When a marketing manager has to select one course of action among a number of alternatives, conclusive research provides him information that helps him to evaluate various alternatives and select among them a course of action. This type of research provides a rational basis for his decisions. The alternatives may be well or vaguely defined. Conclusive research design is characterized by formal research procedures. The research objectives are accurately defined and so are information needs. Conclusive research studies can be classified either as descriptive or experimental.
Descriptive Research:
The research objectives in this type of research are generally describing the features of consumer segment that include demographic, socio-economic, psychographic and benefits sought for. Descriptive studies can also portray buyer perceptions of brands; audience profiles for media types, viz., TV, radio, newspaper, journals and magazines, etc. They can also portray buying power of consumers, availability of distributors, product consumption patterns, price sensitivity of consumers, market share, etc. These are just a few representative studies out of numerous studies that come under descriptive research in marketing. Despite the emphasis on description, it should not be concluded that the studies should be simply fact-gathering expeditions. They can also be used to make predictions about the occurrence of a marketing phenomenon. The data regarding the presence of an association among variables can only be used for productive purposes but statements regarding cause and effect relationships are not possible with descriptive research (As may be possible in experimental or causal research).
The purpose and nature of descriptive research is quite different from that of exploratory research. Many descriptive studies are made with only hazy objectives and with inadequate planning. Much of the data collected in such studies turns out to be useless. Descriptive studies of this type are actually more of exploratory type.
Effective descriptive research is marked by a clear statement of the problem, research objectives and information needs described in detail. The research design should be fairly structured. Since the purpose is to provide information regarding specific questions or hypothesis, the research must be designed to ensure accuracy of the findings. Since descriptive studies may cost huge amount of money to carry out, there is then this necessity of its formal design.
Descriptive research often makes use of survey research design which consists of a cross-sectional research design. It is collecting data on few factors from a number of cases at one point of time. This is the most popular type of research design and is useful in describing the characteristics of consumers and determining the frequency of marketing phenomenon. It is often expensive and requires skillful and competent market researchers to conduct it effectively. It is also termed as statistical method in contrast to case method which focuses on many factors of few cases. This method ceases to focus on individual cases and focuses instead on classes, averages, percentages, measures of dispersion and more sophisticated statistical procedures. For example, cluster analysis can be used to group customers into different classes on the basis of few customer attributes or characteristics. Factor analysis can be used to group attributes into few factors which are important from consumers’ point of view vis-à-vis a product, etc. In fact descriptive research presupposes that a sound causal model of marketing system exists in the mind of the decision-maker in contrast to exploratory research which seeks to generate hypotheses. A market researcher has to have a tentative hypothesis for carrying out descriptive research for the sole reason that he knows what data to collect from his respondents in the survey.
Descriptive research designs can also use one or more of the following sources of data:
(a) Interrogation of respondents
(b) Secondary data
(c) Simulation
The analysis techniques used in descriptive research are those specifically for mass data. Each individual item tends to lose its identity. This is both an advantage and a disadvantage. The advantage lies in the objectivity with which the analysis can be made. Averages or variance can be computed and compared; two independent researchers will arrive at similar results, which is not the case with the case method. But the disadvantage lies in its mobility to prove cause and effect relationships which is the domain of experimental (causal) research. Even the direction of causal effect may not be visible through statistical study, e.g. when advertising and sales co-vary, it is often not clear whether advertising causes sales or sales cause the expenditure of more advertising effort because of greater apparent potential sales results.
Simulation of marketing phenomenon consists in an incomplete representation of the marketing system or some aspect of this system. It is relatively a new source of data, which is largely computer-oriented. Simulation can be used to gain insight into the dynamics of the marketing system by manipulating the independent variables (marketing mix and situational variables) and observing their influence on the dependent variables.
A marketing simulation requires data inputs regarding the characteristics of the phenomenon to be represented and the relationships present. Simulations should neither over-simplify nor over-complicate. The limitations are its validity and time/cost of updating the model as conditions change. Simulation models can be classified on the basis of the purpose they serve, viz., predictive and descriptive.
Very often descriptive research may take the route of analyzing already existing data on respondents, e.g., census data. Secondary source of data is much cheaper to access than carrying out a full-fledged survey and hence a great temptation for a market research. But the validity and appropriateness of such a secondary data descriptive research should always be borne in mind before embarking on one.
Experimental or Causal Research
Although, it is the nature of marketing decision-making that all the conditions allowing the most accurate causal statements are not usually present but in these circumstances, causal inference will still be made by marketing managers. In doing so they would desire to be able to make causal statements about the effects of their activities. For example a new advertising campaign developed has resulted in a certain % increase in sales or the sales discount strategy followed has resulted in % increase in sales, etc. In both of these cases, marketing managers are making a causal statement.
However, the scientific concept of causality is complex and differs substantially from the one held by the common person on the street. The common sense view holds that a single event (the ‘cause’) always results in another event (the ‘effect’) occurring. In science, an event has a number of determining conditions or causes, which act together to make the effect probable. Note that in the common sense notion of causality, the effect always follows the cause. This is deterministic causation in contrast to scientific notion, which specifies the effect only as being probable. This is termed as probabilistic causation. The scientific notion holds that we can only infer causality and never really prove it. That is, the chance of an incorrect inference is always thought to exist. The world of marketing fits the scientific view of causality. Marketing effects are probabilistically caused by multiple factors and we can only infer a causal relationship.
The conditions under which we can make causal inferences are:
(a) Time and order of occurrence of variables.
(b) Concomitant variation.
(c) Deletion of other possible causal factors.
The fundamental research tool used to identify causal relationships is the experiment. The objective of an experiment is to measure the effect of explanatory on a dependent variable, while controlling for other variables that might intervene one’s ability to make causal inferences. Such research leads to questions like:
“Is a certain print advertisement more effective in color than black and white?”
“Which of the different technique of promotion is most effective in selling a particular product?”
“Can we add the sales of a product by getting additional shelf space?”
The utility of experimental design in marketing extends across the functional areas of sales promotion, distribution, price determination and establishing product policies. Whenever marketing management is interested in measuring the effects of alternative courses of action, experimentation may be a practical means of reducing the risk involved in deciding among the alternatives.
An experiment is executed when one or more independent variables are consciously controlled by the person running the experiment and their effect on the dependent variable or variables is measured. In surveys there is no manipulation of independent variables by the researchers. This is a fundamental difference between the experimental and non-experimental research.
The research objectives in this type of research are generally describing the features of consumer segment that include demographic, socio-economic, psychographic and benefits sought for. Descriptive studies can also portray buyer perceptions of brands; audience profiles for media types, viz., TV, radio, newspaper, journals and magazines, etc. They can also portray buying power of consumers, availability of distributors, product consumption patterns, price sensitivity of consumers, market share, etc. These are just a few representative studies out of numerous studies that come under descriptive research in marketing. Despite the emphasis on description, it should not be concluded that the studies should be simply fact-gathering expeditions. They can also be used to make predictions about the occurrence of a marketing phenomenon. The data regarding the presence of an association among variables can only be used for productive purposes but statements regarding cause and effect relationships are not possible with descriptive research (As may be possible in experimental or causal research).
The purpose and nature of descriptive research is quite different from that of exploratory research. Many descriptive studies are made with only hazy objectives and with inadequate planning. Much of the data collected in such studies turns out to be useless. Descriptive studies of this type are actually more of exploratory type.
Effective descriptive research is marked by a clear statement of the problem, research objectives and information needs described in detail. The research design should be fairly structured. Since the purpose is to provide information regarding specific questions or hypothesis, the research must be designed to ensure accuracy of the findings. Since descriptive studies may cost huge amount of money to carry out, there is then this necessity of its formal design.
Descriptive research often makes use of survey research design which consists of a cross-sectional research design. It is collecting data on few factors from a number of cases at one point of time. This is the most popular type of research design and is useful in describing the characteristics of consumers and determining the frequency of marketing phenomenon. It is often expensive and requires skillful and competent market researchers to conduct it effectively. It is also termed as statistical method in contrast to case method which focuses on many factors of few cases. This method ceases to focus on individual cases and focuses instead on classes, averages, percentages, measures of dispersion and more sophisticated statistical procedures. For example, cluster analysis can be used to group customers into different classes on the basis of few customer attributes or characteristics. Factor analysis can be used to group attributes into few factors which are important from consumers’ point of view vis-à-vis a product, etc. In fact descriptive research presupposes that a sound causal model of marketing system exists in the mind of the decision-maker in contrast to exploratory research which seeks to generate hypotheses. A market researcher has to have a tentative hypothesis for carrying out descriptive research for the sole reason that he knows what data to collect from his respondents in the survey.
Descriptive research designs can also use one or more of the following sources of data:
(a) Interrogation of respondents
(b) Secondary data
(c) Simulation
The analysis techniques used in descriptive research are those specifically for mass data. Each individual item tends to lose its identity. This is both an advantage and a disadvantage. The advantage lies in the objectivity with which the analysis can be made. Averages or variance can be computed and compared; two independent researchers will arrive at similar results, which is not the case with the case method. But the disadvantage lies in its mobility to prove cause and effect relationships which is the domain of experimental (causal) research. Even the direction of causal effect may not be visible through statistical study, e.g. when advertising and sales co-vary, it is often not clear whether advertising causes sales or sales cause the expenditure of more advertising effort because of greater apparent potential sales results.
Simulation of marketing phenomenon consists in an incomplete representation of the marketing system or some aspect of this system. It is relatively a new source of data, which is largely computer-oriented. Simulation can be used to gain insight into the dynamics of the marketing system by manipulating the independent variables (marketing mix and situational variables) and observing their influence on the dependent variables.
A marketing simulation requires data inputs regarding the characteristics of the phenomenon to be represented and the relationships present. Simulations should neither over-simplify nor over-complicate. The limitations are its validity and time/cost of updating the model as conditions change. Simulation models can be classified on the basis of the purpose they serve, viz., predictive and descriptive.
Very often descriptive research may take the route of analyzing already existing data on respondents, e.g., census data. Secondary source of data is much cheaper to access than carrying out a full-fledged survey and hence a great temptation for a market research. But the validity and appropriateness of such a secondary data descriptive research should always be borne in mind before embarking on one.
Experimental or Causal Research
Although, it is the nature of marketing decision-making that all the conditions allowing the most accurate causal statements are not usually present but in these circumstances, causal inference will still be made by marketing managers. In doing so they would desire to be able to make causal statements about the effects of their activities. For example a new advertising campaign developed has resulted in a certain % increase in sales or the sales discount strategy followed has resulted in % increase in sales, etc. In both of these cases, marketing managers are making a causal statement.
However, the scientific concept of causality is complex and differs substantially from the one held by the common person on the street. The common sense view holds that a single event (the ‘cause’) always results in another event (the ‘effect’) occurring. In science, an event has a number of determining conditions or causes, which act together to make the effect probable. Note that in the common sense notion of causality, the effect always follows the cause. This is deterministic causation in contrast to scientific notion, which specifies the effect only as being probable. This is termed as probabilistic causation. The scientific notion holds that we can only infer causality and never really prove it. That is, the chance of an incorrect inference is always thought to exist. The world of marketing fits the scientific view of causality. Marketing effects are probabilistically caused by multiple factors and we can only infer a causal relationship.
The conditions under which we can make causal inferences are:
(a) Time and order of occurrence of variables.
(b) Concomitant variation.
(c) Deletion of other possible causal factors.
The fundamental research tool used to identify causal relationships is the experiment. The objective of an experiment is to measure the effect of explanatory on a dependent variable, while controlling for other variables that might intervene one’s ability to make causal inferences. Such research leads to questions like:
“Is a certain print advertisement more effective in color than black and white?”
“Which of the different technique of promotion is most effective in selling a particular product?”
“Can we add the sales of a product by getting additional shelf space?”
The utility of experimental design in marketing extends across the functional areas of sales promotion, distribution, price determination and establishing product policies. Whenever marketing management is interested in measuring the effects of alternative courses of action, experimentation may be a practical means of reducing the risk involved in deciding among the alternatives.
An experiment is executed when one or more independent variables are consciously controlled by the person running the experiment and their effect on the dependent variable or variables is measured. In surveys there is no manipulation of independent variables by the researchers. This is a fundamental difference between the experimental and non-experimental research.
Treatments are independent variables that are manipulated or whose effects are measured. Dependent variables are the measures taken on test units. Test units are the entities to whom the treatments are presented and whose response to the treatments is measured. It is common in marketing for both people and physical entities, such as stores or geographic areas, to be used by as test units. For example, people may be asked to try a product and then have their attitudes towards it measured. Here people are test units, product type is the independent variable and attitude is the dependent variable.
Two concepts of validity are important in experimentation, internal and external. Internal validity is concerned by with the question of whether the observed effects on the test units could have been caused variables other than the treatment. Without internal validity the experiment is confounded. External validity is concerned with the general application of experimental results. A researcher, obviously, would like an experimental design to be strong in both kinds of validity. Unfortunately, it is often necessary to trade off one type of validity for another. There are three true experimental designs in which a researcher is able to eliminate all extraneous variables as competitive hypotheses to the treatment:
(1) The Pre-test—Post-test Control Group Design:
Experimental Group R O1 X1 O2
Control Group R O3 O3
(2) The Solomon Four Group Design:
Experimental Group 1 R O1 X O2
Control Group 1 R O3 O4
Experimental Group 2 R X O5
Control Group 2 R O6
(3) Post-test only Control Group Design:
Experimental Group R X O1
Control Group R O2
Symbols used:
X Represents exposure of a test group to an experimental treatment, the effects of which are to be determined.
O Refers to processes of observation or measurement of the dependent variables on the test units.
R Indicates that individuals have been assigned at random to separate treatment groups or that groups themselves have been allocated at random to separate treatments.
Movement from left to right indicates movement through time. All symbols in any one row refer to a specific treatment group. Symbols that are vertical to one another refer to activities or events that occur simultaneously. O1 X1O2 indicates that one group received or measurement of the dependent variable both prior to (O1) and after (O2) the presentation of treatment (X1). Further the symbols.
RX1 O1
RX 2O1
Indicate that two groups of subjects were randomly assigned to two different treatment groups at the same time. Further, the groups received different experimental treatments and the dependent variables were measured in the two groups at the same time. These purely experimental research designs control extraneous variables like history, maturation, interactive testing effect, instrumentation, statistical regression, and selection bias and test unit mortality.
PROCESS OF PLANNING AN EXPERIMENTAL DESIGN:
(i) Selection of the Problem: Every problem cannot be studied through experimental method. The problem which can be studied through experimental method has to be selected.
(ii) Proper Description of the Selected Problem: After selecting the problem, it must be put in proper language, i.e., the hypothesis must be stated in clear and conceptual terms. The variables that affect the phenomenon must be known and conceptualized.
(iii) Selecting the Setting: The background in which the experiment relating to phenomenon is to be carried out is termed as setting. In case of laboratory experiment it is created artificially and the experimenter decides how it can be done. In case of a field experiment, natural setting has to be located where the experiment can be made.
(iv) Pilot Study: In planning an experiment, a pilot study may be necessary so that the researcher is brought face to face with realities and many problems that he had not thought of. This also will enable him to know more precisely the various causative factors involved, the nature and working of the institution, the extent of co-operation or resistance that he is expected to meet.
(v) Research Design: The most vital past of the research is research design as it lays down the manner in which the researcher will manipulate the situation in order to study the desired effect. This in itself leads to problem or control over the phenomenon.
Two concepts of validity are important in experimentation, internal and external. Internal validity is concerned by with the question of whether the observed effects on the test units could have been caused variables other than the treatment. Without internal validity the experiment is confounded. External validity is concerned with the general application of experimental results. A researcher, obviously, would like an experimental design to be strong in both kinds of validity. Unfortunately, it is often necessary to trade off one type of validity for another. There are three true experimental designs in which a researcher is able to eliminate all extraneous variables as competitive hypotheses to the treatment:
(1) The Pre-test—Post-test Control Group Design:
Experimental Group R O1 X1 O2
Control Group R O3 O3
(2) The Solomon Four Group Design:
Experimental Group 1 R O1 X O2
Control Group 1 R O3 O4
Experimental Group 2 R X O5
Control Group 2 R O6
(3) Post-test only Control Group Design:
Experimental Group R X O1
Control Group R O2
Symbols used:
X Represents exposure of a test group to an experimental treatment, the effects of which are to be determined.
O Refers to processes of observation or measurement of the dependent variables on the test units.
R Indicates that individuals have been assigned at random to separate treatment groups or that groups themselves have been allocated at random to separate treatments.
Movement from left to right indicates movement through time. All symbols in any one row refer to a specific treatment group. Symbols that are vertical to one another refer to activities or events that occur simultaneously. O1 X1O2 indicates that one group received or measurement of the dependent variable both prior to (O1) and after (O2) the presentation of treatment (X1). Further the symbols.
RX1 O1
RX 2O1
Indicate that two groups of subjects were randomly assigned to two different treatment groups at the same time. Further, the groups received different experimental treatments and the dependent variables were measured in the two groups at the same time. These purely experimental research designs control extraneous variables like history, maturation, interactive testing effect, instrumentation, statistical regression, and selection bias and test unit mortality.
PROCESS OF PLANNING AN EXPERIMENTAL DESIGN:
(i) Selection of the Problem: Every problem cannot be studied through experimental method. The problem which can be studied through experimental method has to be selected.
(ii) Proper Description of the Selected Problem: After selecting the problem, it must be put in proper language, i.e., the hypothesis must be stated in clear and conceptual terms. The variables that affect the phenomenon must be known and conceptualized.
(iii) Selecting the Setting: The background in which the experiment relating to phenomenon is to be carried out is termed as setting. In case of laboratory experiment it is created artificially and the experimenter decides how it can be done. In case of a field experiment, natural setting has to be located where the experiment can be made.
(iv) Pilot Study: In planning an experiment, a pilot study may be necessary so that the researcher is brought face to face with realities and many problems that he had not thought of. This also will enable him to know more precisely the various causative factors involved, the nature and working of the institution, the extent of co-operation or resistance that he is expected to meet.
(v) Research Design: The most vital past of the research is research design as it lays down the manner in which the researcher will manipulate the situation in order to study the desired effect. This in itself leads to problem or control over the phenomenon.
No comments:
Post a Comment