Chapter 14

Data Preparation
True/False Questions
1. The data preparation process begins after the fieldwork is done.
(False, moderate, page 402)
2. The data-preparation process is guided by the preliminary plan of data that was
formulated in the research design phase.
(True, moderate, page 402)
3. The initial step in questionnaire checking involves a check of all questionnaires for
completeness and interviewing quality.
(True, easy, page 403)
4. If quotas or cell group sizes have not been met, additional interviews should be
conducted in the underrepresented cells before the data are edited.
(True, moderate, page 403)
5. At the editing stage, the researcher makes a preliminary check for consistency.
(True, moderate, page 403)
6. Unsatisfactory respondents may differ from satisfactory respondents in systematic
ways and the decision to designate a respondent as unsatisfactory may be subjective.
Both of these factors bias the results.
(True, moderate, page 404)
7. When entering respondent data into a spreadsheet, each column contains the data for
one respondent.
(False, easy, page 404)
8. The following codes should be included for each respondent: project code,
interviewer code, date and time codes, category codes, and validation code.
(False, moderate, page 405)
9. If possible, standard codes should be used for missing data.
(True, easy, page 405)
10. The missing value codes should be distinct from the codes assigned to the legitimate
responses.
(True, easy, page 405)
11. A codebook contains coding instructions and the necessary information about
variables in the data set. It guides the coders in their work and helps the researcher to
properly identify and locate the variables.
(True, easy, page 407)
180

12. Given the time and cost constraints, as well as the fact that experienced keypunch
operators are quite accurate, it is sufficient to verify only 15-25 percent of the data
entered by keypunching.
(False, moderate, page 407)
13. The selection of data-transcription method is guided by the type of interviewing
method used and the availability of equipment.
(True, easy, page 407)
14. Data cleaning includes consistency checks and treatment of missing responses. The
checks at this stage are less extensive than the checks made during editing.
(False, moderate, page 409)
15. Computer packages like SPSS, SAS, EXCEL, and MINITAB can be programmed to
identify out-of-range values for each variable and print related code information to
make it easy to check each variable systematically for out-of-range values.
(True, moderate, pages 410)
16. Responses can be logically inconsistent.
(True, moderate, page 410)
17. All extreme values result from errors and point to problems with the data.
(False, difficult, page 410)
18. Missing responses represent values of a variable that are unknown, either because
respondents provided ambiguous answers or their answers were not properly
recorded.
(True, moderate, page 410)
19. Treatment of missing responses poses problems, particularly if the proportion of
missing responses is more than 10 percent.
(True, moderate, page 410)
20. The value 1.0 represents the weighted case.
(False, easy, page 411)
21. The effect of weighting is to increase or to decrease the number of cases in the sample
that possess certain characteristics.
(True, easy, page 411)
22. The purpose of respecification is to create variables that are consistent with the
objectives of the study.
(True, moderate, page 413)
23. Square root and log transformations are seldom used in the respecification of
variables.
(False, moderate, page 413)
181

24. An important variable respecification procedure involves the use of dummy variables.
(True, moderate, page 413)
25. The use of dummy variables refers to a respecification procedure using variables that
take on only two values, usually 0 or 1.
(True, moderate, page 413)
26. Dummy variables are used to respecify categorical variables.
(True, moderate, page 413)
27. The general rule for respecifying categorical variables with K categories is that K + 1
dummy variables are needed.
(False, moderate, page 413)
28. When respecifying a categorical variable with K categories, only K 1 dummy
variables are needed because only K 1 categories are independent.
(True, moderate, page 413)
29. In the Health Care Services example given in your text, the data was transformed by
subtracting the mean value from each response in an effort provide a more accurate
idea of the relative importance of the data.
(True, difficult, page 414)
30. To standardize a scale Xi, we first subtract the mean, X , from each score and divide
by the standard error, sX .
(False, difficult, page 414)
31. Standardized zi scores can be obtained as: zi = (X - X )/s.
(True, moderate, page 414)
32. Only one technique is appropriate for analyzing the data from a given project.
(False, easy, page 416)
33. Statistical techniques can be classified as univariate or multivariate.
(True, easy, page 416)
34. Univariate techniques can be classified based on whether the data are metric or
nonmetric.
(True, easy, page 416)
35. Metric data are measured on a nominal or ordinal scale.
(False, easy, page 416)
36. Nonmetric data are measured on an interval or ratio scale.
(False, easy, page 416)

182

37. When using univariate techniques, the number of samples is determined based on
how the data are treated for the purpose of analysis, not based on how the data were
collected.
(True, moderate, page 416)
38. Multivariate technique can be classified as dependence techniques or interdependence
techniques.
(True, easy, page 416)
39. In interdependence techniques, the variables are classified as dependent or
independent.
(False, moderate, page 417)
40. The major technique for examining variable interdependence is factor analysis.
(True, easy, page 417)
41. In order to perform individual-level analysis enough data must be obtained from each
individual to allow analysis at the individual level.
(True, easy, pages 418-419)
42. Within-country analysis is also referred to as intercultural analysis.
(False, moderate, page 419)
43. Pancultural and cross-cultural analysis are types of cultural unit analysis.
(False, moderate, page 419)
44. When doing international research, differences in means, differences in distribution,
and differences in variance should be assessed.
(True, moderate, page 419)
45. Ethical issues that arise during the data preparation and analysis step of marketing
research process pertain mainly to the researcher.
(True, moderate, page 418)
Multiple Choice Questions
46. Which of the following is not a step in the data-preparation process?
a. editing
b. select a sampling technique
c. transcribing
d. data cleaning
(b, moderate, page 402)

183

47. A questionnaire returned from the field may be unacceptable for several reasons.
Which of the reasons below is not one of those reasons?
a. Parts of the questionnaire may be incomplete.
b. The returned questionnaire is physically incomplete.
c. The questionnaire is answered by someone who does not qualify for participation.
d. The questionnaire does not uplift, motivate, or encourage the respondent to
become involved.
(d, moderate, pages 403)
48. If you discover that a respondent has consistently checked the same response on a
series of scales or that skip patterns were not followed on a questionnaire, more than
likely you are participating in the _____ step of the data-preparation process.
a. coding
b. data cleaning
c. questionnaire checking
d. editing
(c, moderate, page 403)
49. The _____ step of the data-preparation process consists of screening questionnaires to
identify illegible, incomplete, inconsistent, or ambiguous responses.
a. coding
b. data cleaning
c. questionnaire checking
d. editing
(d, moderate, page 403)
50. Unsatisfactory responses uncovered in the editing stage of the data-preparation
process are commonly handled by doing any of the following except:
a. returning to the field to get better data
b. assigning missing values
c. discarding unsatisfactory responses
d. All of the above are done to handle unsatisfactory responses.
(d, moderate, page 403)
51. Which treatment of unsatisfactory responses is particularly attractive for business and
industrial marketing surveys where the sample sizes are small and the respondents are
easily identifiable?
a. returning to the field to get better data
b. classifying and counting questionnaires accordingly
c. discarding unsatisfactory responses
d. assigning missing values
(a, moderate, page 403)

184

52. Which treatment of unsatisfactory responses is desirable if (1) the number of

respondents with unsatisfactory responses is small; (2) the proportion of
unsatisfactory responses for each of these respondents is small; or (3) the variables
with unsatisfactory responses are not the key variables?
a. returning to the field to get better data
b. classifying and counting questionnaires accordingly
c. discarding unsatisfactory responses
d. assigning missing values
(d, difficult, page 403)
53. Which treatment of unsatisfactory responses is desirable if (1) the unsatisfactory
respondents do not differ from satisfactory respondents in obvious ways (e.g.,
demographics, product usage characteristics), (2) the proportion of unsatisfactory
responses for each of these respondents is large, or (3) responses on key variables are
missing?
a. returning to the field to get better data
b. classifying and counting questionnaires accordingly
c. discarding unsatisfactory responses
d. assigning missing values
(c, difficult, page 403)
54. _____ is the assignment of a code to represent a specific response to a specific
question along with the data record and column position that code will occupy.
a. Coding
b. Data cleaning
c. Questionnaire checking
d. Editing
(a, easy, page 404)
55. A code in which the number of records for each respondent are the same, and the
same data appear in the same columns for all respondents, is a _____.
a. respondent code
b. fixed-field code
c. category code
d. record code
(b, moderate, page 405)
56. The general guidelines for coding unstructured questions and questionnaires in
general include all of the following except:
a. Category codes should be mutually exclusive and collectively exhaustive.
b. Category codes should be assigned for critical issues even if no one has
mentioned them.
c. Answers should be abbreviated as much as possible.
d. Data should be coded to retain as much detail as possible.
(c, moderate, pages 406-407)

185

57. A codebook generally contains all of the following information except:

a. record number
b. category number
c. instructions for coding
d. variable name
(b, moderate, page 407)
58. _____ involves transferring the coded data from the questionnaires or coding sheets
onto disks or magnetic tapes, or directly into computers by keypunching.
a. Coding data
b. Editing data
c. Transcribing data
d. Data Cleaning
(c, moderate, page 407)
59. If the data have been collected via _____, transcribing data is unnecessary because
the data are entered directly into the computer.
a. CATI
b. mark sense forms
c. CAPI
d. both a and c are correct
(d, moderate, page 407)
60. _____ require responses to be recorded with a special pencil in a pre-designated area
coded for that response.
a. CATI or CAPI
b. Computerized sensory systems
c. Mark sense forms
d. Optical scanning
(c, moderate, page 407)
61. _____ involves direct machine reading of the codes and simultaneous transcription.
a. CATI or CAPI
b. Computerized sensory systems
c. Mark sense forms
d. Optical scanning
(d, moderate, page 407)
62. _____ utilizes questions that appear on a computerized gridpad. The responses are
recorded directly into the computer using a sensing device.
a. CATI or CAPI
b. Computerized sensory systems
c. Mark sense forms
d. Optical scanning
(b, moderate, page 407)

186

63. Respondents have been asked to express their degree of agreement with a series of
lifestyle statements on a 1-to-5 scale, assuming that 9 has been designated for missing
values, data values of 0, 6, 7, and 8 are out of range. Where in the data-preparation
process might any out-of-range data be caught?
a. coding
b. transcribing
c. data cleaning
d. statistically adjusting the data
(c, moderate, pages 409-410)
64. Respondents have been asked to express their degree of agreement with a series of
lifestyle statements on a 1-to-5 scale, assuming that 9 has been designated for missing
values, data values of 0, 6, 7, and 8 are out of range. Where in the data cleaning
process might any out-of-range data be caught?
a. consistency checks
b. returning to the field
c. treatment of missing responses
d. both a and c are correct
(a, moderate, pages 409-410)
65. Which of the following is not an option available for the treatment of missing values?
a. substitute a neutral value
b. casewise deletion
c. substitute an imputed response
d. returning to the field
(d, moderate, page 410)
66. If the researcher uses the mean response for the missing value, he/she is using which
option for the treatment of missing values?
a. returning to the field
b. casewise deletion
c. substitute an imputed response
d. substitute a neutral value
(d, moderate, page 410)
67. Which option for the treatment of missing values involves the researcher using the
respondents pattern of responses to calculate a suitable response to the missing
questions?
a. returning to the field
b. casewise deletion
c. substitute an imputed response
d. substitute a neutral value
(c, easy, page 410)

187

68. Which option for the treatment of missing values involves cases, or respondents, with
any missing responses being discarded from the analysis?
a. returning to the field
b. casewise deletion
c. pairwise deletion
d. substitute a neutral value
(b, easy, page 411)
69. Which option for the treatment of missing values involves the researcher using only
cases or respondents with complete responses for each calculation?
a. returning to the field
b. casewise deletion
c. pairwise deletion
d. substitute a neutral value
(c, moderate, page 411)
70. In _____, each case or respondent in the database is assigned a weight to reflect its
importance relative to other cases or respondents.
a. standardization
b. variable respecification
c. scale transformation
d. weighting
(d, easy, page 411)
71. You would use weighting under which circumstance?
a. to create new variables that are composites of several other variables
b. to adjust the sample so that greater importance is attached to respondents with
certain characteristics
c. to compare attitudinal scores measured on a seven-point semantic differential
scale with lifestyle or image scores measured on a five-point Likert scale
d. none of the above
(b, moderate, page 411)
72. If in conducting a study to determine what modifications should be made on an
existing product and the researcher assigns weights of 4.0 to heavy users, 3.0 to
medium users, 2.0 to light users, and 1.0 to nonusers, which groups opinion has
greater importance?
a. nonusers
b. light users
c. medium users
d. heavy users
(d, easy, page 411)

188

73. _____ involves the transformation of data to create variables or modify existing
variables.
a. Standardization
b. Variable respecification
c. Scale transformation
d. Weighting
(b, moderate, page 413)
74.

75.

76.

77.

Dummy variables are also called all of the names below except:
a. binary
b. instrumental
c. quantitative
d. dichotomous
(c, difficult, page 413)
You would use variable respecification under which circumstance?
a. to create new variables that are composites of several other variables
b. to adjust the sample so that greater importance is attached to respondents with
certain characteristics
c. to compare attitudinal scores measured on a seven-point semantic differential
scale with lifestyle or image scores measured on a five-point Likert scale
d. none of the above
(a, moderate, page 413)
_____ involves a manipulation of scale values to ensure comparability with other
scales or otherwise make the data suitable for analysis.
a. Standardization
b. Variable respecification
c. Scale transformation
d. Weighting
(c, moderate, page 414)
You would use scale transformation under which circumstance?
a. to create new variables that are composites of several other variables
b. to adjust the sample so that greater importance is attached to respondents with
certain characteristics
c. to compare attitudinal scores measured on a seven-point semantic differential
scale with lifestyle or image scores measured on a five-point Likert scale
d. none of the above
(c, moderate, page 414)

189

78.

Which statement is not true about standardization?

a. The standardized scale will have a mean of zero and a standard deviation of 1.
b. Standardization allows the researcher to compare variables that have been
measured using different types of scales.
c. Standardization is the process of correcting data to reduce them to the same scale
by subtracting the sample mean and dividing by the standard error.
d. Standardization is the process of correcting data to reduce them to the same scale
by subtracting the sample mean and dividing by the standard deviation.
(c, moderate, page 414)

79. _____ are appropriate when there is a single measurement of each element in the
sample, or there are several measurements of each element but each variable is
analyzed in isolation.
a. Interdependence techniques
b. Dependence techniques
c. Multivariate techniques
d. Univariate techniques
(d, moderate, page 416)
80. _____ are suitable for analyzing data when there are two or more measurements of
each element and the variables are analyzed simultaneously.
a. Interdependence techniques
b. Dependence techniques
c. Multivariate techniques
d. Univariate techniques
(c, moderate, page 416)
81. Data that are interval or ratio in nature are _____.
a. metric data
b. independent data
c. nonmetric data
d. paired data
(a, moderate, page 416)
82. Data derived from a nominal or ordinal scale are _____.
a. metric data
b. independent data
c. nonmetric data
d. paired data
(c, moderate, page 416)

190

83. Univariate techniques can be classified based on _____.

a. whether the data are metric or nonmetric
b. whether one, two, or more than two samples are involved
c. whether interdependence techniques or dependence techniques are to be used
d. a and b are correct
(d, difficult, page 416)
84. When utilizing univariate techniques, for the purpose of analysis, data pertaining to
different groups of respondents, e.g., males and females, are generally treated as
_____ samples while the samples are _____ when the data for the two samples relate
to the same group of respondents.
a. metric; paired
b. independent; metric
c. independent; paired
d. paired; independent
(c, moderate, page 416)
85. The samples are _____ if they are drawn randomly from different populations.
a. metric data
b. independent
c. nonmetric data
d. paired
(b, moderate, page 416)
86. The samples are _____ when the data for the two samples relate to the same group of
respondents.
a. metric data
b. independent
c. nonmetric data
d. paired
(d, moderate, page 416)
87. For metric data, when there is (are) _____ the z test and the t test can be used.
a. only one sample
b. two independent samples
c. more than two independent samples
d. a and b are correct
(d, difficult, page 416)
88. For metric data, when there is (are) _____, one-way analysis of variance (one-way
ANOVA) can be used.
a. only one sample
b. two independent samples
c. more than two independent samples
d. two or more related samples
(c, difficult, page 416)

191

89. For metric data, when there is (are) _____, the paired t test can be used.
a. only one sample
b. two independent samples
c. more than two independent samples
d. two or more related samples
(d, difficult, page 416)
90. For nonmetric data, when there is (are)_____, chi-square, Mann-Whitney, Median, KS, and K-W one-way ANOVA tests can be used.
a. only one sample
b. two independent samples
c. more than two dependent samples
d. two or more related samples
(b, difficult, page 416)
91. For nonmetric data, when there is (are)_____, sign, McNemar, and Wilcoxon tests can
be used.
a. only one sample
b. two independent samples
c. more than two dependent samples
d. two or more related samples
(d, difficult, page 416)
92. Multivariate techniques can be classified based on _____.
a. whether the data are metric or nonmetric
b. whether one, two, or more samples are involved
c. whether interdependence techniques or dependence techniques are to be used
d. a and b are correct
(c, difficult, page 416)
97. _____ are appropriate when one or more variables can be identified as dependent
variables and the remaining as independent variables.
a. Interdependence techniques
b. Dependence techniques
c. Multivariate techniques
d. Univariate techniques
(b, moderate, page 416)
98. _____ are statistical techniques that attempt to group data based on underlying
similarity, and thus allow for interpretation of the data structures. No distinction is
made as to which variables are dependent and which are independent.
a. Interdependence techniques
b. Dependence techniques
c. Multivariate techniques
d. Univariate techniques
(a, moderate, page 417)

192

95. For multivariate statistical techniques, when there is _____, cross-tabulation, analysis
of variance and covariance, regression, two-group discriminant analysis, and conjoint
analysis can be used.
a. variable interdependence
b. one dependent variable
c. more than one dependent variable
d. interobject similarity
(b, difficult, page 417)
96. For multivariate statistical techniques, when there is _____, multivariate analysis of
variance and covariance and canonical correlation, and multiple discriminant analysis
can be used.
a. variable interdependence
b. one dependent variable
c. more than one dependent variable
d. interobject similarity
(c, difficult, page 417)
97. For multivariate statistical techniques, when there is _____, factor analysis can be
used.
a. variable interdependence
b. one dependent variable
c. more than one dependent variable
d. interobject similarity
(a, difficult, page 417)
98. For multivariate statistical techniques, when there is _____, cluster analysis and
multidimensional scaling can be used.
a. variable interdependence
b. one dependent variable
c. more than one dependent variable
d. interobject similarity
(d, difficult, page 417)
99. _____ requires that the data from each respondent be analyzed separately.
a. Cross-cultural analysis
b. Individual-level analysis
c. Pancultural analysis
d. Intracultural analysis
(b, moderate, page 419)
100. _____ is within-country analysis of international data.
a. Cross-cultural analysis
b. Individual-level analysis
c. Pancultural analysis
d. Intracultural analysis
(d, moderate, page 419)
193

101. _____ is also known as across-countries analysis. The data of all the respondents
can be pooled and analyzed.
a. Cross-cultural analysis
b. Individual-level analysis
c. Pancultural analysis
d. Intracultural analysis
(c, moderate, page 419)
102. _____ is also known as across-countries analysis. The data can be aggregated for
each country and these aggregate statistics analyzed.
a. Cross-cultural analysis
b. Individual-level analysis
c. Pancultural analysis
d. Intracultural analysis
(a, moderate, page 419)
103. _____ is similar to that conducted in domestic marketing research. The objective
is to gain an understanding of the relationships and patterns existing in each country
or cultural unit.
a. Cross-cultural analysis
b. Individual-level analysis
c. Pancultural analysis
d. Intracultural analysis
(d, moderate, page 419)
104. The objective of this level of analysis is to assess the comparability of findings
from one country to another.
a. Cross-cultural analysis
b. Individual-level analysis
c. Pancultural analysis
d. Intracultural analysis
(a, moderate, page 419)
105. Which statement is true concerning ethical issues concerned with data
preparation?
a. Researchers should try to get some idea about the quality of the data.
b. Discarding respondents after analyzing the data raises ethical concerns.
c. Any departure from the assumptions underlying the statistical techniques used to
analyze the data raises ethical concerns.
d. All of the above statements are true.
(d, moderate, page, 419, 420)

194

Essay Questions
106.

Give a brief overview of the data-preparation process.

Answer
The data-preparation process is shown in Figure 14.1. The entire process is guided by
the preliminary plan of data analysis that was formulated in the research design phase.
The first step is to check for acceptable questionnaires. This is followed by editing,
coding, and transcribing the data. The data are cleaned and a treatment for missing
responses prescribed. Often, statistical adjustment of the data may be necessary to
make them representative of the population of interest. The researcher should then
select an appropriate data analysis strategy. The final data analysis strategy differs
from the preliminary plan of data analysis due to the information and insights gained
since the preliminary plan was formulated. Data preparation should begin as soon as
the first batch of questionnaires is received from the field, while the fieldwork is still
going on. Thus if any problems are detected, the fieldwork can be modified to
incorporate corrective action.
(moderate, page 402)
107. Write an example question and its associated coding for (1) a structured question
with a single response with less than nine possible choices and (2) a structured
question that permits multiple responses.
Answer
A single column is sufficient to code a structured question with a single response if
there are less than nine possible responses. In questions that permit a large number of
responses, each possible response option should be assigned a separate column. The
student should provide his/her own examples. Example responses follow.
Do you have a currently valid passport?
1.Yes
2. No (1/54)
For this question, a Yes response is coded 1 and a No response 2. The numbers in
parentheses indicate that the code assigned will appear on the first record for this
respondent in column 54. Because only one response is allowed and there are only
two possible responses (1 or 2), a single column is sufficient.
Which accounts do you now have at this bank? (X as many as apply)
Regular savings account _ (162)
Regular checking account _ (163)
Mortgage
_ (164)
Now account
_ (165)
In this example, suppose a respondent checked regular savings, regular checking, and
now accounts. On record #9, a 1 will be entered in the column numbers 162, 163, and
165. Column 164 will receive a 0. Since there is only one record per respondent, the
record number has been omitted.
(difficult, page 406)
195

108.

Discuss the procedures for coding unstructured or open-ended questions.

Answer
Respondents verbatim responses are recorded on the questionnaire. Codes are then
developed and assigned to these responses. Sometimes, based on previous projects or
theoretical considerations, the researcher can develop the codes before beginning
fieldwork. Usually, this must wait until the completed questionnaires are received.
Then the researcher lists 50 to 100 responses to an unstructured question to identify
the categories suitable for coding. Once codes are developed, the coders should be
trained to assign the correct codes to the verbatim responses. Category codes should
be mutually exclusive and collectively exhaustive. Category codes should be assigned
for critical issues even if no one has mentioned them. Data should be coded to retain
as much detail as possible.
(moderate, pages 406-407)
109. Discuss the pros and cons of the various options for the treatment of missing
responses.
Answer
1. Substitute a Neutral Value. If the mean is substituted for the missing value, the
mean of the variable remains unchanged and other statistics, such as correlations, are
not affected much. Although this approach has some merit, the logic of substituting a
mean value (say 4) for respondents who, if they had answered, might have used either
high ratings (6 or 7) or low ratings (1 or 2) is questionable.
2. Substitute an Imputed Response. This approach requires considerable effort and
can introduce serious bias. Sophisticated statistical procedures have been developed
to calculate imputed values for missing responses.
3. Casewise Deletion. Because many respondents may have some missing responses,
this approach could result in a small sample. Throwing away large amounts of data is
undesirable, because it is costly and time consuming to collect data. Furthermore,
respondents with missing responses could differ from respondents with complete
responses in systematic ways. If so, casewise deletion could seriously bias the results.
4. Pairwise Deletion. Different calculations in an analysis may be based on different
sample sizes. This procedure may be appropriate when (1) the sample size is large,
(2) there are few missing responses, and (3) the variables are not highly related. Yet,
this procedure can produce results that are unappealing or even infeasible.
The different procedures for the treatment of missing responses may yield different
results, particularly when the responses are not missing at random and the variables
are related. Hence, missing responses should be kept to a minimum. The researcher
should carefully consider the implications of the various procedures before selecting a
particular method for the treatment of nonresponse.
(difficult, pages 410-411)

196

110. List the procedures for statistically adjusting data and briefly discuss the uses for
each.
Answer
In weighting, each case or respondent in the database is assigned a weight to reflect
its importance relative to other cases or respondents. Weighting is most widely used
to make the sample data more representative of a target population.
Variable respecification involves the transformation of data to create new variables or
modify existing variables. The purpose of respecification is to create new variables
that are consistent with the objectives of the study.
Scale transformation involves a manipulation of scale values to ensure comparability
with other scales or otherwise make the data suitable for analysis.
(difficult, pages 411, 413-414)
111. Discuss how the researcher should go about selecting a data analysis strategy
(Figure 14.5).
Answer
The selection of a data analysis strategy should be based on the earlier steps of the
marketing research process, known characteristics of the data, properties of statistical
techniques, and the background and philosophy of the researcher.
The selection of a data analysis strategy must begin with a consideration of the earlier
steps in the process: problem definition (Step I), development of an approach (Step
II), and research design (Step III). The preliminary plan of data analysis prepared as
part of the research design should be used as a springboard.
Changes may be necessary in light of additional information generated in subsequent
stages of the research process. The next step is to consider the known characteristics
of the data. The measurement scales used exert a strong influence on the choice of
statistical techniques and, the research design may favor certain techniques.
It is also important to take into account the properties of the statistical techniques,
particularly their purpose and underlying assumptions. Some statistical techniques are
appropriate for examining differences in variables, others for assessing the
magnitudes of the relationships between variables, and others for making predictions.
The techniques also involve different assumptions, and some techniques can
withstand violations of the underlying assumptions better than others.
Finally, the researchers background and philosophy affect the choice of a data
analysis strategy. The experienced, statistically trained researcher will employ a range
of techniques, including advanced statistical methods. Researchers differ in their
willingness to make assumptions about the variables and their underlying
populations. Researchers who are conservative about making assumptions will limit
their choice of techniques to distribution-free methods.
(moderate, pages 414-416)
197

112.

Explain the difference between univariate and multivariate techniques.

Answer
Multivariate techniques differ from univariate techniques in that they shift the focus
away from the levels (averages) and distributions (variances) of the phenomena,
concentrating instead upon the degree of relationships (correlations or covariances)
among these phenomena.
(moderate, page 416)

198