Post on 26-Nov-2014
Definition of Marketing Research
It is the process of systematic collection and analysis of market information and data for the purpose of important decision making in marketing
Relevance Of Marketing Research Today
Today’s business decisions are extremely complex and a large variables are involved
Globalization and liberalization has intensified competition and survival has been at stake
Optimization of all levels to reduce costs
Relevance Of Marketing Research Today
Employees and shareholders are becoming increasingly aware of their rights to participate in decision making
More and more organizations are turning from selling to marketing concept
The tools used for research have increased and organizations are increasingly practicing data mining
What Constitutes Good Market Research
Clearly defining the problem and purpose Process of research well described Research design thoroughly planned Maintaining high ethical standards Limitations clearly revealed Adequate analysis of the data Findings presented unambiguously Conclusions justified Integrity of researcher
Why Market Research Sometimes Fails When the problem is not carefully designed
When researchers are less competent, not well trained and lack creativity
When sufficient time and finance is not provided
When there exists a lack of co-ordination and understanding between researchers and marketing managers
Pros & cons of In-house M R Activity
Better knowledge of the company, its history, background and current problems
Better control and liaison
Commitment may be higher
More assurance of confidentiality
Pros & cons of In-house M R Activity
Commissioning costs can be lower
Lack of objectivity due to close relations with clients
Friendship and company politics
Pros of Outsourced M R Activity
Good value
A wider skill base
Objectivity
Anonymity
Pros of Outsourced M R Activity
Superior facilities
Agency specialization
Less chance of bias since a fresh look is possible
Evaluation of external agency
Personal recommendation Recommendations by trade associations Experience and Qualifications of key
personnel Matching agency specializations of the
agency to the need Agency orientation towards client’s
decision makers
Evaluation of external agency
Facilities and resources of the agency Level of creativity and innovation in
previous research Time conscious and organization culture Stability and reliability of the agency
personnel
Qualities of a good Researcher
Must be innovative and ready to look at any problem with fresh perspective
Must have skill and knowledge and prepared to try out new ideas
Must be familiar with recent developments in industrial sectors, markets, products etc.
Must be familiar with and absorb recent trends in research techniques
Must have considerable selling skills to convince the top management or sponsor regarding the accuracy and soundness of recommendations
Research Proposal Defined
It is a researcher’s offer to render market research service to a potential buyer or a sponsor
Purpose Of Proposal
To present management question to be researched and its importance and relevance
To indicate the research efforts of other researchers on related management questions
To highlight the data necessary for solving the management questions
To present the methodology of the research To state the means of gathering and treating
the data and interpretation of the data
Exploration Exploration
Discover Management Dilemma
Define Management Question
Define Research Question
Refine Research Question
Research Proposal
Research Design
Design Strategy (type, purpose, time frame, scope)
Data CollectionDesign
SamplingDesign
Question & InstrumentPilot Testing
Instrument Revision
Data Collection& Preparation
Data Analysis& Interpretation
Report
ManagementDecision
Benefits Of Making Proposal
It helps the researcher to have a tentative work plan
Charts the logical steps needed to accomplish the stated goals
It helps the researcher to review the project steps
It helps the researcher to assess previous approaches to the problem and revise the plan accordingly
Benefits Of Making Proposal
It provides an opportunity to spot errors in assumptions and correct them if necessary
It gives the time and budget estimates It becomes a milestone for the sponsor
to review the progress of the study It serves as a logical guide for
investigation
Why Ethical Considerations in MR
• To ensure confidentiality for the sponsor organization
• To protect the rights of the respondents
• To ensure the client or sponsor organization receives genuine findings from the study
• To protect the safety of the researcher and the tea
• To ensure integrity in MR
More About Research Ethics
Ethical practices are the means to ensure that no one is harmed or suffers adverse consequences from the research activities
There is no single approach to ethics
Strict adherence to a set of laws is not practical since researcher is subjected to unforeseen constraints
More About Research Ethics
Many design based ethical problems can be eliminated by careful planning and constant vigilance
Responsible research anticipates ethical dilemmas and adjusts the design and procedures to overcome them
Be fair and frank about the limitations of research with the client as also respondents
Ethical practices towards respondents
Before starting data collection, explain the benefits to the respondents. Do not overstate or understate the benefits
Assure that the rights and well being of respondents will be protected. Show by destroying names etc.
Take informed consent of the respondents
Clearly state that sensitive questions may not be answered
Survey Research
The communication approach involving questioning people and recording their responses for analysis
It is the most versatile approach for the collection of abstract information of all types
An economical and efficient means of learning opinions and attitudes, intentions and expectations of respondents
Survey Research
It depends greatly upon the ability and willingness of respondents
Interpretation of the question by the respondent can greatly vary their responses
Observation Survey
A method of collecting past and current data by recording observation
It can be used to collect data even when respondents are hostile or unable to give information
Data can be collected as it occurs in its natural environment
Observation Survey
It can record the information which most respondents would ignore
It is less demanding and has less bias involved
It is a slow and expensive process and involves interpretation skills and needs to be used with care and understanding
Questionnaire Design
A researcher first needs to answer 4 basic questions –
What type of data is needed to answer management questions?
What communication approach will be used? Should the questions be structured,
unstructured or a combination? Should the questioning be undisguised or
disguised and if so, to what degree?
Forms of Questionnaire
Structured : It is a formal list of questions framed to generate facts
Unstructured : It is one in which questions are not structured and the order in which they are to be asked is left to the researcher
Disguised : A questionnaire where the object of inquiry is not revealed to the respondent. It can be structured, or unstructured
Undisguised : It is one where the object of inquiry is revealed to the respondent
Types of Questions
Open ended questions : They give respondent complete freedom to decide the form, length and details of the answer
Dichotomous questions : They have only 2 answers in the form of yes or no, true or false etc.
Multiple choice questions : Here the respondent is offered two or more choices and the respondent has to indicate the one applicable in his case.
Phrasing and Ordering of Questions
Difficult words should be avoided as far as possible
Lengthy questions should be avoided
Two or more questions should not be combined into one
Phrasing and Ordering of Questions
Questions lacking specificity should be avoided
Simple questions should be asked in the beginning of the interview
Specialized questions needing in-depth information should be left at the end
Characteristics of informed consent Respondent must be capable of giving
free consent
Consent must be voluntary, free from coercion and unethical practices
Consent must be adequately informed to make a decision
Respondents must be told of the risks involved if any
Classification of Research Designs
The degree to which the question has been crystallized (exploratory or formal)
The method of data collection (observational or communicative)
Purpose of the study (finding who, what, when and why)
The time dimension (once or repetitive study)
The topical scope (case study or statistical study)
Exploratory Research It is a preliminary investigation where the
researcher is not sufficiently knowledgeable
The focus is to bring out different ideas relating to the management problem
It is generally based on readily available secondary data
It is an informal and unstructured design
It is independent of the size of the research
Exploratory Research
A researcher is asked for a help by the new management of an oldest and most reliable bank concerned about the erosion of profitability
Management Dilemma: How can we improve our profit?
Exploratory Research
This management Dilemma does not specify what kind of research needs to be done
This question is broad and is strictly managerial in thrust
It does not indicate the route to be taken by the bank to increase profit
Exploratory Research
The route to be taken by the bank to increase profit may be via increase in deposits, increase in loan disbursements, downsizing the manpower, outsourcing less value adding jobs outside, or by some other means
Exploratory Research
Now the management question becomes a two sub-question based:
How can we improve deposits?
How can we reduce costs?
Exploratory Research
A small focus group is conducted among employees and data is analyzed to compare financial and operating statistics with profitable banks.
It becomes obvious that the bank operations are not as progressive as its competitors’ and has its costs well in line.
Exploratory Research
Now the revised management question becomes:
What should be done to make the bank more competitive?
Exploratory Research
Now the Research question becomes:
Should the bank position itself as a modern, progressive institution with appropriate changes in services or policies
Or
Maintain its image as the oldest, most reliable institution
Exploratory Research
The investigative questions become:
1. What is the public’s position regarding financial services and their use?
a. What specific financial services are used?
b. How attractive are various services?c. What bank-specific and environmental-specific factors
influence a person’s use of a particular service?
Exploratory Research
2. What is the bank’s competitive position?a. What are the geographic patterns of our customers
and of our competitors’ customers?b. What are demographic differences between our and
competitors customers?c. How aware is the public of the bank’s promotional
efforts?d. What opinion does the public hold of the bank and its
competitors?e. How does growth in services compare among competing institutions?
Focus Group Discussion It is a formal discussion between representative
samples on a particular subject Ideally, the group should consist of 8-10
representatives The representatives should be homogeneous to
have a meaningful discussion It should be conducted by an experienced specialist
whose role is of a catalyst FGD is fast and cheap and is the method for
generating the hypothesis They are flexible in approach, direct and easily
understandable to the client The limitation is that the sample may be inadequate
since it may be drawn on the basis of convenience
Advantages of F G D
Ability to quickly and inexpensively grasp the core issues of a topic
They are brief and extremely flexible Provide an opportunity to observe reactions
to the research questions in an open ended group setting
They bring out surprise information and new ideas
Descriptive Research
• The focus of this research is to answer the “who, what, when, where and how” of the management dilemma
• These studies are factual and can be complex at times
• They are formal and well structured and hence the research process should be well defined in the planning stage
• Descriptive research can be of 2 types –• Cross sectional studies which are concerned with a
sample of elements from a given problem• Longitudinal studies which are based on panel data
and panel methods
Descriptive Research
In the bank example, we may be interested in developing a profile of savers by analyzing data:
• Size of accounts• No. of accounts for minors• No. of accounts opened in last six months• Amount of activity in the account• Gender of account holders• Nearness to workplace/home
Causal Research
John Stuart Mill formulated a set of principles for Causal research
When two or more cases of a given phenomenon have one and only one condition in common then that condition may be regarded as the cause of the phenomenon
Causal Research
A B C Z
D EC Z
ZC
Causal Research
Advantage: Helps ruling out irrelevant factors
Limitation: Other factors not considered may be influencing the phenomenon
Causal Research
John Stuart Mill improved the logic by having a negative canon
If a change in the amount of one variable is accompanied by a comparable change in the amount of another variable in two or more cases, and the latter change does not occur in the absence of the first change, one change is the cause of the other
Causal Research
A B C Z
NonC
NonZ
ZC
A B
Causal Research It investigates the cause and effect relationship
between 2 or more variables The study is based on the 3 possibilities viz
symmetrical, reciprocal and asymmetrical relationships
A symmetrical relationship exists when 2 variables fluctuate together in spite of being independent of one another
A reciprocal relationship exists when 2 variables mutually influence or reinforce each other
Asymmetrical relationship exists when changes in the dependent variable are responsible for changes in the independent variable.
Causal Research Stimulus-response relationship: Represents an event
that results in a response from object e.g. price rise results in sales reduction, change in Govt. policy results in company performance
Property-disposition relationship: A property is an enduring characteristic of a subject that does not depend on circumstances for its activation. Age, gender, family status ethnic group are personal properties. A disposition is a tendency to respond in a certain way under certain circumstances. It includes attitudes, opinions, habits and drives. E.g. effect of age on saving, social class and taxation
In-depth Interview It involves a detailed probing of the respondent by
specialist interviewer An informal approach where appropriate
questions are developed during the interview The focus is to identify and record subtle reactions
of the respondent It discovers the hidden motivations affecting
consumer behaviour It provides a strong stimulus to the insights of the
researcher Being unstructured, the information can be
subjective leading to different interpretations by different researchers of the same information
Sampling
Sampling – It is the process of selecting some elements in the population which are true representatives of the population
Population – It is the total collection of elements about which we wish to make some inferences
Census – It is a count of all elements in a population
Sampling
Good sample – It is one which well represents the characteristics of the population it purports to represent
It must have accuracy that is, the degree to which bias is absent from the sample and precision that is, precision of estimate. The smaller the standard error of estimate, the higher is the precision of sample
Sample Size – It is a function of the variation in the population parameters under study
Decide the target population
Develop aframe
ChooseSamplingtechnique
Decide the sample size
Choose sampling methods
Sampling Frame
Decide on each type of element that will be covered to represent the population
Sampling Technique
Probability based i.e. statistical
Non-probability i.e. judgmental sampling
Sample Size
Calculate the standard deviation of population
Decide the sampling error acceptable to the specific study under consideration
On the basis above, derive a sample size arithmetically
Sampling Methods
Simple random – It is the method in which each element of the population has an equal chance of selection
Systematic sampling – Here, every kth element in the population is sampled, beginning with a random start of an element in the range of 1 to k
Sampling Methods
Stratified random sampling – Here the population is divided into mutually exclusive and exhaustive strata and then a simple random sample is selected within each of the strata
Cluster sampling – It is one where entire group or clusters are selected at random instead of selecting individual units from the population
Sampling Methods
Quota sampling involves fixation of certain quotas to be fulfilled
For example, we want to conduct a survey of households numbering 2,00,000 in totality
A quota of 1% is fixed i.e. 2000 households
1200 in rural area and 800 in urban
Sampling Methods
Rural 1200 Rich 150
Urban 800 Middle class
650
Poor 1200
Total 2000 Total 2000
Independent Controls
Sampling Methods
Inter-related Controls
Rural Urban Total
Rich 100 50 150
Middle class
400 250 650
Poor 700 500 1200
Total 1200 800 2000
Advantages of Quota Sampling
Economical
Administratively convenient
Time saving
Suitable when a sampling frame is not available
Disadvantages of Quota Sampling
Calculation of standard error may be difficult
May not be representative
Quality may suffer
Bias creeps in into the process
Other forms of Judgmental Sampling
Master sample
Panel sample
Convenience sample
Types Of Scales
Nominal scales : Numbers serve as labels to identify persons, objects or events. The identification can be made on the basis of a certain characteristic or attribute. Here, a set is split into subsets which are mutually exclusive and collectively exhaustive
Ordinal scales : These are ranking scales which besides having the unique characteristics of the nominal scale, indicate the order by distinguishing elements on the basis of single direction
Types Of Scales
Interval scale : They are the combination of nominal and ordinal scales with additional strength of determination of the equality of differences.
Ratio scales : They possess order, distance and unique origin and are the most superior amongst all scales.
Criteria for Ascertaining Good Scale
Reliability : It is the extent to which repeat measurements made by it under constant conditions will yield the same result.
Validity : It is the success of the scale in measuring what it sets out to measure, so that the difference between individual’s scores can be taken as representing true differences in the characteristic under study.
Attitude Scales
Paired comparison scale : It requires the respondent to select two objects of a set with regard to an attribute of interest
Thurstone scale : It is a technique in which a large number of statements in a random order are administered to one or more respondents in the form of a questionnaire. Depending upon the acceptance or rejection of a statement by the respondent, his scale score is determined by averaging the scale values for the statements accepted by him
Attitude Scales
Summated rating scale : Researcher collects a large number of
statements and eliminates those which are ambiguous, irrelevant or deficient
The remaining statements are administered to one or more respondents for their reaction using a five point rating system i.e. strongly approve, approve, undecided, disapprove, and strongly disapprove.
These categories are assigned value 5,4,3,2 and 1 respectively
Attitude Scales
In case of negatively worded statements, this scoring is reversed
The correlation between the statement scores and the total score is ascertained
Statements with a high correlation with total score are selected for final scale
Attitude Scales• Scalogram analysis : This method of scaling is
based on the assumption that, an individual with a more favourable attitude score than another must be just as favourable or more favourable in his response to every statement in the set than the other individual
• Semantic differential : This technique involves three dominant factors viz. evaluation, potency and activity factor. It is a bi-polar scale conforming to the basic concept of motivation, attraction or repulsion from an object
Data Processing
• Editing : It is the process of identifying, examining errors and omissions in the collected data and making necessary corrections. The errors can be in the form of inconsistency in the responses, partial or a vague response
Data Processing
• Coding : It is the procedure of classifying the answers to a question into meaningful categories to carry out subsequent operations of tabulating and analysis. In the absence of coding, it will be impossible to reduce a large number of heterogeneous responses into meaningful categories resulting in a weak and ineffective data analysis.
Data Processing
• Tabulation : It comprises sorting of the data into different categories and counting the number of cases that belong to each category. Tabulation can be –
Univariative Bivariative Multivariative
Report Writing
• The findings and interpretation of the research study needs to be reported to the management or the sponsor in the form of a formal report, which should ideally be in the form of a written document
• Principles of good report : Report must be written objectively Written in a concrete and impersonal style Organized and coherent Absolute clarity of ideas Precise, to the point, neat and tidy in approach
Report Format• An ideal market research report follows the following
format – Title page Name of sponsor Index of contents Introduction Methodology Statistical tables, charts and illustrations Findings Limitations Summary and conclusions Executive summary/recommendations Appendices and bibliography
Statistical Methods for Data Analysis
• Correlation – It is a statistical technique used for measuring
the relationship or interdependence of two or more variables
A graph of the two series in question is drawn, known as cattle diagram
Values of two variables are plotted such that a point on the graph indicates a pair of values
Independent variable is taken on the X axis and the dependent on the Y axis
Statistical Methods for Data Analysis• Regression analysis – It is a technique used for estimating or predicting
the value of one variable given the value of another.
When relationship between two variables is linear, a change in the independent variable is followed by a constant absolute change in the dependent variable.
The model does not necessarily imply any cause and effect relationship between two variables.
The relationship is given by the equation Y=a+bX
Statistical Methods for Data Analysis
• Cluster analysis – Used to classify persons or objects into a small
number of mutually exclusive and exhaustive groups.
There needs to be high homogeneity within the cluster and high heterogeneity between clusters.
Used extensively in marketing research for resolving the problem of classifying consumers, products etc.
Applications of Marketing Research contd
• Advertising effectiveness –Recognition study Carried out for printed advertisement and referred
to as a readership study. Respondents are shown a particular page of the
magazine and the following measures of recognition are taken –
o Noted : The percentage of readers who have seen the ado Seen – associated : The percentage of readers who read a
part of the ad indicating the brando Read must : Percentage of readers who read a major part of
the ad Scores are assigned to these and related to the
expenditure incurred.
Applications of Marketing Research
• Advertising effectiveness –Recall test – Respondents are asked to recall specifics of an
ad Copies of test magazines are sent to a sample of
respondents for normal reading Respondents are read out a list of ads and
asked to identify those they remember and the extent to which they can recall
Scores are assigned to the ability of the respondent to remember name of the product, underlying message, and favourable attitude.
Applications of Marketing Research
• Advertising effectiveness –Sales test – This method measures the effect of an ad
on sale of the products Assumption is that, changes in sales are a
result of ad Isolate the influence of other factors while
determining the impact of an ad
Experimental Design Experimental Treatment – It is a study of alternatives
whose effects are to be measured and compared
Test units – The individuals or organizations whose response to the experimental treatment is being studied
Experimental Error – A measure of the apparent variation in performance of test units treated alike by the experimenter after the effect of extraneous forces is removed from the data.
Extraneous forces – All other forces and conditions which affect the response of test units to experimental treatments.
Experimental Design It originated in the field of agricultural research
It is the process in which the researcher intervenes to manipulate the indirect or exploratory variable in a setting and observe how it affects the subject being studied
There must be a complete agreement of independent and dependent variables
The 2 variables must occur almost simultaneously
Other extraneous variables should not influence the dependent variable
Advantages of Experimental Design
Researcher’s ability to manipulate the independent variable
Effect of extraneous variables can be controlled more effectively
Convenient and cost effective design
Replication of an experiment with different subject groups and conditions is possible
Disadvantages of Experimental Design Artificiality of the laboratory
Generalization from non probability samples can pose problems despite random assignment
Despite lower costs of experimentations, many applications of experimentation far outrun the budgets for other primary data collection methods
Experimental studies of the past are not feasible
There are limits to the type of manipulation and controls that are ethical