1.00

Question 36

If two variables are perfectly correlated, the correlation coefficient will have a value of

A) +1.00
B) -1.00
C) 0.00
D) +1.00 or -1.00

Question 37

A correlation coefficient of -.87 means that there is

A) a strong but not perfect relationship
B) a perfect relationship
C) a very weak relationship
D) no relationship

Question 38

Which of the following correlation coefficients indicates the strongest relationship between two variables?

A) +.42
B) +.89
C) +1.26
D) -.91

Question 39

Which of the following correlations represents the strongest relationship?

A) +0.90
B) +0.45
C) -0.37
D) -0.93

Question 40

Which scatterplot demonstrates the strongest relationship between variables X and Y?

A)
B)
C)
D)

Question 41

The strength of the relationship between two variables is best described by the size of the

A) observational coefficient
B) correlational coefficient
C) experimental coefficient
D) coefficient of determination

Question 42

When the correlation coefficient is squared, the resulting number supplies

A) a direct measure of the strength of the relationship
B) an indirect measure of the strength of the relationship
C) information about the direction of the relationship
D) all of the above

Question 43

A correlation between two hypothetical measures is +.80. What percentage of the variability in one measure is accounted for by its relationship with the other measure?

A) 80%
B) 64%
C) 36%
D) 89%

Question 44

A correlation of -0.6 indicates that _______ of the variation in Y is explained by X.

A) 6%
B) 12%
C) 36%
D) 60%

Question 45

If the correlation between X and Y is -.70, then the percentage ofvariation in Y accounted for by variation in X is

A) 7
B) 35
C) 49
D) 70

Question 46

If the correlation between X and Y is -.50, then the coefficient of determination is

A) .50
B) -.50
C) .25
D) .10

Question 47

Find the median for the following observations 6,4,8,1,2.

A) 3
B) 4
C) 5
D) 8

Question 48

Using inferential statistics on correlational data

A) is impossible
B) is totally unnecessary
C) allows you to make causal inferences from a random sample
D) allows you to generalize from a random sample to its population

Question 49

Strictly speaking, you should use inferential statistics on correlational data

A) never
B) only when you are interested in inferring causality
C) only when you have a random sample of a population
D) always, as long as your data is valid

Question 50

Using median splits so that one can do a t test on correlational data is:

A) a common practice
B) less powerful than testing the significance of the correlation coefficient
C) better than testing the significance of the correlation coefficient because it allows the investigator to determine if a variable caused a certain effect
D) both a and b

Question 51

If your data is nominal rather than interval, conventional wisdom suggests you should use

A) ANOVAs
B) Chi-square tests
C) median splits
D) Pearson rs

Question 52

. Statistically significant results mean that:

A) the results are important, provided the data are a random sample of some population
B) a causal relationship exists in the sample, provided the data are a random sample of some population
C) a causal relationship exists in the population
D) the results would generalize to the population, provided the data are a random sample of some population

Question 53

With a restricted range, the researcher

A) may obtain a very low correlation
B) may not have a broad range of scores
C) may mistakenly assume no relationship among variables
D) all of the above

Question 54

Null results may be due to all of the following EXCEPT:

A) conducting too many statistical tests
B) restriction of range
C) too few observations
D) a nonlinear relationship among the variables

Question 55

A researcher is likely to obtain statistically significant results even when variables aren't related if

A) the researcher conducts too many statistical tests
B) there is restriction of range.
C) there is a nonlinear relationship among the variables
D) the researcher fails to collect enough observations.

Question 56

Descriptive research allows you to explain behavior.

Question 57

In order to accurately describe behavior, you must use a representative sample of the population you are studying.

Question 58

Before you can predict behavior accurately, you need to be able to describe it.

Question 59

Archival research uses data that you collected before you came up with your hypothesis.

Question 60

Observational techniques are only used in correlational research.

Question 61

If you want to increase the construct validity of your correlational research, it is often best to use validated tests.

Question 62

A positive correlation of .57 is stronger than a negative correlation of -.63.

Question 63

The coefficient of determination tells you the strength of the relationship between two variables.

Question 64

If you uncover a relationship that is statistically significant at the .001 level, you can conclude that your variables are most likely causally linked.