The information obtained attempts to find out the ‘how’ and

CHAPTER 5: MARKETING ‘why’ of a situation, rather than ‘how many’.

RESEARCH Analysis may be difficult owing to the depth and complexity of the
data collected and so it should be carried out by experienced and
I. Introduction trained researchers.
The main emphasis of this chapter will be on the decisions that
=> New product development and creative development studies.
must be made as part of survey design; from the definition of the
research problem to the design of the questionnaire and data  Individual Interviews
collection. An individual interview can be conducted over the phone,
The application of marketing research to product development and Skype, or in person. The idea is to ask the ideal user (or
trend forecasting international marketing research issues and the an existing customer) a series of questions and follow-ups
impact of the Internet will also be considered. to learn what motivates them to buy a product like yours.

II. The purpose of marketing research  Focus Groups

Focus groups are generally conducted in-person. These
1) What is marketing research? groups are meant to provide a safe and comfortable
environment for your users to talk about their thoughts
Kotler (2000) defines marketing research as ‘the systematic design, and feelings surrounding your product.
analysis and reporting of data and findings relevant to a specific
marketing situation facing the company’.
 Observations
Marketing research is a broad concept, covering investigation into An in-person observation of shopping behavior allows
all aspects of the marketing of goods or services, such as product you to actually watch the consumer react to your product
research and development, pricing research, advertising in-store. This way, you get to see their actual shopping
research, distribution research, as well as all the aspects of market behavior, as opposed to just what they would claim on a
analysis covered by market research. written survey.

2) Why is information necessary?

Quantitative research provides information to which numbers
Aggressive competitors pose serious threats for both large and can be applied. Quantitative research is the best-known face of
small businesses in the constant fight to maintain and increase their marketing research and its main survey method is what most
market share. To maximize opportunities the successful business, people recognize as marketing research.
person must make the right decisions at the right time. The
consequence of making the wrong decision can be financial ruin. The collection of information is a process that must be planned.
There are many different areas in which planning decisions need to
Sound market information provides the basis for marketing be made, so good organization is vital.
decisions. Marketing research, properly designed and
implemented, will provide this information. III. An overview of the marketing research
3) Marketing research as part of a marketing process
information system 1) Stages in the research process
Research procedures will vary depending on the nature of the
Successful companies operate marketing information systems to
research problem, but in general, the process of marketing research
gather accurate, up-to-date information, analyze it and disseminate
can be seen to be made up of a number of stages. They are:
the results to appropriate decision-makers in time to allow the
1. Define the research problem and set the research objectives.
company to maximize its opportunities and to avoid potential
2. Design the research. This includes:
threats.
(a) data sources;
Marketing research can assist management in the decision-making (b) select the sampling method;
process across the full range of marketing activities, from (c) select the data collection method;
description of a market segment to prediction of future trends. (d) design the data collection form (questionnaire).
3. Test the research design (pilot).
4) Types of research 4. Collect the data.
5. Analyze the data and interpret the results.
Although marketing research techniques can be applied to all areas
6. Present the findings.
of marketing, not all techniques are appropriate to every situation.
Broadly speaking, there are two types of research

- qualitative and quantitative. 2) Problem definition and setting research

Qualitative research uses techniques such as group discussions, Defining the research problem is the most critical step in the
individual depth interviews, projective techniques and observation. research process. Unless the problem is accurately defined, the
information collected will be of limited or no use.
In other words, what is the problem and what do we want to find
out to try to solve it? This preliminary planning is important as it
has implications for the design of the research and the quality

3) Research design
There are three types of research design:
- Exploratory,
- Descriptive
- Causal.
The choice of research design will depend on the problem
previously defined. Some examples of external sources are listed below:
◆ Government statistics: Census data, family expenditure surveys,
a) Exploratory research trade and manufacturing trends.
This is most useful in the early stages of research, particularly if ◆ Trade information: Trade press, e.g. Fashion Weekly, Drapers
the researcher is not familiar with the subject area. Record; trade associations, e.g. CBI, trade surveys, company
The aim of exploratory research is to uncover any variables that reports and competitors’ accounts.
may be relevant to the research project as well as an investigation ◆ Financial institutions: Many major banks publish reports on
of the environment in which the research will take place. regional and national industries.
These are some of the objectives that warrant the need for ◆ Commercial research: Many market research companies
exploratory research: undertake continuous research and omnibus surveys covering an
– Define an ambiguous problem more precisely like why sales of a extremely broad range of topics, including consumer, media and
specific product are declining. retail
– Gain a better understanding of an issue.
– Generate new ideas like what one can do to improve customer b) Primary sources
relationships. Most marketing research projects will involve the collection
– Develop hypotheses that can explain the occurrence of specific of more up-to-date information than is available from secondary
phenomena. sources. Primary sources of information may include consumers,
designers, buyers, manufacturers, retailers, and so on, depending
b) Descriptive research upon the research problem
The purpose of descriptive research is to provide an accurate
description of the variables uncovered by the exploratory stage. 5) Practical sampling methods
This could be used to investigate the market share of a company’s
products or the demographic characteristics of the target market In designing research a major decision that the researcher must
(age, gender, income, etc.). make concerns the selection of a sampling method.
Data are usually obtained from secondary data sources or from
Sampling is a very important tool in marketing research. It
surveys.
involves selecting a small number of people from the larger survey
population whose characteristics, attitudes and behaviour are
c) Causal research representative of the larger group
Causal research is used to determine the relationship between
variables, e.g. the relationship between advertising and repeat a) Deciding sample size
purchases.
Deciding how many people to include in your sample is as
important a decision as how they should be selected.
4) Data sources
Data come from two sources, primary and secondary. Secondary Factors such as cost, time and staff availability, level of accuracy
data sources consist of information that has already been collected required, data collection method and location of the population all
for other purposes and primary sources of information are those play a part in deciding sample size.
used for the purpose of collecting information specifically for the
current research project. When selecting a sample it is important that there is a high level of
confidence that the sample is representative of the research
a) Secondary sources population as a whole. The sample must be large enough to
provide accurate results, without being so large as to increase
These provide the researcher with a starting point for data
research costs unnecessarily. It is possible to calculate confidence
collection. It may be possible to solve the research problem either
levels for different sample sizes and there are several texts that
wholly or in part by using secondary data. This reduces the cost of
cover this adequately
a research project as secondary data are cheaper than collecting
primary data. - In statistics analysis, there are 2 cases which is necessary
to collect data to make comments:

+ Collecting the data to control characteristics of

process/product/service
+ Collecting the data to determine/estimate characteristics of 3000
process/product/sevice n= =60
50
If the random number picked from the tables was 35, for example,
then the first item selected from the sampling frame would be 35.
Every 60th number after that would be selected until a sample size
of 50 was achieved

 Stratified random sampling : One way to try to overcome

this type of sampling error is to use stratified random
sampling. This is used when it is felt that different groups
within the population have characteristics that are likely
to lead to different types of answers. The population is
divided into distinguishable groups (strata) who have
similar characteristics..

Stratification factors should be as relevant as possible to the survey

(e.g. consumer surveys are often stratified by age, gender, socio-
economic group, and so on). A random sample is then taken from
b) Choice of a sampling method each stratum.
The two main types of sampling method: probability methods and  Proportionate and disproportionate sampling If all the
non-probability methods strata are equally important to the survey, a proportionate
sample would be taken, i.e. the same number selected
from each stratum. Frequently, some strata are more
important to the research than others.

 Cluster sampling :Cluster sampling is a variation of

stratified random sampling and may be used when the
survey population is concentrated in a relatively small
number of groups (clusters) that are considered typical of
the market in question. A random sample of these clusters
is then taken. A random sample of units from within these
clusters is then taken.

Sampling frames : When using probability sampling methods it is

 Probability methods necessary to use a sampling frame. This is a list of every element
in the survey population. The sample is drawn from this list. A
Statistically speaking, these are the best types of sampling method sampling frame is essential for probability-based techniques, as
as each respondent has a known chance of being selected, so bias each element must have a known chance of selection, and so must
is minimized. They also allow the accuracy of the results to be be included in the sampling frame
estimated statistically.
A sampling frame must have the following characteristics:
The main types of probability sample are simple random sampling, ◆ Each element should be included only once.
systematic random sampling, stratified random sampling and ◆ No element should be excluded.
cluster sampling ◆ The frame should cover the whole of the population.
 Simple random sampling Items can be selected from the ◆ The information used to construct the frame should be up-to
sampling frame by using the lottery method, e.g. taking date and accurate.
numbers out of a hat. In the UK, ERNIE the computer ◆ The frame should be convenient to use.
selects Premium Bond winners, and does so by using Non-probability methods
simple random sampling. Random number tables are
generated by computer and often used in marketing With non-probability sampling methods, some element of
research. judgment enters the selection process. The extent to which
 Systematic random sampling: with larger samples it is judgment is used, and therefore the element of bias introduced,
more convenient to divide the population by the sample varies in these methods. Non-probability methods do not require a
size to calculate the sampling interval (n). A random sampling frame and the chance of each unit being selected is
starting point is selected using random number tables and unknown.
every n time after that is selected.
Statistical estimates of the size of the sampling error cannot
Example If the sample size is 50, and the population size is 3000, therefore be made.
then the sampling interval is calculated as:
 Non-probability methods
With non-probability sampling methods, some element of c) Mail (postal) questionnaire
judgment enters the selection process. The extent to which
judgment is used, and therefore the element of bias introduced, If the survey population is widely dispersed, it may be more useful
varies in these methods. Non-probability methods do not require a to send the questionnaire by mail than to have an interviewer call
sampling frame and the chance of each unit being selected is on the respondent.
unknown.
The advantage over the telephone interview with this method is
Statistical estimates of the size of the sampling error cannot that the questionnaire can be lengthy and ask for detailed
therefore be made. information.

 Convenience sampling: IV. Questionnaire design

Items are selected that are close or easily available. This
is useful in the exploratory stage of research, giving the The questionnaire is a vital part of most surveys and great care
researcher a ‘feel’ for the subject. Despite being very must be taken with its design.
cheap and quick to carry out, the level of error and bias
Many factors will affect the design of the questionnaire, such as
with this method is likely to be very high and so it should
the nature of the data required (qualitative or quantitative) and how
be used with caution.
the questionnaire is to be administered (by personal interview,
telephone, mail or other self-completion, or whether electronic
 Judgment sampling
instruments will be used).
-NItems are selected by the researcher that are felt to be
representative of the survey population. This method Questionnaires are notoriously difficult to construct, but Tull and
attempts to be more representative than convenience Hawkins (1997) suggest that a convenient way of tackling the
sampling. design is by breaking up the task into a number of decision areas,
- Experts also may be consulted for advice on which items namely:
are likely to be more appropriate for the survey. For ◆ preliminary decisions;
example, in a survey of textile manufacturers, a staff ◆ question content;
specialist such as a product developer may provide useful ◆ question wording;
advice on which manufacturers would be suitable for ◆ response format;
selection. ◆ question sequence;
◆ questionnaire layout;
 Quota sampling
◆ pretest and revise.
- This is the most likely non-probability method to
produce a representative sample as items selected are
based on known characteristics of the population. 1) Preliminary decisions

Example: Assume that your survey population has the following These include decisions on:
characteristics: - what information is required,
- who will be included in the survey
Age: 16–29: 26%; 30–64: 58%; 65 : 16% - how they will be contacted
Gender: Male : 48%; Female: 52%
2) Question content
6) Data collection methods This section is concerned with the content of individual questions:
If a survey is to be conducted, there are a number of methods what to include, rather than how to phrase the question. Points to
available for the collection of data and each has its relative consider include the following:
advantages and disadvantages.
 Is the question necessary? The first decision to be made
The three main traditional methods are by personal interview, here is whether or not the question is actually necessary.
telephone interview and mail questionnaire. If the question is not necessary for the purposes of
meeting the survey objectives, then leave it out.
a) Personal interview  How many questions are needed? If the question is
necessary, one must take care that the information you
Face-to-face interviewing is still the most widely used method of elicit will answer the question without ambiguity.
primary data collection.
For example, if you asked a respondent: ‘Do you think woollen
This method is labor-intensive and costly, but is more likely to trousers are comfortable and warm to wear?’ and the response
result in a satisfactorily completed questionnaire than any of the was ‘No’
other methods.
 Has the respondent the information to answer the
b) Telephone interviewing question? Sometimes respondents are asked questions on
subjects about which they are not informed.
Interviewing is done from a central location, cutting the costs of
Some respondents will attempt to answer questions without being
fieldwork considerably, providing the sample size is large. It is not
adequately informed, which will affect the validity of the results.
a cost-effective method for small samples.
  Is the respondent able to articulate the response? Even if 2) The Likert scale
the respondent has the necessary information to answer a
question, they may not always be able to articulate their Respondents are asked to indicate their level of agreement or
responses successfully. disagreement with a series of statements about a subject or an
Using aids such as descriptions and pictures makes it easier for the object.
respondent to answer the questions, and so complete the interview. The statements used are identified as either positive or negative,
and scores are allocated for particular responses.
 Asking questions beyond the memory span of the
2) The Likert scale
respondent Asking questions about behaviour over a long
time span may not produce accurate information. The list of possible responses is usually:
For example, asking the respondent how much they spent on tights ◆ strongly agree;
in the last year would result in an answer that was pure guesswork. ◆ agree;
Asking how much they spent on tights in the last fortnight would ◆ don’t know/neutral;
be more likely to provide accurate information. ◆ disagree;
◆ strongly disagree.
3) Question wording
Great care must be taken with the wording of questions. This is of
particular importance when conducting cross-cultural or 2) The Likert scale
international marketing research. Decisions about question
The Likert scale is not an interval scale, so it is not possible to
wording include:
infer that ‘strongly agree’ is twice as strong an attitude as ‘agree’.
+ Does the word mean the same to all respondents?
+ The use of vague or ambiguous words also should be avoided The scores achieved by individual respondents are only relative to
those achieved by other respondents. Likert scales are popular as
4) Response format they are easy to construct and give reliable information about the
degree of respondents’ feelings.
There are a number of types of response format that may be used.
The most commonly used are dichotomous, multiple-choice and
open-ended formats. Most questionnaires contain a mixture of
these

 Dichotomous Only two responses are allowed, such as

‘yes’ or ‘no’, ‘male’ or ‘female’. A neutral ‘don’t know’
category is sometimes included.
 Multiple-choice : Here the respondent is presented with a
choice of several possible answers to the question.
Frequently, the list of choices is shown to the respondent
on a card. The order of the alternative answers should be
rotated to avoid bias.
 Open ended The respondent has complete freedom of
choice in the response given with this format. This format
is often used where little information exists to construct a
multiple choice list, or when great detail is required.

V. Attitude measurement and rating scales

1) Types of attitude scale
 Interval scales

These scales have regular calibrations.

The advantage of this scale is that it can be used to measure the

strength of a particular attitude. It also allows the use of statistical
measures such as standard deviation, correlation coefficients and
significance testing.