Deck 14: Data Analysis I-- Data Preparation and Presentation

Question 1

A questionnaire is printed in a layout that includes spaces along the border for writing the code. Researchers refer to this design as:

A)edge coding .
B)transfer sheets.
C)boundary data.
D)automated entry form.

Answer 1

A

Question 2

In a codebook, the term "value label" refers to:

A)the value of a numeric variable.
B)a designation in words of what each value of a variable represents.
C)the name that has been given to a variable in a computer file.
D)a designation of what is to be done with missing values in a data file.

Answer 2

B

Question 3

The skewness of a frequency distribution could not be determined if the distribution:

A)involved a variable measured at the interval level.
B)involved numerical data.
C)involved a variable measured at the nominal level.
D)involved a variable measured at the ordinal level.

Answer 3

C

Question 4

You could do all of the following with a variable at the nominal level of measurement EXCEPT:

A)determine the skewness of the distribution.
B)construct a simple frequency distribution.
C)construct a grouped frequency distribution.
D)you could do the procedures mentioned by all the answers.

Question 5

Coding in data analysis refers to:

A)converting numbers into observations.
B)converting observations into Morse code.
C)categorizing a variable into a limited number of categories.
D)using numbers instead of names to protect confidentiality.

Question 6

In addition to designating a variable as string or numeric, statistical software will also often enable you to specify subtypes of string or numeric variables. Such subtypes would include all of the following EXCEPT:

A)currency.
B)name of respondent.
C)decimal.
D)time or date.

Question 7

A grouped frequency distribution might be constructed for all of the following reasons EXCEPT:

A)when a variable has too many categories to be understandable.
B)when some categories of a variable have few or no cases in them.
C)when you want to achieve a higher level of measurement.
D)all of the answers are reasons for grouping.

Question 8

A positively skewed distribution is one in which:

A)the "tail" of the distribution extends toward the higher values of the variable.
B)the "tail" of the distribution extends toward the lower values of the variable.
C)the distribution has no "tail".
D)equal numbers of cases fall on each side of the center of the distribution.

Question 9

When data are first collected, they are in the form of a:

A)raw data distribution.
B)simple frequency distribution.
C)grouped frequency distribution.
D)contingency distribution.

Question 10

In data analysis, the term graph refers to:

A)a frequency distribution in a table format.
B)a visual representation of a frequency distribution.
C)a grouped frequency distribution designed for interval-ratio data.
D)the statistics generated by spreadsheet software.

Question 11

Which of the following is mentioned in the text as a way of entering raw data into a computer file

A)use the data entry facility of a statistical application.
B)use spreadsheet or database software.
C)use optical scanning technology.
D)all of the answers are mentioned in the text.

Question 12

In many statistical analysis programs, a "compute" procedure is used to:

A)detect errors that result during data entry.
B)create new variables from one or more existing variables.
C)decide whether to reject a null hypothesis.
D)determine which kind of graph would be most appropriate for the data at hand.

Question 13

In a grouped frequency distribution, the size of the grouped intervals is called:

A)the computed size.
B)the valid interval.
C)the interval width.
D)the raw interval.

Question 14

In data analysis, spreadsheet programs can be very useful for:

A)importing data into a statistical application.
B)deciding on the appropriate level of measurement for variables.
C)storing qualitative data but not quantitative data.
D)all of the answers are uses of spreadsheet programs.

Question 15

A skewed data distribution is:

A)asymmetrical.
B)symmetrical.
C)fully ordered.
D)at the ratio level of measurement.

Question 16

The term "simple frequency distribution" refers to:

A)a distribution of all the values of all the separate cases on a variable.
B)a distribution in which each value of a variable is listed once, along with the number of cases that have that value.
C)a distribution in which a large number of categories of a variable are combined together to form a smaller number of categories.
D)a distribution that has been placed into some form of graph.

Question 17

A codebook is most useful at which stage of the research process

A)problem formulation
B)literature review
C)sampling
D)data analysis

Question 18

The purpose of the "data cleaning" step in data analysis is to:

A)determine which statistical test is appropriate.
B)detect errors that result during data entry.
C)compute more complex variables from existing variables.
D)eliminate variables that will not be used in the analysis.

Question 19

Coding categories should have which of the following two characteristics

A)overlapping and non-exhaustive.
B)ratio level measurement and reactivity.
C)overlapping and exhaustive.
D)mutually exclusive and exhaustive.

Question 20

__________ refers to deriving some meaning from the observation that have been made during a research project.

A)Data analysis
B)Contingency analysis
C)Table elaboration
D)Data transformation

Question 21

In a bar graph, the vertical axis represents:

A)the frequencies or number of cases in each category.
B)the number of cases in the population.
C)the categories of the variable.
D)the raw data distribution.

Question 22

A partial relationship would best be described as:

A)a relationship between an independent and dependent variable in one category of a test variable.
B)a relationship between an independent and dependent variable without considering any test variables.
C)a relationship between an independent and dependent variable when replication is the outcome.
D)a relationship between an independent and dependent variable when specification is the outcome.

Question 23

In table elaboration, when partial tables show an even stronger relationship between two variables than is found in the zero-order table, this is known as:

A)interaction.
B)spuriousness.
C)suppression.
D)percent difference.

Question 24

In table elaboration, a "test variable" could be best described as:

A)a control variable.
B)an independent variable.
C)a dependent variables.
D)a control variable and an independent variable.

Question 25

Table elaboration is most clearly a part of:

A)multivariate analysis.
B)univariate analysis.
C)bivariate analysis.
D)grouped frequency distributions.

Question 26

If a contingency table is a third-order partial table, then the number of variables controlled is:

A)two.
B)three.
C)four.
D)it can't be determined from the information given.

Question 27

Suppose that in a multivariate analysis the zero-order table shows a strong relationship between two variables while one partial table shows a strong relationship, and another partial table shows no relationship. This outcome is called:

A)interpretation.
B)explanation.
C)specification.
D)spuriousness.

Question 28

The graphing procedure in which a line graph is used to connect the midpoints of each category of the variable is called:

A)a pie chart.
B)a frequency polygon.
C)a histogram.
D)a bar chart.

Question 29

If you find data cast into a contingency table, then which of the following statements is true

A)the analysis may be univariate.
B)the analysis may be bivariate.
C)the analysis may be multivariate.
D)the analysis may be bivariate, or the analysis may be multivariate.

Question 30

You ask people two open-ended questions in a survey: what is your religious affiliation, and list the names of the magazines that you subscribe to. Develop a coding scheme for each of these variables and justify it as a good coding scheme.

Question 31

Which of the following is NOT one of the forms of graphical display of data presented in the text

A)grouped frequency distributions.
B)bar charts.
C)frequency polygons.
D)all of the answers are forms of graphical presentation.

Question 32

All of the following terms refer to outcomes that can occur when conducting table elaboration EXCEPT:

A)interpretation.
B)explanation.
C)interdependence.
D)spuriousness.

Question 33

According to the text, bar graphs are especially useful for representing data that:

A)is at the interval or ratio level of measurement.
B)is at the nominal or ordinal level of measurement.
C)is in numeric rather than alphanumeric form.
D)is continuous rather than discrete.

Question 34

The two preferred methods for graphing continuous data are:

A)pie charts and frequency polygons.
B)bar charts and frequency polygons.
C)pie charts and bar charts.
D)frequency polygons and histograms.

Question 35

When the zero-order relationship is all or partly spurious and the test variable is temporally located before either the independent or the dependent variables, the situation is referred to as:

A)spurious order relationship.
B)specification.
C)antecedent variable relationship.
D)percent difference.

Question 36

In table elaboration, when partial tables show a reduced relationship between the independent and dependent variables than was found in the zero-order table, this could be due to:

A)suppression.
B)spuriousness.
C)interaction.
D)multivariateness.

Question 37

The form of graph that is often used to show how budgets are divided over a number of categories is the:

A)histogram.
B)frequency polygon.
C)bar chart.
D)pie chart.

Question 38

According to the conventions used in this text, contingency tables:

A)contain three or more variables.
B)are applied only to univariate statistics.
C)represent the independent variable with the vertical columns of the table.
D)represent the independent variable with the horizontal rows of the table.

Question 39

A percentage difference derived from a percentage table:

A)is a crude indicator of the strength of a relationship.
B)indicates the amount of error made when converting a contingency table to a percentage table.
C)is meaningless without another one to compare with it.
D)indicates the percentage of the population represented by the sample.

Question 40

The term bivariate statistics is used to describe statistics that:

A)are in graphical form.
B)describe the distribution of a single variable.
C)describe the relationship between two variables.
D)are used for variables at the interval or ratio level.

Question 41

Use the data in Essay Question 3 to construct the most appropriate graph to display the results. Justify why your choice of graph is the most appropriate.

Question 42

What are some of the steps involved in computerizing a set of data Include in your answer a discussion of a codebook and what it is used for.

Question 43

Using the example above on the variables of gender and car accidents in Essay Question 7, describe what table elaboration is and what it might do with this data.

Question 44

Describe each form of graphical presentation discussed in the text. What level of measurement is each most appropriate for What are some of the rules for constructing each one

Question 45

Fifty minor offenders, arrested for shoplifting, were used as subjects in a research project. One half were placed on probation and the other half were given a special shoplifter group counseling program. The results were as follows for those on probation: 20 re arrested, 5 no offenses. For the group counseling subjects, 10 were re arrested and 15 had no offenses. Prepare a contingency table for this data, followed by a percentage table. Finally, compute a percentage difference between the two conditions.What does the percentage difference say about the relationship between the two variables

Question 46

Define and show clearly the distinctions between univariate statistics, bivariate statistics, and multivariate statistics.

Question 47

Identify and describe four possible outcomes that can be found in partial tables. Explain what each means.