Deduction is the process of reasoning from general to specific points. A specific conclusion is drawn from general principles and premises. Microeconomics is dominated by deductive models in that conclusions are reached concerning single firms, industries, or persons on the basis of general theories. Deduction is in fact a more theoretical approach than induction, because the beginning of a deductive argument is a general theory that has been proven accurate in the past.

A simple example of deduction follows. Every business manager is profit oriented. William is a business manger. Therefore, William is profit oriented.

The first statement in the example is a general statement, theoretical in nature, yet based on previous results of empirical research. The researcher studying William in this case did not necessarily do the research concerning all business managers, he might be simply using other people’s results and applying them to his particular problem. This is the reason scientists build up a body of theory, so that others can use it and not have to repeat the work each time a similar situation arises. The second statement in the example comes directly from the observations of the researcher. It is between the second and third statements that the deductive reasoning has taken place. To be valid the third statement must seem ‘logical’ to other people in that it must appear to follow directly from the previous statements.

Induction

Induction is the process of reasoning from a set of particular cases to a general conclusion. The general conclusion refers to all similar cases, even though all cases cannot be tested directly. Macroeconomics is dominated by inductive-empirical studies. The fact that induction uses empirical evidence as the beginning of the reasoning process makes it a more applied approach than deduction.

Induction and inductive reasoning are based on the premise that nature is orderly and uniform. In this way, general conclusions can be drawn from incomplete data. The distinctive feature of an inductive argument, however, is that there is at least one statement that will yield a false conclusion. An example of induction will illustrate this point:

1. Whenever there is an economic recession, Company A reduces its inventory.

2. Whenever there is an economic recession, Company B reduces its inventory.

3. Whenever there is an economic recession, Company C reduces its inventory.

Conclusion: Whenever there is an economic recession , all firms reduce their inventories.

The variable nature of human behavior makes both induction and deduction very difficult. The conclusions reached through both types of reasoning can rarely be phrased in absolute terms and remain accurate for long.

In a research project the deductive form of reasoning says that each statement should have in the back of it a justification in terms of a fact that has been observed as part of the research, a clearly understood assumption, or an accepted rule of logic. The deductive process in empirical research ideally is analogous to the plane geometry proofs of high school, except that empirical facts usually form some of the premises in the former case.

Deductive arguments in some areas of economics dealing with pricing come close to this ideal through the use of diagrams and mathematical reasoning.

In a deductive system, all propositions will be consistent with each other. If the conclusion is at odds with some other accepted proposition or principle, the researcher needs to seek further explanation.

With an inductive argument the generalization may be about a single entity, about a set of objects, or about casual relations and explanatory hypotheses. Research arguments or reasons are developed in support of a conclusion in the inductive method. It is not possible to establish any purely formal test for the adequacy of an inductive explanation.

Reasoning by analogy may be either deductive, inductive, or a combination of both. As a straightforward example of deductive analogy, the analog computer is a good illustration. Electric power systems for utilities have been simulated by analog systems so that the effects of changes in design or breakdowns may be investigated by changing components in the analog system. The analog and its interpretation are based on purely deductive laws of physics.

Inductive analogies appear in several forms. In one case a prediction is made on the hypothesis that the present situation is analogous to a previous situation

Problems of definitions and semantics in the present state of the social sciences make vagueness and ambiguity obstacles to providing proof. Yet there are a number of methods for testing and strengthening arguments. Logical analysis, when applicable to deductive arguments is helpful. The resort to mathematical symbols and their manipulation yields rigorous reasoning subject to the weakness of abstract representation of the real world. Inductive techniques, analogical reasoning offer means for reasoning to highly likely conclusions.

A major polling organization indicates that almost 80% of Americans believe that the collecting and giving out of personal information without their knowledge is a serious violation of their privacy.

Hence the right to privacy is an important question in business research. This issue involves the subject’s freedom to choose whether or not to comply with researchers requests. Traditionally researchers have assumed that individuals make an informed choice.

However, critics have argued that the poor, the old, the poorly educated, and other underprivileged individuals may not be aware of their right to choose.

Further, they have argued that the interview may begin with some vague explanation of its purpose, initially ask questions that re relatively innocuous, and then move to questions of a highly personal nature. It has been suggested that subjects be informed of their right to be left alone or break off the interview at any given time.

The nature of the research process requires open lines of communication between the client and vendor if the inquiry is to be effective. When a client enters into a working relationship with a research vendor, certain types of proprietary data will be discussed. The vendor has an ethical responsibility to hold the nature of their relationship confidential. In addition the client has the right to expect complete anonymity, as long as it does not in some way harm or deceive respondents or others involved in the research endeavor.

Another aspect of the privacy issue is illustrated by the question, "Is the telephone call that interrupts someone’s favorite television program an invasion of privacy?".

The answer to this issue, and to most privacy questions, lies in the dilemma of determining where the rights of the individual and the needs of society for better scientific information on citizen preference take over.

Generally, certain standards of common courtesy have been set by interviewing firms, for example, not to interview late in he evening and at other inconvenient times. However there are several critics who may never be appeased. The computerized interviewer has stimulated increased debate over this aspect of the privacy issue. As a practical matter, respondents may feel more relaxed about privacy issues if they know who is conducting a survey. Thus it is generally recommended that field researchers indicate that they are legitimate researchers by passing out business cards, wearing name tags, or in other ways of identifying the name of their company.

It has been argued that the subjects have the right to be informed of all aspects of the research. This includes information about its purpose and sponsorship. The argument for the researchers obligation to protect this right is based on the academic tradition of informing and enlightening the public.

Further, debriefing experimental subjects may relieve stress resulting from deception or other questionable procedures.

A pragmatic argument for providing respondents with information about the nature of the study concerns the long run ability of researchers to gain cooperation from the respondents. If the public understands why survey or experimental information has been collected and that the researchers may be trusted with private information, it may be easier to conduct research.

A number of clients might be very desirous of a list of favorable industrial prospects generated from a research survey. It is the researchers responsibility to ensure that the privacy and anonymity of the respondents are preserved. If the respondents name and address are known, this information should not under any circumstances be forwarded to the sponsoring organization. In formation that research supplier obtains about a client’s general business affairs should not be disseminated to other clients or third parties. The client or users of business research have a number of rights and obligations. Their primary right is to expect objective and accurate data from the research supplier. They should also expect that their instructions of confidentiality have been carried out.

The Privacy Act of 1974 has lent itself to various interpretations. One question that remains unanswered deals with the act’s stipulation that information about an individual cannot be extracted without securing his or her informed, expressed consent. Certain privacy advocates have suggested that this provision indicates that written consent of the respondent is required. If this becomes the case, it is very likely that potential respondents may be reluctant to cooperate if they are required to sign some type of agreement.

The right to be let alone, poses a dilemma for business research. While the continued application of research can have a positive effect on the efficiency of business in meeting society’s needs, it can also have a negative effect. For example, technology has made it possible to compile large household lists that include names, addresses, and a great deal of individual specific data. This database technology has the potential to threaten privacy in a free society. Society has the right to expect that personal information gained through research endeavors will not be disclosed.

Business researchers commonly use three types of surveys, which vary in terms of how data are gathered:

1. Questionnaire (including mail surveys)

2. Personal interview

3. Telephone interview.

Each type of survey involves slightly different administrative techniques and has to deal with different problems.

Questionnaire surveys are usually lower in cost and best for dealing with sensitive topics. Personal interviews provide the highest response rates and are best for dealing with complex topics. Telephone surveys take the least time to complete.

Mail questionnaires are highly standardized, and the questions are quite structured. If questions or instructions are difficult to comprehend, respondents must use their own interpretations, which may be wrong. Interviewing on the other hand allows for feedback from the interviewer regarding the respondents comprehension of the questionnaire.

Surveys that are boring, unclear, or too complex get thrown in the wastebasket. A poorly designed survey may be returned by only 15% of those sampled; in other words it will have a 15% response rate.

The major limitations of mail questionnaires relate to response problems. Respondents who answer the questionnaire may not be typical of all people in the sample. Individuals with a special interest in the topic are more likely to respond to a mail survey than those who are indifferent.

A researcher has no assurance that the intended subject will fill out the questionnaire.

Mail survey respondents tend to be better educated than nonrespondents. Poorly educated respondents who cannot read or write may well skip open ended questions where respondents are required to write out their answers.

Rarely will a mail survey have the 80-90% response typical of personal interviews.

Personal interview surveys tend to be highest in sampling error because they usually have smaller samples and often use cluster sampling. These characteristics, in turn, are due to the high cost per case of personal interviews. Because of the presence of an interviewer, these surveys are lowest in measurement error and nonresponse bias. An important characteristic of personal interviews is the opportunity to probe. If a respondent’s answer is brief or unclear, the researcher may probe for a clearer or more comprehensive explanation.

Depending on the research purpose, personal interviews vary in the degree to which questions are structured and in the amount of probing required. The personal interview is especially useful for obtaining unstructured information. Complex questions that cannot easily be asked in a telephone or mail survey can be handled by skillful interviewers.

If the research objective requires an extremely lengthy questionnaire, personal interviews may be the only alternative. Generally telephone interviews last fewer than ten minutes, whereas a personal interview can be much longer, perhaps and hour and a half. Control over interviewers is important to assure that difficult, embarrassing and time consuming questions are handled properly.

Personal interviews are generally more expensive than mail and telephone interviews. The geographic proximity of respondents, the length and complexity of the questionnaire , and the number of people who are nonrespondents because they could not be contacted (not at homes) all influence the cost of the personal interview.

The presence of an interviewer heightens the likelihood of social desirability bias and often makes respondents feel uncomfortable and answering sensitive or status-related questions. The interviewer also introduces inadvertent unreliability and measurement bias by varying the phrasing or tone of questions, and by deciding what to record, what to probe, and what to omit.

Interviewers in turn are influenced by respondents; it is difficult to treat respondents identically when they may be powerful and wealthy, poor with little education, or tired, or nervous.

Telephone interview surveys are intermediate on measurement error and nonresponse bias. Because they involve simpler sampling designs, their sampling error is generally low as well. The time restrictions on telephone interviews mean that the researchers may have to ignore potentially important variables. This is likely to introduce the errors associated with unmeasured variables.

Research agencies and interviewing services typically operate from a central location. Lines can be purchased from a long distance carrier at a fixed charge so that telephone calls can be made throughout the country or within a specific geographical area. Such central location interviewing allows firms to hire professional interviewers and to supervise and control the quality of interviewing more effectively.

In telephone interviewing the speed of data collection is also an advantage. Rather than wait several weeks for data collection by mail or personal interviews, hundreds of telephone interviews can be collected literally overnight. When the interviewer enters the respondents’ answers directly into a computerized system, data processing is sped up even more.

Telephone interviews are more impersonal than face to face interviews, embarrassing or confidential questions may be answered more willingly in a telephone interview than a personal interview. However a mail survey is the best method for gathering extremely sensitive information because it is anonymous.

Although telephone calls may be less threatening because an interviewer is not present, a limitation of this type of interviewing is the absence of face-to-face contact. The respondent cannot see that the interviewer is still writing down the previous comment and may continue to elaborate on an answer. If the respondent pauses to think about an answer the interviewer may not realize the person is thinking and go on to ask an additional question. Hence there is a greater tendency for ‘no answers’ and ‘incomplete answers’ to be recorded in telephone interviewing than in personal interviewing.

Each of the three survey types is prone to some research errors. Questionnaire studies are highest on measurement error because of the lack of opportunity to probe or explain the meaning of questions to respondents. They also have the highest levels of nonresponse bias because response rates are lower and because there is less opportunity to assess, and therefore correct for, bias of this kind.

Qualitative research deals with matters that cannot be meaningfully quantified. It is often viewed as soft research and sometimes thought of as having no value in business research.

Qualitative studies are usually in-depth investigations of an unstructured nature, using a very limited sample (eight to twelve individuals are usually sought in focus groups). The analyses make no attempt to be to be representative of the entire population; they are frequently meant to be impressionistic rather than definitively analytical. The goal is to understand the nature of the phenomenon in a holistic sense, rather than to numerically measure and analyze a predetermined set of variables. This requires high degree of skill on the interviewers part and usually involves long interview sessions.

The nature of qualitative research makes it ideal in exploratory stages of the research process, when the researcher possesses a low degree of understanding of the problem situation. Qualitative research has been used in the form of in-depth interviews, both group and individual, to help in the development of new products and advertisements, to understand the nuances of customer service in health care settings, to identify problem areas within profit and nonprofit organizations and to investigate behaviors that are considered to be sensitive and socially undesirable, such as compulsive shopping and shoplifting.

The two main qualitative research techniques are individual in-depth interviews and group in-depth interviews, also know as focus groups. By all accounts, in-depth interviews, particularly in the form of focus groups, are the most popular data collection methods in this type of research endeavor.

As the name suggests, an individual in-depth interview is used for intensive probing into one persons feelings about some subject. An interview may last an hour or even longer. Usually the interviewer does not have a set of structured or pre-specified questions. Rather the interviewer is given considerable flexibility to probe and stimulate responses from the subject. Several questioning strategies have been proposed to help the interviewer obtain the best possible information. These are laddering; which is designed to find the underlying attributes that distinguish things, people, or organizations. A sample questions could be " what differentiates Ford Motor Co. From General Motors and Chrysler?".

The interviewer would then identify salient attributes as the interviewee ‘ladders’ from GM to Chrysler. These attributes would subsequently be probed with further questioning to identify the significance and meaning of the attributes to the image of Ford Motor Co.

Hidden issue questioning; is designed to uncover and individual’s feelings about sensitive issues, such as employee theft, job security, or shoplifting. The questioning process would then attempt to uncover common themes across various interviewees.

Symbolic questioning; can be fruitfully used to identify underlying meanings. A respondent could be asked to name the opposite of yogurt. If the response is candy, the interviewer probes for the attributes of candy. Common responses might be sweet, unhealthy, fattening, and bad for you. Conversely then, yogurt would have the symbolic meanings of not sweet, healthy, nonfattening and good for you. Thus contextual meaning is given to the food class ‘yogurt’.

While individual in-depth interviewing can be very enlightening there are significant drawbacks. The cost and time associated with one-to-one interviews can be prohibitive from a resource standpoint. In addition the demands on the interviewer are tremendous. It does not take long for fatigue and information overload to cloud the interpretation ability of one interviewer. For this reason focus groups are perhaps the most popular of all qualitative research methods.

Focus groups.

A focus group usually consists of eight to twelve interviewees, drawn from a sampling of the population under study. Groups under eight tend to be too small, with little diversity and insufficient group interaction. Groups larger than twelve can be difficult to control because of sheer size and oftentimes group interaction is not achieved.

Group composition is critical because of the small number of respondents in the study, so the selection and recruitment of interviewees is all-important for the successful completion of the study.

The moderator is essential to the successful execution of a focus group interview. A good moderator is usually highly experienced in group dynamics and personal interaction. A moderator should neither lose control of the respondents in the focus group and should provide the structure for the group participants.

More structure means less spontaneity from the participants, but their comments are more directed to the researchers goals. Less structure means more opportunity for respondents to speak, but what they say is not necessarily of particular interest to the study’s objectives. The ideal structure depends on the objectives agreed upon between the moderator and the client before the focus group is conducted.

Therefore without a sensitive and effective moderator a single, self appointed, participant may dominate the session.

All in all, focus groups have a great potential to be used and abused in contemporary business research. This data collection method is extremely flexible, can produce insights other collection methodologies cannot, and tends to be more readily accepted by managers than many quantitative approaches, because the managers are frequently more involved in the focus group research.

On the other hand, serious abuses have occurred with focus groups. The involvement of management in the research process and the lack of generalizability can spell disaster for the unwary.

Because focus groups may have been successful in one setting with one set of objectives achieved it does not mean they will be successful with all settings and all objectives.

Exploratory research cannot take the place of quantitative, conclusive research. Nevertheless a number of firms use what should be exploratory studies as the final, conclusive research project. This has lead to incorrect decisions. Most of the limitations of exploratory research techniques is qualitative and the interpretation of the findings is typically judgmental.

Conclusions based on qualitative research may also be subjective to considerable interpreter bias. Findings from focus groups may also be ambiguous, how is a facial expression or nod of the head to be interpreted?

Have subjects fully grasped the idea or concept behind a nonexistent product? Have respondents overstated their interest because they tend to like all new products?

Because of such problems in interpretation, exploratory findings should be considered preliminary. Another problem with exploratory studies involves the projectability of the findings. Most studies use techniques which utilizes small sample sizes, which may not be representative because they have not been selected on a probability basis.

This is a written statement of the research design.

It includes a statement explaining the purpose of the study (research objectives) or a definition of the problem. It systematically outlines the particular research methodology and details the procedures that will be utilized at each stage of the research process. Normally a schedule of costs and deadlines will be included in the research proposal.

From the managers perspective a well written proposal can provide the following benefits:

Guarantees that the researcher understands the managerial problem. Document must clearly state the problem to be addressed and the results expected.

Acts as a control mechanism. When the research proposal is accepted, it establishes what the researcher will provide to management. In the case of outside research contractors it serves as a contractual obligation

Allows the manager to assess the proposed research methods. A preliminary assessment of the research methods can help to ensure that the study will result in the information desired by the manager.

Helps management judge the relative value and quality of the proposed investigation. Management may be forced to prioritize research projects because of limited budgets, several suppliers may be bidding the project and the proposal is particularly useful to compare the relative merits of each suppliers bid.

The following can be considered the major structural components of a mid range business proposal:

1. Letter of transmittal states the nature of the proposal and the main reasons the study should be conducted.

2. Title, this is usually short, descriptive and imaginative.

3. Background, this should briefly present the circumstances surrounding the immediate problem and the events leading to the development of the proposal should be reviewed. Depending upon the type of proposal the literature relating to the situation is presented here.

4. Statement of problem and research objectives must be clearly stated no matter what type of proposal is being prepared.

5. Research strategy and methodology presents the research methods and procedures to be used to accomplish the research objectives. Specifically the steps to be taken to achieve the desired results.

6. Nature of the final report specifies the anticipated results of the study and the means by which they will be communicated to the manager desiring the research.

7. Budget and schedule for completing the research.

8. Background of consultants if the proposal originated outside the organization requesting the information, the qualifications of the researchers should accompany.

Business research can take many forms but systematic inquiry is a common thread.

Systematic inquiry requires careful planning in an orderly investigation. Business research, like other forms of scientific inquiry is a sequence of highly interrelated activities.

The stages of the research process overlap continuously, and it is somewhat of an oversimplification to state that every research project follows a neat and ordered sequence of activities. Nevertheless, business research often follows a generalized pattern and the stages of this pattern are:

Defining the problem. The research task may be to clarify a problem or define an opportunity, therefore the initial stage is problem discovery, rather than definition. Often only symptoms are apparent to begin with. Profits may be declining but management may not know the exact nature of the problem. Thus the problem stated is in general terms. What is to be investigated is not specifically identified.

Planning the research design. After the researcher has formulated the research problem the research design must be developed. This is a master plan identifying the methods and procedures for collecting and analyzing the needed information. It is a framework of the research plan of action. The objectives of the study determined during the early stages of the research are included in the design to ensure that the information collected is appropriate for solving the problem. The research investigator must also determine the sources of information, the design techniques, the sampling methodology, and the schedule and cost of the research.

Sampling. Although the sampling plan is included in the research design, the actual sampling is a separate stage of the research process. We normally associate sampling with surveys. In business research, however, sampling is used in all kinds of research including pilot projects and evaluation studies. Unlike basic and general applied research, business research is always relevant to a specific population, time and place. If we cannot measure all the elements in that time and place, we sample. Researchers take the first step in sampling when they specify research boundaries as part of the research problem. These boundaries delineate the target population. The target population is the total set of cases lying within the research problem boundaries. These are the cases about which the researcher hopes to draw conclusions.

Data collection. Once the research design and the sampling plan has been formalized the process of gathering information may begin. Obviously, because there are many research techniques, there are many methods of data collection. When the survey method is used the some form of direct participation by the respondent is necessary during the process. Data collection can also be by the many types of surveys, differing interviews techniques or by questionnaires.

Data processing and data analysis. Once the field work has been completed the data must be converted into a format that will answer the decision makers questions. Data processing generally begins with the editing and coding of the data. Editing involves checking the data for omissions, legibility, and consistency in classification. Analysis is the application of logic to understand and interpret the data that has been collected about a subject.

Analysis may involve determining the consistent patterns and summarizing the appropriate details revealed in the investigation. The appropriate analytical technique for data analysis will be determined by management’s information requirements, the characteristics of the research design, and the nature of the data collected.

Conclusions and report preparation. The research report should communicate the research findings effectively. All too often the report is a complicated statement of the study’s technical aspects and sophisticated research methods. Often, management is not interested in detailed reporting of the research design and statistical findings but wishes only a summary of the findings. The managers information needs will determine how much detail is provided in the written report. The report is also a historical document that will be a source of record for later usage, such as repeating the survey of provide a basis for building upon the survey findings.

The purpose of experimental research is to allow the researcher to ‘control’ the research situation so that casual relationships among variables may be evaluated. The experimenter therefore manipulates a single variable and holds constant all other relevant, extraneous variables (Also events may be controlled in an experiment in a way that is not possible in a survey).

The experimenter either creates an artificial situation or deliberately manipulates a situation.

An example of experimentation concerns unit pricing. The purpose of unit pricing is to help shoppers avoid confusion in attempting to compare prices of comparable products, especially if the products are in different size packages. In the simplest form of experiment the researcher’s purpose may be to compare the effectiveness of the traditional shelf tag display with the list format, which lists prices in order of increasing prices for a given quantity.

The hypothesis would be that a single list of all brands sizes and their unit prices is an effective arrangement of unit price information. A shift toward the purchase of less expensive items can be measured by average price paid per unit, the dependent variable.

Say a supermarket chain has four stores in a city which has never used unit price information - the researcher may resold sales for a product over a period of time - then two of the stores may be assigned unit prices on shelf tags and the other two stores display unit price information in the list format. So the researcher can determine if unit price information as a list determines if it is effective in moving prices toward the less expensive units.

The researcher can hold constant variables and manipulates the experiment.

Selection and assignment of test units is possible.

Control over extraneous variables can be made.

Randomization assures the researcher that repetitions of an experiment will show true effects if they exist.

It is not always in business possible to control everything that would be required for the ‘perfect’ experiment.

The researchers may be constrained by management’s greater concern with efficiency than with research.

Random sampling errors and sample selection errors may occur in experimentation, if repetitions of the basic experiment sometimes favors one experimental condition than the other.

Surveys require asking people questions in order to gather information, which can be done verbally or in writing. Interviews or questionnaires are used to gather this information through the mail, over the telephone, or face to face. The type of information gathered varies and is dependent upon the objectives of the survey. Most survey research is descriptive in nature because it wants to find out what is happening or to learn the reasons for a particular business activity.

They can provide a quick method of gathering data, especially with the telephone.

Surveys can be flexible with questionnaires being designed to gather certain types of information and face to face interviewing can be unstructured so the interviewer can steer the interviewee toward the objectives of the survey.

Survey research techniques have been refined to become quite scientific and accurate.

Research projects must be estimated for the accuracy of the results by the manager and the two main sources of error are random sampling error and systematic error.

Statistical errors will occur because of chance variation, especially with a small sample being studied. To avoid this error the sample will have to be increased, this will add to time and cost for the survey and may put it outside the allocated budget if care is not taken in controlling this.

Systematic error will result from some imperfect aspect of the research design or from a mistake in the execution of the research.

The respondents being asked the questions may not tell the truth or co operate fully with the objectives of the survey, hence there will be error.

There may be no, or very low, response to the survey or indeed refusal to participate.

 

Much fieldwork is done by research suppliers who specialize in data collection, it is rare for the person who designed the survey to carry out the fieldwork.

When the second party is employed, the job of the study designer is to hire a research supplier and to establish supervisory controls over the field service.

In some cases even a third party may be involved because the firm subcontracts the field work to a field service.

The field workers are, usually, well trained and consistent.

Respondents may respond better to the survey if they have a person trained in questioning techniques who will put them at ease and thereby they are more likely to be cooperative.

Questions can be clarified by the interviewer at the time to save misunderstandings.

Probing of respondents is made easier.

Interviewers from the field service firm may not be either trained or supervised correctly.

Interviewers may cheat by just filling in the forms rather than questioning respondents.

Interviewers may fake answers when they find the questions embarrassing.

Interviewers may not have been selected properly in the first place, they may not be up to the rigors of the job, or not healthy enough for it.

Supervisors may not be up to the job wither, not verifying the interviewers properly, or not encouraging or not explaining properly what is required of the interviewer.

Validity addresses the problem of whether a measure measures what it is supposed to measure. There are two types of validity, these are internal validity and external validity.

This has to do with the cause and effect relationship in the experiment. It refers to whether the experimental treatment was the sole cause of observed changes in the dependent variable. If the observed results are influenced by the confounding effect from extraneous factors the researcher has problems making valid conclusions about the relationship between the experimental treatment and the dependent variable.

There are six major types of extraneous variables which may jeopardize internal validity:

history, where the effect refers to a specific event in the external environment that is beyond the control of the experimenter.

Maturation, where people change over time in some way that will effect the experimental results.

Testing, the effect of the pretest in a before and after study may sensitize respondents when they take the test a second time.

Instrumentation, to avoid the effects of testing an alternative form of test can be given during the post or after measurement. This may reduce the effect of testing, since there is a change in the measuring instrument.

Selection, the selection effect is a sample bias resulting in differential selection of respondents for the comparison groups

mortality, if the experiment is conducted over time there may be sampling error due to morality or sample attrition, this can occur when some respondents withdraw.

This problem concerns the researchers ability to generalize results from the experiment to the external environment, such as other corporate offices or the marketplace.

External validity is the quality of being able to generalize beyond the data of the experiment. In essence, it is a sampling question. To what extent can the results of a simulated negotiating experiment be transferred to the real world of union-management relations?

Problems of external validity generally relate to the possibility that a specific but limited set of experimental conditions may not deal with the interactions of untested variables in the real world.

The experimental situation may be artificial and it may not represent the true setting and conditions in which the investigated behavior takes place. If the study lacks external validity, it will be difficult to repeat the experiment if different subjects, settings or time intervals are used.

Researchers are often interested in finding differences in data and the detection of group differences is very useful. The manager can be presented with a large amount of data from this type of analysis so therefore the proper analytical method is related to the level of measurement of the data.

The Chi-square test allows the researcher to test for significance in the analysis of frequency distributions. This means that categorical data on variables such as sex, education, or dichotomous answers may be statistically analyzed.

The following methodology best describes how the Chi-square test is incorporated into the analysis and is a plan the manager, or researcher, can use when deciding which analytical method to use:

1. State the hypothesis, maybe where no this hypothesis is a statement that no relationship exists in the data.

2. Choose the statistical test, there are many tests from which to choose and the Chi-square test being just one of them, and there are many criteria that can be used in choosing a test, one criteria being the power efficiency of the test.

3. Select the desired level of significance, this choice is made before the data is collected because it affects the decision concerning the amount of data that is needed, the sample size.

The exact level is really determined by how much risk the researcher is willing to take in coming to an incorrect conclusion.

4. Compute the calculated difference value, where the value for the test, such as Chi-square is determined.

5. Obtain the critical test value, this is the criterion that defines the region of rejection as separate from the region of acceptance of the null hypothesis. These values are published in tables for each test so that all a researcher has to do is to look up the value for the appropriate test and significance level.

6. Make the decision, for most tests if the calculated value is larger than the critical value the null hypothesis is rejected and the conclusion is reached that the alternative hypothesis is supported.

This is where the Chi-square test fits into the plan, now to focus on the use of the actual test itself;

If the researcher wants to test the null hypothesis that the number of consumers aware of a certain product brand equals the number unaware of the brand. The logic inherent in the Chi-square test allows the researcher to compare the observed frequencies with the expected frequencies based on his theoretical ideas about the population distribution in his presupposed proportions. The technique will test whether the data come from certain probability distribution, it will test the ‘fit’ of the observed distribution with the expected distribution.

Using the plan detailed previously the researcher can:

Formulate the null hypothesis and determine the expected frequency of each answer.

Determine the appropriate significance level.

Calculate the Chi-square test value using the observed and expected frequencies from the sample.

Make the statistical decision by comparing the Chi-square value with the critical Chi-square value, which comes from the tables previously mentioned.

Like many other probability distributions the Chi-square distribution is not a single probability curve but a family of curves. These curves vary according to the degrees of freedom. Degrees of freedom refer to the number of observations that can be varied without changing the restraints or assumptions associated with a numerical system..

One of the simplest techniques for describing sets of relationships is cross-tabulation. A cross-tabulation is a joint frequency distribution of observations on two or more sets of variables . This generally means that tabulation of subgroups will be conducted for purposes of comparison. The Chi-square distribution provides a means for testing the statistical significance of cross-tabulation. This allows the researcher to test for differences in two groups distribution across categories.

The process of analysis begins after the data have been collected. During the analysis stage several interrelated procedures are performed to summarize and arrange the data.

The goal of most research is to provide information. There is a difference between raw data and information. Information refers to a body of facts that are in a format suitable for decision making, whereas data are simply recorded measures of certain phenomena. The raw data collected in the field must be transformed into information that will answer the managers questions. The conversion of raw data into information requires that the data be edited and coded so the data may be transferred to a computer or other data storage media.

The research steps following data collection related to processing and analysis are:

Editing

Coding

Data conversion (keyboarding)

Data analysis, and from this step follows;

Descriptive analysis

Univariate analysis

Bivariate analysis

Multivariate analysis

Which all lead to:

Interpretation.

After the data is collected it must be edited by a qualified person, personal interviews require more immediate editing than survey data because of comprehension and memory problems that may arise from the personal interview. Consistency and completeness is checked at this stage, missing data is checked for, interviewers asked for verification of details they have logged. However no matter how detailed this is carried out items can still ‘fall through the cracks’.

Sometimes the use of open ended questions can make it necessary to create categories not originally anticipated. Open ended questions , with no planned coding rules, present two problems for the researcher. First determining the number of mutually exclusive and exhaustive classes is not an easy task. Second ‘data dredging’ should only be used in exploratory or descriptive research with no strict hypotheses. Even the use of open-ended questions with preplanned coding can present problems for the researcher. If there are several coders, this reduces the consistency of the coded data. If a significant number of anomalous responses are found in the questions, another category might have to be added.

Without good data analysis planning the results of the study may be of little value to the manager, regardless how good the measurement and data collection processes are, errors of incompleteness and inconsistency in the data could ruin the analysis and ultimately corrupt the results of the study.

Once the data is stored is a data storage medium and the analysis run the ‘garbage in, garbage out’ rule applies. It also means that the proper analysis must be used for the level of measurement. If the data is improperly analyzed it can never be useful decision information.

1. Construct a clearly defined preanalytical plan which will ensure completeness, clarity, and consistency in the collected data before it is analyzed. Also it will increase the likelihood that the data will be converted to useful decision information.

2. Draw up a data analysis plan before the actual analysis is performed. This will ensure that the analysis will be in line with the research problem, and the information provided the analysis will be useful in solving the problem.

3. Require tabulations and/or scatter diagrams to be run on all variables whenever possible. This will ensure last minute accuracy of the data before the analysis, additionally valuable information, useful for the interpretation of the results, can be obtained from this basic analysis.

4. Require the researcher to explain the basic tabulations or scatter diagrams of the variables and how they might affect the analysis. This makes the researcher think about the analysis again and helps ensure that the analysis will provide the results that are needed.

5. Require that the researcher justify all results and conclusions from the analysis. This forces the researcher to communicate at the same level as the manger, it also ensures that the analysis yields the information that management needs to solve its problem.

Technology is having a tremendous impact on developments in the field of business research, in particular the proliferation of microcomputers is drastically changing the way mangers researchers collect, analyze, access, use, and report information. The widespread availability of distributed microcomputers and their links to more powerful main frames allow an unprecedented number of users access to a wealth of information. The following outlines the advantages and disadvantages of such computer data banks.

Most studies with a large sample size require a computer for data processing, it has been estimated that a research system that uses a microcomputer can reduce a three to four week research study to a five day project. This type of in line tabulating automatically stores and tabulates responses as they are collected, which substantially reduces errors of the clerical type that occur during the editing and coding processes.

The researcher must decide on the method of collection of the data and in some instances, such as passive data collection, the computer can outpace the human element. This method could be used in gauging the number of vehicles passing a point, which the computer will do more accurately with out getting bored.

Among the important primary data collection vendors are those that supply the services of a panel. This a group who are measured repeatedly over a period of time. Computerized forms of panel data collection is being used, which can be electronic devices to measure television and radio audiences, and universal product code scanning technologies both in the home and the business.

Predictive modeling is important for business decision making, the use of scenarios is widely used by organizations to see what would happen under various sets of assumptions. Useful business simulation models can be either deterministic, such as repeated application of break even models under different scenarios, stocastic, such as risk analysis for new product development assuming probabilistic demand for sales.

Simulation models are experimental devices that after validation, are used to calculate what would happen to the outcome variable under certain sets of assumptions. Computerization of simulation models has given firms and invaluable, cost effective tool for exploration of ‘what would happen if’ questions.

Most libraries now use computers to retrieve information. The ability of computers to store, scan, and print data has made the formerly tedious task of searching through masses of available literature faster and more efficient. Many of the periodical indexes and abstracts are available for searching on line, this enables a computer to search its data base for any stored information that relates to the subject area of interest. The process can take minutes rather than days and weeks.

Computer programs for solving integer and mixed-integer linear programs are now widely available. Such codes are generally reliable for problems involving up to 100 or more integer variables and have often been used to solve specially structured problems with a few thousand variables.

Computer analysis is dependent upon the data structure - the way the respondents information is positioned in a storage medium.

For purposes of data analysis three points regarding data structure should be considered: 1) the stored data must be consistent with the software used for analysis, 2) the data must contain codes for missing values consistent with the software used for analysis, 3) all coding in the data structure must conform to the specifications in the codebook.

In general, the reliability of software is established by its longevity and use in the marketplace. A careful check must be made of these factors as the proper interpretation of the results depends upon this. A lot of software, especially new software, contains bugs in the program and a check has to be made to see if the supplier has ironed all these out.

Computers tend to give the perception of being too impersonal and respondents may not like being denied the human face to face ‘personal’ aspect of data collection. Therefore respondents may ‘fudge’ answers or not cooperate as they think they are only affecting the computer which after all does not matter as much as contact with another person.

Data banks within computers can have a tendency to ‘loose’ data or to corrupt it making it more difficult, and sometimes impossible to carry out analysis of stored information. Retrieval of information may not be easy to the manager and may require the services of a programmer

which has to be factored into the budget.

Computer data banks require maintenance and regular checking which also adds to the cost and it has to be factored whether the cost of keeping a computer may make the firm less competitive than others in the data storage and analysis business.

Overcomplicated programs tend to be less ‘user friendly’ and difficult for managers to use. Hence this may be an off putting factor in the decision whether to use such systems.

Not only programs can be over complicated but the actual computer itself has to be carefully selected to ensure it is the right one for the jobs in hand, and with the speed of technological advances for future jobs. The wrong decision here may be expensive.

Research planning may not be carried out correctly. Planning errors are reflected in the setup of the design to collect information. A critical error results from ill-structuring or misspecification of the problem, and this will result in information that cannot solve management’s problems.

Planning errors are also associated with the selection of an inappropriate research design. The research design process is a highly interactive stepwise procedure, in which decisions in earlier stages affect the outcomes in later stages.

Careful planning and critical review of the research proposal can minimize planning errors.

Collection errors arise during the actual collection of the data, these types of errors can serve to magnify errors that are a result of inadequate planning. The researcher must ensure that the data collect is of the right type and quality to answer the research question. If the data is not collected appropriately then even the best planning cannot yield information that is truly valuable to management.

Research errors are due to the incorrect interpretation of the results of the study and generally have to do with putting erroneous meaning into the relationships and numbers identified at the data analysis stage of research, such errors can be either intentional or unintentional..

Intentional errors are ethical concerns, but unintentional errors are more difficult to detect and usually result from ignorance or honest mistake. Critical and unbiased reviews can help minimize unintentional errors.

One or more of the following strategies can be used to reduce errors:

1. Planning error reduction strategy. Development of well thought out proposals that clearly specify the methods, costs, and value of the research to be undertaken. Forces researchers to clearly state the research problem and the means by which they are going to solve it (design). This strategy can be ensured by having the proposals evaluated by impartial and research competent individuals and committees.

2. Collection error reduction strategy. This requires the careful and thoughtful execution of the specified research design outlined in the proposal. This strategy will keep the researchers orientated to the plan and focuses their efforts on the research problem identified.

3. Analytical error reduction strategy. This requires a justification of the analytical procedures used in summarizing and manipulating the data. A justification of the procedures to be utilized requires the researcher to think through the implications of the analysis, and to be sure to ask why a particular analytical technique is being used to answer the research problem.

4. Reporting error reduction strategy. This requires unbiased and knowledgeable reviewers to examine the results of the study. Professional researchers are accustomed to having their work subjected to public review and criticism, this is a fundamental characteristic of all research .

These planning steps should help to reduce the errors throughout the process which will ultimately become evident in the data analysis and interpretations stage.

From a management perspective the qualitative meaning of the data and the managerial implications are important aspects of the interpretation. In all too many instances data are merely reported and not interpreted, research firms may provide reams of computer output that do not state what the data mean. At the other extreme there are researchers who tend to analyze every possible relationship between each and every variable in the study; the have usually not defined the problem in the earlier stages of the research. Researchers who have a clear sense of the purpose of the research do not request statistical analysis of data that may have little or nothing to do with the primary purpose of the research.

The preanalytical process incorporates personal and computerized means to ensure that the data is purified before analysis which will reduce inaccuracies and fallacious elements in the data.

The role of business research in decision making is providing the decision maker with relevant and useful information with which to make decisions.

The growth of the information industry coupled with the developments of the computer give organizations and researchers more reasons to search for the best information possible before making a decision.

Managers are faced with a number of problems that they must solve through specific decisions, they want quick, low cost, timely, and accurate information to solve these problems.

The applied decision making orientation of managers in business demands that a wider range of studies and investigations be included under the designation of business research. The scope of business research is limited by one’s definition of ‘business’, certainly research in the production, finance, marketing, or management areas of a for profit corporation is within the scope of business research.

In a number of situations researchers will know exactly what their problems are and will design studies to test specific hypotheses. In another situation management may be totally unaware of a problem, in such as case exploratory research may be necessary to gain insights into the nature of the problem.

Another good reason for using business research is to find out why something went wrong. Detailed information about specific mistakes or failures is frequently sought.

If a general problem area is identified, breaking industry sales volume and a firm’s sales volume into geographic areas may provide an explanation of various problems, and exploring problems in greater depth may indicate which managerial decisions were erroneous.

Change is constant in the business environment and information about changing environments helps managers to make decisions on how they should change the firms strategy, position, market share and so on. Predicting and forecasting future trends and conditions aids managers at the same time.

Sometimes the purpose can be just to describe what is going on in the organization or the environment.

Business research is the systematic and objective process of gathering, recording, and analyzing data for decision making. The research must be systematic, not haphazard. It must be objective to avoid the distorting effects of personal bias. Managers can use business research to define problems, to identify opportunities, or to clarify alternative courses of action.

Research is also used to explain what went wrong with managerial efforts in the past, to describe current events in the marketplace and to forecast future conditions.

Globalization is causing us to question traditional ways of viewing cultural identity, as our organizational contexts become more complex, and simultaneously more multicultural what is the most efficacious way to conceptualize culture? Embedded in this question are assumptions about the nature and mandate of scientific inquiry.

Research in the cross-cultural comparative stream has provided us with great momentum and allowed us to make initial steps toward understanding cultural differences. It has motivated the development of various typologies of culture for understanding these differences.

The focus on a finite set of dimensions of culture has allowed the inclusion of cultural variables into cross national research in other management disciplines such as strategy and marketing.

As a body or research, the organization culture literature exhibits great variation in focus and methods. Conceptualizations of culture range from viewing it as an independent variable that managers can manipulate for desired ends, that is something and organization ‘has’ to viewing it as something an organization ‘is’. The latter viewpoint carries the interpretive approach to its logical conclusion.

Researchers face serious theoretical and methodological dilemmas when defining culture by national boundaries and representing the unwieldy construct even more so. Representations of culture as a social construct tend to overlook existing variation and incongruencies. This is especially likely if we trying to reach a management oriented academic and practitioner audience that favors a clear, systematic, economical representation.

Issues of representation imply issues of method and additionally of research paradigm. The discussion of cross national intercultural interaction highlights insights gained from intensive ethnographic research. Significant obstacles inhibit wide adoption from this naturalistic research paradigm.

The globalization of the world economy has spurred a rapid increase in cross border alliances. Firms seeking to gain competitive advantage are doing so by collaborating with local counterparts in regions where they have no experience. Similar alliances by researchers can be advantageous to advancing the field of international business. Such alliances can reduce the amount of travel required, thus streamlining the expense of the research. Local expertise can facilitate the navigation of bureaucratic labyrinths and expedite the evaluation of the external factors that may affect the project. Collaboration with a national from the region under investigation can enhance the comprehension of cultural factors that may impede or invalidate research findings. The western researcher will also have a guide in understanding local protocol and assistance with translation, interpretation, and administration.

For the local researcher such alliances can provide the opportunity to engage in projects that may otherwise be impractical due to time or resource constraints. This approach could be a significant stepping stone toward the publication and dissemination of investigations that contribute to the theoretical development of globally applicable paradigms, and internationally transferable management practices.

There are however some problems which have to be overcome by the researcher when conducting studies in another culture, some of these are now considered:

Mail questionnaires in less industrialized countries tend to have a very poor response rate. Factors such as the reliability of the postal system, the expense of mailing a survey to another country, the motivation of the respondents to oblige an unknown source, lack of literacy all serve to constrain the effectiveness of this type of survey.

Interviews which involve face to face contact with open ended questions seem to elicit the richest information. Successful interviews are dependent upon gaining access to organizations and their principal mangers. In many countries access can be a formidable obstacle to foreign researchers.

Investigators need to establish contacts and formulate specific data collection strategies well before their arrival

With participant observation the practicality once again depends on the type of contacts the researcher can establish and the country where the research is being conducted. Assuming permission can be obtained direct observation at various organizations can provide benefits. The opportunity to observe the normal routines of an organization will sensitize the researcher to the culture of the country and the organization and facilitate the development of boundaries and parameters for experimental research.

Experimental research is not a popular mode of investigation in less industrialized countries. Since this form or research almost always involves the disruption of daily routines, it is difficult to persuade managers and government officials to participate. The scarcity of resources infers that there is little slack in the system and scheduling time for experimental tests may be difficult.

The evolution a global marketplace is redefining the arena for international business. The economic, political, social, and technological changes taking place in the world today are transforming the rules of competition. These rapid and dramatic changes have the potential to render much of the international business literature obsolete. Our theories, methods and practices exhibit a significant ‘western’ influence that may not be applicable to countries such as China, Mexico, Brazil, and Hong Kong which are growing in strategic importance. To maintain the relevance and usefulness of international business research, it is incumbent upon researchers to take these points into account and take a systematic look at forgotten locations.

Click here to go back to the Home page....