1
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
Chapter 3
Secondary Data, Literature Reviews, and
Hypotheses
Learning Objectives (PPT slide 3-2)
1. Understand the nature and role of secondary data.
2. Describe how to conduct a literature review.
Key Terms and Concepts
1. Alternative hypothesis
2. Causal hypotheses
3. Conceptualization
4. Construct
5. Consumer panels
6. Dependent variable
2
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
24. Variable
Chapter Summary by Learning Objectives
Understand the nature and role of secondary data.
The task of a marketing researcher is to solve the problem in the shortest time, at the least cost,
with the highest level of accuracy. Therefore, before any marketing research project is conducted,
the researcher must seek out existing information that may facilitate a decision or outcome for a
company. Existing data are commonly called secondary data. If secondary data are to be used to
assist the decision-making process or problem-solving ability of the manager, they need to be
Describe how to conduct a literature review.
A literature review is a comprehensive examination of available information that is related to your
research topic. When conducting a literature review, researchers locate information relevant to the
research problems and issues at hand. Literature reviews have the following objectives: provide
Identify sources of internal and external secondary data.
Internal secondary data are obtained within the company. Company internal accounting and
financial information is a major source. These typically consist of sales invoices, accounts
receivable reports, and quarterly sales reports. Other forms of internal data include past marketing
research studies, customer credit applications, warranty cards, and employee exit interviews.
3
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
sources of data; and tabulate all the data extracted. The most common sources of external
secondary data are popular, scholarly, government, and commercial.
Discuss conceptualization and its role in model development.
Literature reviews can also help you conceptualize a model that summarizes the relationships you
hope to predict. If you are performing purely exploratory research, you will not need to develop a
model before conducting your research. Once you have turned your research objectives into
Understand hypotheses and independent and dependent variables.
A variable is an observable item that is used as a measure on a questionnaire. A construct is an
unobservable concept that is measured by a group of related variables. Some examples of
commonly used constructs in marketing research include service quality, value, customer
satisfaction, and brand attitude. Constructs that represent characteristics of respondents may also
be measured, for example, innovativeness, opinion leadership, and deal proneness. There are two
kinds of hypotheses: descriptive and causal. Descriptive hypotheses are possible answers to
Chapter Outline
Opening Vignette: Will Brick-and-Mortar Stores Eventually Turn into Product
Showrooms?
Research conducted during the recent holiday shopping season provided evidence of an important
and growing trend in shopping: using one’s mobile phone while shopping in a store as an aid in
4
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
Following were the three specific ways the mobile phones were used.
• Fully 38 percent of respondents reported using the cell phone to call a friend for advice while
in a store.
• 24 percent of respondents used mobile devices to search for online product reviews.
• One in four shoppers used their mobile phone to compare prices, for example, using
Amazon’s barcode scan service to search for more competitive prices.
Pew Internet and American Life’s research on mobile phone use while shopping is available at
their website and is published at several other media sites as well. Their study is an example of
secondary data, and in this case, the research is available for free.
I. Value of Secondary Data and Literature Reviews
A literature review is an important step in developing an understanding of a research topic and
supports the conceptualization and development of hypotheses.
A. Nature, Scope, and Role of Secondary Data (PPT slide 3-3 and 3-4)
The term secondary data refers to data gathered for some other purpose than the immediate
study. Sometimes it is called “desk research” while primary research is called “field research.
The two types of secondary data are as follows.
• Internal secondary data are collected by a company for accounting purposes, marketing
programs, inventory management and so forth.
• External secondary data are collected by outside organizations such as federal and state
governments, trade associations, nonprofit organizations, marketing research services, or
academic researchers.
5
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
The role of secondary data in marketing research has changed in recent years. Traditionally,
secondary data were viewed as having limited value. The job of obtaining secondary data often
was outsourced to a corporate librarian, syndicated data collection firm, or junior research
II. Conducting a Literature Review (PPT slide 3-5)
A literature review is a comprehensive examination of available secondary information related to
the research topic. Secondary data relevant to the research problem obtained in a literature review
should be included in the final report of findings. This section of the report typically is labeled
“background research,” or “literature review.” Secondary research alone may sometimes provide
the answer to the research question, and no further research will be required.
But even if analysts plan to conduct primary research, a literature review can be helpful because it:
• provides background and contextual information for the current study
• clarifies thinking about the research problem and questions that are the focus of the study
There are several other reasons for conducting a literature review. These are as follows.
• It helps researchers stay abreast of the latest thinking related to their topic of interest.
• It can help clarify and define the research problem and research questions.
• It can uncover other published studies on the topic of consumer advertising engagement, as
well as the different ways to define and measure consumer engagement.
6
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
A. Evaluating Secondary Data Sources (PPT slide 3-6)
With the increasing emphasis on secondary data, researchers have developed criteria to
evaluate the quality of information obtained from secondary data sources. Following are the
criteria used to evaluate secondary data.
• Purpose—because most secondary data are collected for purposes other than the one at
hand, the data must be carefully evaluated on how it relates to the current research
objective.
• Accuracy—it is enhanced when data are obtained from the original source of the data. A
lot of information in industries may be cited repeatedly without checking the original
source.
B. Secondary Data and the Marketing Research Process (PPT slide 3-7)
In many areas of marketing research, secondary research plays a subordinate role to primary
research. In product and advertising concept testing, focus groups, and customer satisfaction
surveys, only primary research can provide answers to marketing problems. But in some
situations, secondary data by itself can address the research problem. If the problem can be
solved based on available secondary data alone, then the company can save time, money, and
effort.
7
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
III. Internal and External Sources of Secondary Data (PPT slide 3-8 to 3-13)
Secondary data are available both within the company and from external sources.
A. Internal Sources of Secondary Data (PPT slide 3-8 and 3-9)
Generally, internal secondary data consists of sales, accounting, or cost information. Exhibit 3.2
lists common sources of internal secondary data (PPT slide 3-8).
• Sales invoices
• Accounts receivables reports
Exhibit 3.3 outlines other potential sources of internal secondary data (PPT slide 3-9)
B. External Sources of Secondary Data (PPT slide 3-10 to 3-12)
Primary sources of external secondary data are as follows.
• Popular sources
• Scholarly sources
Many popular sources are available both in the library and on the Internet. Examples of popular
sources include Bloomberg Businessweek, Forbes, Harvard Business Review, Wired, and so on.
Most popular articles are written for newspapers and periodicals by journalists or freelance
writers.
Many business students are already familiar with the business articles and resources offered by
ABI/Inform or Lexus/Nexus. These databases can be searched through online library gateways
at most colleges and universities.
A great deal of information is available on the Internet without subscription to library
8
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
databases. Search engines continually catalog this information and return the most relevant and
most popular sites for particular search terms. Google, Yahoo!, and Bing are all good at locating
published studies.
All popular sources found on the web need to be carefully evaluated. The “About Us” portion of
the website should be checked to see who is publishing the articles or studies to see if the source
of the material is reputable. Another issue to consider is that marketing studies found on
websites sometimes promote the business interest of the publisher.
Google has a specialized search engine dedicated to scholarly articles called Google Scholar.
Using Google’s home page search function rather than Google Scholar will identify some
scholarly articles, but will include many other kinds of results that make scholarly articles
difficult to identify.
Detail, completeness, and consistency are major reasons for using U.S. government documents.
Exhibit 3.5 lists some of the common sources of secondary data available from the U.S.
government (PPT slide 3-12).
There are two notes of caution about census data and often other types of secondary data.
• Census data are collected only every 10 years with periodic updates, so researchers
always need to consider the timeliness of census data.
• Census data include only a limited number of topics. As with other secondary data,
predefined categories of variables such as age, income, and occupation may not always
9
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
meet user requirements.
A major trend in marketing research is toward a greater dependency on syndicated data sources.
The rationale for this is that companies can obtain substantial information from a variety of
industries at a relatively low cost.
Syndicated data, or commercial data, are market research data that are collected, packaged,
and sold to many different firms (PPT slide 3-11). The information is in the form of tabulated
reports prepared specifically for a client’s research needs, often tailored to specific reporting
divisions. Reports also may be organized by geographic region, sales territory, market segment,
product class, or brand. Suppliers have traditionally used the following two methods of data
collection.
• Consumer panels
• Store audits
A third method that has increased rapidly in recent years is obtained using optical-scanner
technology. Scanner data are most often obtained at the point of purchase in supermarkets,
drugstores, and other types of retail outlets.
Consumer panels consist of large samples of households that have agreed to provide detailed
data for an extended period of time.
Benefits associated with panel data are as follows.
• Lower cost than primary data collection methods
• Rapid availability and timeliness
10
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
• Accurate reporting of socially sensitive expenditures, for example, beer, liquor,
cigarettes, and generic brands
• High level of specificity
There are two types of panel-based data sources.
• Those reflecting actual purchases of products and services
• Those reflecting media habits
• Mintel tracks consumer spending across 34 countries and provides syndicated data for
specific such as beauty, food and drink, and household and personal care.
The following list describes additional syndicated data companies and the consumer panels
they maintain:
• J. D. Power and Associates maintains a consumer panel of car and light-truck owners to
provide data on product quality, satisfaction, and vehicle dependability.
• GfK Roper Consulting offers subscription to their syndicated services.
Media panels and consumer panels are similar in procedure, composition, and design. They
differ only in that media panels primarily measure media consumption habits as opposed to
product or brand consumption.
Nielsen Media Research is by far the most widely known and accepted source of media panel
data. It collects data on television viewing habits through a device, called a people meter,
connected to a television set.
Nielsen Audio conducts ongoing data collection for media, including radio, television, cable,
and out of home. They are perhaps best known for radio audience measurement.
11
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
brand movement in return for detailed activity reports and cash compensation to the retailer.
The audits then operate as a secondary data source. Store audits provide two unique benefits.
• Precision
Many of the biases of consumer panels are not found in store audits. By design, store audits
measure product and brand movement directly at the point of sale (usually at the retail level).
Key variables measured in a store audit include the following:
• beginning and ending inventory levels
• sales receipts
Collectively, these data allow users of store audit services to generate information on the
following factors.
• Product/brand sales in relation to competition.
• Effectiveness of shelf space and POP displays.
C. Synthesizing Secondary Research for the Literature Review (PPT slide 3-13)
Divergent perspectives and findings need to be included in the literature review. It is likely that
the findings of some studies will be inconsistent with each other. These differences may include
estimates of descriptive data, for example, the percentage of people who buy from catalog
marketers, the amount of dollars spent on advertising, or online retail sales numbers. Reports
may also disagree as to the nature of theoretical relationships between variables.
Researchers often need to dig into the details of the methodology that is used to define variables
and collect data. For example, differences in estimates of online retail spending are caused by
several factors. Three major causes of discrepancies in online retail estimates are as follows.
i. The inclusion (or not) of travel spending, which is a major category of online spending
12
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
It is necessary to make intelligent judgments about the sources that may be causing the
differences in the findings.
IV. Developing a Conceptual Model (PPT slide 3-14 to 3-15)
In addition to providing background for a research problem, literature reviews can also help
conceptualize a model that summarizes the relationships one hopes to predict. If a researcher is
performing purely exploratory research, he or she will not need to develop a model before
conducting his or her research. Once the research objectives have been turned into research
A. Variables, Constructs, and Relationships (PPT slides 3-15)
To conceptualize and test a model, researchers must have the following three elements.
• Variables
• Constructs
• Relationships
A variable is an observable item that is used as a measure on a questionnaire. Variables have
concrete properties and are measured directly. Examples of variables include gender, marital
status, company name, number of employees, how frequently a particular brand is purchased,
and so on. In contrast, a construct is an unobservable, abstract concept that is measured
indirectly by a group of related variables. Some examples of commonly measured constructs in
marketing include service quality, value, customer satisfaction, and brand attitude.
A literature review will help identify, define, and measure constructs. Nevertheless, after
conducting a literature review and consulting secondary research an analyst may believe there
is not enough information to design a full-scale study. There may be several sources of
uncertainty.
• The definition of important constructs
13
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
• The identification of variables or items that will measure each construct
• The identification of constructs that may have an important role in affecting an outcome
or dependent variable of interest
B. Developing Hypotheses and Drawing Conceptual Models (PPT slide 3-16 to 3-20)
Coming up with hypotheses is often relatively straightforward. There are two types of
hypotheses:
• Descriptive—descriptive hypotheses address possible answers to specific business
problems (PPT slide 3-16). Developing descriptive hypotheses involves three steps.
o Reviewing the research problem or opportunity
o Writing down the questions that flow from the research problem or opportunity
Causal hypotheses help businesses understand how they can make changes that, for example,
improve awareness of new products, service quality, customer satisfaction, loyalty, and
repurchase.
Hypotheses can suggest positive or negative relationships. A positive relationship between
two variables is when the two variables increase or decrease together. Negative relationships
suggest that as one variable increases, the other one decreases.
Good hypotheses have several characteristics (PPT slide 3-18).
• Hypotheses follow from research questions.
• It must be written clearly and simply. If the hypothesis is a causal hypothesis, it must have
both an independent and dependent variable in the statement.
To more effectively communicate relationships and variables, researchers follow a process
called conceptualization (PPT slide 3-19). Conceptualization involves the following (PPT
14
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
slide 3-19):
• identifying the variables for the research
• specifying hypotheses and relationships
V. Hypothesis Testing (PPT slides 3-24 and 3-25)
A hypothesis is an empirically testable though yet unproven statement developed in order to
explain a phenomena. It suggests relationships between variables.
If the difference between two means is large, one would be more confident that there is, in fact, a
true difference between the sample means of the two groups. But another important component to
consider is the size of the sample used to calculate the means because the size of the sample and the
variance in the sample affect sample error.
In marketing research the null hypothesis is developed so that its rejection leads to an acceptance
of the alternative hypothesis. Usually, the null hypothesis is notated as H0 and the alternative
hypothesis is notated as H1. If the null hypothesis (H0) is rejected, then the alternative hypothesis
(H1) is accepted. The alternative hypothesis always bears the burden of proof.
Marketing Research in Action
The Santa Fe Grill Mexican Restaurant (PPT slide 3-26 and 3-27)
The Marketing Research in Action focus in this chapter returns to the Santa Fe Grill. Recall that
15
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
better understand these core concepts.
Answers to Hands-On Exercise
1. Based on your understanding of the material presented in Chapter 3 and the above key
research questions, should the owners of the Santa Fe Grill Mexican restaurant go back and
restate their questions? If “no,” why not? If “yes,” why? Suggest how the research questions
could be restated.
Back in Chapter 1, we learned that the Santa Fe Grill owners had the following questions.
• Are our customers satisfied with the restaurant?
• Are there problems with food, atmosphere, or some other attribute?
Now that the owners understand the meaning of satisfaction, image, and loyalty, these
research questions can certainly be restated. Students’ responses will vary. Given below are
some of the possibilities.
• When customers choose the Santa Fe Grill as their restaurant of choice, what are their
expectations regarding food quality, price, atmosphere, and image?
2. Regarding the owners’ desire to understand the interrelationships between customer
satisfaction, restaurant store image, and customer loyalty, develop a set of hypotheses that
might be used to investigate these interrelationships.
Students’ responses will vary. The following are the possible hypotheses.
• The more positively a store image is perceived by customers, the higher their
satisfaction with the experience is.
Answers to Review Questions
1. What characteristic separates secondary data from primary data? What are three sources of
secondary data?
16
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
The primary characteristic which distinguishes secondary data from primary data is that
secondary data “already exists” (or has already been collected by someone). Primary data is
information that hasn’t been collected yet. Another way of viewing the distinction is that
secondary data is “arms–length” in nature—it may have been collected by another company
or group of individuals, or it may be out of date, or “out of sync” with the needs of the
decision makers (for example, data was collected on beer consumption in liters and this has
to be transcribed into ounces). Primary data is more “up close and personal.” It requires a
direct connection with the respondents. Secondary data collection is more cost-effective and
efficient when compared with gathering primary data. The following are the three primary
sources of secondary data.
• Internal data—which exists within the company such as sales reports, profit and loss
statements, employee reviews, and databases
2. Explain why a company should use all potential sources of secondary data before initiating
primary data collection procedures.
A company should use all potential sources of secondary data for a number of reasons. First,
the role of secondary data in marketing research has changed during the past two decades.
3. List the six fundamental principles used to assess the validity of secondary data.
The following are the six fundamental principles that are used to assess the validity of
secondary data.
• Purpose: The data must be related to the research objective.
• Accuracy: The party which collected the data, the time period, and sources need to be
evaluated.
17
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
• Credibility: The source and parties collecting the data need to be well regarded and
truthful.
4. What are the various reasons to conduct a literature review?
The following are some of the reasons to conduct a literature review.
• It can help clarify and define the problems or questions.
• It can uncover existing studies on a topic.
5. What should you look for in assessing whether or not an Internet resource is credible?
To assess a source, one can check the “about us” portion of a website to see who is
publishing the information and whether the source is reputable. One should also consider
6. A researcher develops hypotheses that suggest consumers like ads better when they (1) are
truthful, (2) creative, and (3) present relevant information. Picture the conceptual model that
would show these relationships. Which variables are the independent and which the
dependent variables?
The dependent variable is the perceived liking of the ad. The independent variables are the
degree of truthfulness perceived, the degree of creativity, and the degree of relevant
7. What are relationships? What is a positive relationship? What is a negative relationship?
Give an example of a positive and a negative relationship.
Relationship is the nature of the association between variables. A positive relationship
means that the variables move in tandem. A negative relationship means that as one
18
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
relationship increases or decreases, the other variable moves in the opposite direction.
8. What is the difference between a parameter and a sample statistic?
A parameter is the true value of a variable, while a sample statistic is the value of a variable
that is estimated from a sample.
Answers to Discussion Questions
1. It is possible to design a study, collect and analyze data, and write a report without
conducting a literature review? What are the dangers and drawbacks of conducting your
research without doing a literature review? In your judgment, do the drawbacks outweigh
the advantages? Why, or why not?
The primary danger of conducting a research without doing a literature review is that one
may not even need to collect primary data. The answer could already be available but the
2. EXPERIENCE MARKETING RESEARCH. Visit several of the marketing blogs listed
in Exhibit 3.4. Do these blogs have any information that might be relevant to practitioners
who are conducting research in the topic areas that the blogs address? Why, or why not?
These blogs typically are written from the perspective of an opinion piece commenting on
some news story or recent finding or experience. From that perspective, yes, the comments
3. EXPERIENCE MARKETING RESEARCH. Using Google Scholar, identify ten or so
references that are relevant to the topic of service quality. Identify four or five that you think
would be helpful to designing a survey to measure the service quality received at a restaurant.
19
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
Nearly three million results appear when one searches “service quality” using Google
Scholar. Not surprisingly, many of these references direct the user to the work of
Parasuraman, Zeithaml, and Berry. The following are samples students might select.
• Servqual: A Multiple-item Scale for Measuring Customer Perceptions of Service
Quality
A. Parasuraman, Valarie A. Zeithaml, Leonard L. Berry, Marketing … – 1986 –
Marketing Science Institute
Students’ responses will vary, but students should be encouraged to provide a complete
profile and discuss the marketing implications of the profile.
5. EXPERIENCE MARKETING RESEARCH. Go to the home page of the U.S. census,
This link provides thirteen indicators, ranging from manufacturer’s shipments, inventories,
and orders to housing vacancies and homeownership with comparative data profile, a
summary which explains the “trending” associated with the particular indicator, and a
graphical representation which helps clarify the information.
6. You are thinking about opening a new quick service restaurant on your campus after you
graduate. What information about your potential consumers would you try to find from
secondary research to help you understand your target market?
20
Chapter 03 – Secondary Data, Literature Reviews, and Hypotheses
Students’ responses will vary. Some of them may say that they will go through secondary
research to see what kind of food the students generally seem to appreciate, what is the
amount the students are willing to spend on their food, etc.
7. You are planning to open a coffee shop in one of two areas in your local community.
Conduct a secondary data search on key variables that would allow you to make a logical
decision on which area is best suited for your proposed coffee shop.
8. Based on your experiences in college, draw and label a conceptual model that shows the
factors that led to your satisfaction (or dissatisfaction) with a course.
Students’ responses will vary based on their individual experiences, but some of the factors
might include the following:
• relevance of course to major and intended career
• cost of book