RISET BISNIS Oleh Prof Dr Djumilah Zain SE

RISET BISNIS Oleh: Prof. Dr. Djumilah Zain, SE

RESEARCH METHODS FOR BUSINESS: A Skill Building Approach UMA SEKARAN

CONTENTS INTRODUCTION TO RESEARCH THE RESEARCH PROCESS: STEPS 1 TO 3 THE RESEARCH PROCESS: STEPS 4 AND 5 THE RESEARCH PROCESS: STEP 6 EXPERIMENTAL DESIGNS MEASUREMENT OF VARIABLES DATA COLLECTION METHODS SAMPLING A REFRESHER ON SOME STATISTICAL TERMS AND TESTS 10. DATA ANALYSIS AND INTERPRETATION 11. THE RESEARCH REPORT 1. 2. 3. 4. 5. 6. 7. 8. 9.

CHAPTER 1 INTRODUCTION TO RESEARCH

What is Research? A systematic and organized effort to investigate a specific problem that needs a solution. It is a series of steps designed and followed, with the goal of finding answers to the issues that are of concern to us in the work environment.

Business Research In business, research is usually primarily conducted to resolve problematic issues in, or interrelated among, the areas of accounting, finance, management, and marketing.

Types of Research Applied research Basic or fundamental research

Managers and Research Being knowledgeable about research and research methods helps professional managers to: 1. Identify and solve small problems in the work setting. 2. Know how to discriminate good from bad research. 3. Appreciate and constantly remember the multiple influences and multiple effects of factors impinging on a situation. 4. Take calculated risks in decision making, knowing full well the probabilities attached to the different/possible outcomes. 5. Prevent possible vested interests from operating in a situation. 6. Relate to hired researchers and consultants more

Scientific Investigation Scientific research has the focused goal of problem solving and pursues a step by step logical, organized, and rigorous method to identify problems, gather data, analyze the data, and draw valid conclusions therefrom.

The Hallmarks of Scientific Research 1. Purposiveness 2. Rigor 3. Testability 4. Replicability 5. Precision and confidence 6. Objectivity 7. Generalizability 8. Parsimony

The Building locks of Science in Research Observation Identification of problem area Theoretical framework or Network of association Refinement of theory (pure research) or Implementation (applied research) Hypotheses Interpretation of data Constructs Concepts Operational definition Analysis of data Data Collection Research design

The Hypothetico Deductive Method 1. Observation 2. Preliminary information gathering 3. Theory formulation 4. Hypothesizing 5. Further scientific data collection 6. Data analysis 7. Deduction

Summary In this chapter we have examined what research is, considered the two types of research, tried to understand scientific investigation, what the hypothetico deductive method or research involves, why a manager should know about research, and the advantages and disadvantages of hiring internal and external teams of researchers or consultants. We examine the research process in the next two chapters.

CHAPTER 2 STEPS 1 TO 3: THE BROAD PROBLEM AREA PRELIMINARY DATA GATHERING PROBLEM DEFINITION

The Research Process for Basic and Applied Research ① OBSERVATION Broad area of research interest identified ③ PROBLEM DEFINITION Research problem delineated ② PRELIMINARY DATA GATHERING Interviewing literature survey ④ THEORETICAL FRAMEWORK Variables clearly identified and labeled ⑦ ⑤ GENERATION OF HYPOTHESES ⑥ SCIENTIFIC RESEARCH DESIGN DATA COLLECTION ANALYSIS, AND INTERPRETATION ⑧ DEDUCTION Hypotheses substantiated? Research question answered?

Broad Problem Area The broad problem area refers to the entire situation where one sees a possible need for research and problem solving. The specific issues that need to be researched within this situation may not be identified at this stage. Such issues might pertain to: 1. Problems currently existing in an organizational setting that need to be solved. 2. Areas in the organization that a manager believes need to be improved. 3. A conceptual or theoretical issue that needs to be tightened up for the basic researcher to understand certain phenomena. 4. Some research questions that a basic researcher wants to answer empirically.

Preliminary Data Collection Nature of data to be collected § Background information on the organization § Information on management philosophy and structural factors § Perceptions, attitudes, and behavioral responses Literature survey § Reasons for literature survey § Conducting the literature survey Identifying the relevant sources Bibliographical indexes § Extracting the relevant information § Writing up the literature review

Problem Definition A problem does not necessarily mean that something is seriously wrong with a current situation, which needs to be rectified immediately. A “problem” could simply indicate an interest in an issue where finding the right answers might help to improve an existing good situation. Thus, it is fruitful to define a problem as any situation where gap exists between the actual and the desired ideal state.

Summary In this chapter, we learned about the first three steps in the research process: identification of the broad problem area to be researched, preliminary data gathering through interviews and literature survey, and problem definition. The appendix to this chapter offers information on on line databases, bibliographical indexes, APA format for references, referencing previous studies and quoting original sources in the literature review section, and some of the most frequently cited business journals. In the next chapter we will examine the next two steps in the research process: theoretical framework and hypotheses.

CHAPTER 3 THE RESEARCH PROCESS: STEPS 4 AND 5: THEORETICAL FRAMEWORK HYPOTHESIS DEVELOPMENT

The Research Process for Basic and Applied Research ① OBSERVATION Broad area of research interest identified ③ PROBLEM DEFINITION Research problem delineated ② PRELIMINARY DATA GATHERING Interviewing literature survey ④ THEORETICAL FRAMEWORK Variables clearly identified and labeled ⑦ ⑤ GENERATION OF HYPOTHESES ⑥ SCIENTIFIC RESEARCH DESIGN DATA COLLECTION ANALYSIS, AND INTERPRETATION ⑧ DEDUCTION Hypotheses substantiated? Research question answered?

The Need for a Theoretical Framework A theoretical framework is a conceptual model of how one theorizes the relationships among the several factors that have been identified as important to the problem. The theoretical framework discusses the interrelationships among the variables that are deemed to be integral to the dynamics of the situation being investigated. Developing such a conceptual framework helps us to postulate and test certain relationships so as to improve our understanding of the dynamics of the situation.

Variables A variable is anything that can take on differing or varying values. The values can differ at various for the same object or person, or the values can differ at the same time for different objects or persons. Examples of variables are exam scores, absenteeism, and motivation. Types of variables: 1. The dependent variable (also known as the criterion variable). 2. The independent variable (also known as the predictor variable). 3. The moderating variable. 4. The intervening variable.

Dependent variable The dependent variable is the variable of primary interest to the research. The researcher’s goal is to explain or predict the variability in the dependent variable.

Independent variable An independent variable is one that influences the dependent variable in either a positive or a negative way. That is, when the independent variable is present, the dependent variable is also present, and with each unit of increase in the independent variable, there is an increase or decrease in the dependent variable also. New product success Independent variable Managerial values Independent variable Stock market price Dependent variable Power distance Dependent variable

Moderating variable The moderating variable is one that has a strong contingent effect on the independent variable relationship. That is the presence of a third variable (the moderating variable) modifies the originally expected relationship between the independent and the dependent variables. Number of books Reading abilities Independent variable Dependent variable

Number of books Reading abilities Dependent variable Independent variable Parents literacy Moderating variable Organizational effectiveness Workforce diversity Dependent variable Independent variable Managerial expertise Moderating variable

Intervening variable Willingness to learn An intervening variable is one that surfaces between the time the independent variables operate to influence the dependent variable and their impact on the dependent variable. There is thus a temporal quality or time dimension to the intervening variable. Training programs Growth needs

Willingness to learn Effects for those high in growth needs Effects for those low in growth needs Training programs

Time: t 1 t 2 Creative synergy Workforce diversity Independent variable Time: Intervening variable t 1 t 2 Creative synergy Workforce diversity Independent variable Intervening variable Managerial expertise Moderating variable t 3 Reading abilities Dependent variable

Theoretical Framework The theoretical framework is the foundation on which the entire research project is based. It is logically developed, described, and elaborated network of associations among variables that have been identified through such processes as interviews, observations, and literature survey.

The components of theoretical framework 1. 2. 3. 4. 5. The variables considered relevant to the study should be clearly identified and labeled in the discussions. The discussions should state how two or more variables are related to each other. This should be done for the important relationships that are the orized to exist among the variables. If the nature and direction of the relationships can be theorized on the basis of the findings from previous research, then there should be an indication in the discussions as to whether the relationships would be positive or negative. There should be a clear explanation of why we would expect these relationships to exist. The arguments could be drawn from the previous research findings. A schematic diagram of theoretical framework should be given so that the reader can visualize theorized relationships.

Communication among Cockpit members Communication between Ground control and cockpit Air safety violations Decentralization Training of cockpit crew Independent variables Dependent variables

Communication among Cockpit members Communication between Ground control and cockpit Air safety violations Decentralization Training of cockpit crew Independent variables Nervousness And diffidence Intervening variables Dependent variables

Communication among Cockpit members Communication between Ground control and cockpit Air safety violations Decentralization Training Independent variables Moderating variables Dependent variables

Hypotheses Development Definition of hypothesis An hypothesis is an educated guess about a problem’s solution. It can be defined as a logically conjectured relationship between two or more variables expressed in the form of testable statements. Statement of hypotheses: formats If then statements → To examine whether the conjectured relationships or differences exist or not, these hypotheses can be set either as propositions or in the form of if-then statements. Null and alternate hypotheses The null hypothesis is a proposition that state a definitive, exact relationship between two variables.

Hypotheses Testing with Qualitative Research: Negative Case Analysis Hypotheses can also be tested through qualitative data. For example, let us say that a researcher has developed theoretical framework after extensive interviews, that unethical practices by employees are a function of their being ignorant of what is right or wrong, or because of a need for more money, or because of the organization’s indifference to such practices. To test the hypothesis that these three factors are the primary ones that influence unethical practices, the researcher will look for data that would disconfirm the hypothesis. When even a single case disconfirms the hypothesis, he will revise theory.

Summary In this chapter we examined the four types of variables: dependent, independent, moderating, and intervening. We also discussed how theoretical frameworks are developed and testable hypotheses are generated therefrom as they relate to both qualitative and quantitative research. In the next chapter we will examine the basic research design issues.

CHAPTER 4 THE RESEARCH PROCESS: STEPS 6: ELEMENTS OF RESEARCH DESIGN

The Research Design The issues pertinent to research design relate to where the study will be conducted (i. e, the study setting), what type of a study it would be (type of investigation), the extent to which the researcher manipulates and controls the study (extent of researcher interference), the duration of the study (time horizon), and at what level the data will be analyzed (unit of analysis), as well as deciding what the sample would be (sampling design), how the data would be collected (data collection methods), how variables will be measured (measurement), and how they will be analyzed to test the hypotheses (data analysis).

The Research Process ① OBSERVATION Broad area of research interest identified ③ PROBLEM DEFINITION Research problem delineated ② PRELIMINARY DATA GATHERING Interviewing literature survey ④ THEORETICAL FRAMEWORK Variables clearly identified and labeled ⑦ ⑤ GENERATION OF HYPOTHESES ⑥ SCIENTIFIC RESEARCH DESIGN DATA COLLECTION ANALYSIS, AND INTERPRETATION ⑧ DEDUCTION Hypotheses substantiated? Research question answered?

The Research Design DETAILS OF STUDY Purpose of the data PROBLEM STATEMENT Exploration Description Hyphothesis Types of investigation Extent of researcher interference Establishing: Causal relationships Correlations Group differences, ranks, etc. Minimal: Studying events as they normally occur Manipulation and/or control and/or simulation MEASUREMENT Study setting Contrived Noncontrived Measurement and measures Operational definition Items (measure) Scaling Categorizing Coding DATA ANALYSIS 1. Feel for 2. data 2. Goodness of data Unit of analysis (population to be studied) Individuals Dyads Groups Organizations Machines Etc. Sampling design Probability/ nonprobability Sample size (n) Time horison One shot (cross sectional) Data collection method Observation Interviews Questionnaire Physical measurement Unobtrusive 3. Hyphotesis 4. testing

The Purpose of the Study Exploratory study Descriptive study Hypotheses testing

Type of Investigation Causal When the researcher wants to delineate the cause of a problem, then the study is called a causal study. Noncausal When the researcher is interested in delineating the important variables that are associated with the problem, it is called a correlational study. Whether a study is a causal or a correlational one thus depends on the type of research questions asked and how the problem is defined.

Extent of Researcher Interference with the Study The extent of researcher interference has a direct bearing on whether a causal or correlational study is undertaken. A correlational study is conducted in the natural environment of the organization with the researcher interfering minimally with the normal flow of events. In causal studies conducted to establish cause → effect relationships, the researcher tries to manipulate certain variables so as to study the effects of such manipulation on the dependent variable of interest. In other words, the researcher deliberately changes certain variables in the setting and interferes with the normal flow of events as they usually occur in the organization.

Study Setting Contrived and Noncontrived Research can be done in the natural environment where events normally occur – that is, in noncontrived settings – or in artificial, contrived settings. Correlational studies are invariably conducted in noncontrived settings, whereas rigorous causal studies are done in contrived lab settings.

Unit of Analysis Individuals Dyads Groups Organizations Cultures

Time Horizon Cross-Sectional Studies A study can be done in which data are gathered just once, perhaps over a period of days or weeks or months, in order to answer a research question. Such studies are called one shot or cross sectional studies. Longitudinal studies When data are gathered at two different points in time, it is not a cross sectional or a one shot study, but it is a study carried longitudinally across a period of time.

Summary In this chapter we examined the basic research design issues and the choice points available to the researcher. We also discussed the circumstances in which each design decision will be appropriate. In the next chapter we discuss how experimental designs are set up and the ways in which cause → effect relationships can be determined.

CHAPTER 5 EXPERIMENTAL DESIGN

Causal Versus Correlational Analysis A Correlational study is distinguished from and experimental design in that the former is concerned with identifying the important correlates that explain the environment variance in the pendent variable, and the study is conducted in the environment where event naturally occur without any artificial constraints being imposed in the setting.

The Laboratory (Lab) Experiment and The Field Experiment The Lab Experiment As stated earlier, when a cause effect relationship is to be clearly established between in independent a dependent variable of interest, then all other variables that might contaminate or confound the relationship have to be tightly controlled.

Manipulation of the Independent Variable In order to examine the causal effects of independent variable on dependent variable, certain manipulations need to be tried. Manipulation simply means that we create different levels of the independent variable to assess the impact on the independent variable.

Controlling The Contaminating or “Nuisance” Variables Matching groups Randomization

Ethical Issues in Research and Lab Experiments The following practices are considered unethical: § Putting pressure on individuals to participate in research (through coercion, applying social pressure, etc. ) § Asking demeaning questions that diminish their self respect. § Deceiving subjects by deliberately misleading them as to the true purpose of the research. § Exposing participants to physical or mental stress. § Not allowing them to withdraw from the research when they want to. § Using the research results to disadvantage the participants, or for purposes that the participants would not like. § Withholding benefits from control groups.

The Field Experiment A field experiment is an experiment done in the natural environment in which events normally occur, with treatments given to one or more groups.

Factors Affecting Internal Validity 1. History effects 2. Maturation effects 3. Testing effects 4. Instrumentation effects 5. Selection bias effects 6. Statistical regression 7. Mortality

Factors Affecting External Validity Whereas internal validity raises questions about whether it is the treatment alone or some extraneous factor that causes the effects, external validity raises issues about the generalizability of the findings to other settings.

When Are Experimental Designs Necessary? Some questions that need to be addressed are the following : 1. Are causal relationships necessary to be identified, or would tracing the correlates that account for the variance in the dependent variable be enough? If the latter would do, experimental designs are not really needed. 2. If causal relationships are important to be identified, is there a greater need for internal validity or external validity or both? If internal validity alone is important, a carefully designed lab experiment would be the answer; if generalizability is the more important criterion, then a field experiment would be called for; if both are equally important, then a lab study be first undertaken, followed by a field experiment. 3. Is cost an important factor in the study? If cost is a primary consideration, would a less sophisticated rather than a more sophisticated experimental design do?

Decision points for embarking on an experimental design Is tracing causal affects necessary? Yes and if No Internal validity is more important than external validity Generalizability is more important than internal validity Engage in a lab experiment Engage in a field experiment Do not undertaken an experimental design study Both internal validity and external validity are important First do a LAB experiment, then, a FIELD experiment Are there cost constraints? Yes Engage in a simpler experimental experiment No Engage in a more sophisticated design

Types of Experimental Designs and Internal Validity Pretest and posttest experimental group design Posttests only with experimental and control group Pretest and posttest experimental and control group designs Solomon four group design

Major Threats to Internal Validity in Different Experimental Designs When Members Are Randomly Selected and Assigned Types of Experimental Designs 1. Pretest and posttest with one experimental group design 2. Posttests only with experimental and one Major Threats to Internal Validity Testing, history, maturation Maturation control group 3. Pretest and posttest with one Mortality experimental and one control group designs 4. Solomon four group design Mortality

Simulation An alternative to lab and field experimentation currently being used in business research is simulation. Simulation uses a model building technique to determine the effects of changes, and computer based simulations are becoming popular in business research.

Summary This chapter covered experimental designs, with particular reference to the lab and field experiments. Issues of internal and external validity and the seven factors that could affect internal validity were discussed. Also, some types of experimental designs that can be used to test cause → effect relationships and the usefulness of these in the context of validity versus practicality were examined. The next chapter discusses how the variables– whether in a field survey or in an experimental design– can be measured.

CHAPTER 6 MEASUREMENT OF VARIABLES

How Variables Are Measured Objects that can be physically measured by some calibrated instruments pose no problem. Data representing several demographic characteristics of the office personnel are also easily obtained by asking employees simple, straight forward questions, foe example: – How long have you been working in this organization? – How long have you been working on this particular assignment? – What is your job title? – What is your marital status?

Operational Definition Operationalizing, or operationally defining a concept so that it becomes measurable, is achieved by looking at the behavioral dimensions, facets, or properties denoted by the concept, and categorizing these into observable and measurable elements.

Scales and Measurement Scales § Nominal scale § Ordinal scale § Interval scale § Ratio scale

Developing Scales Rating scale § Graphic rating scale § Itemized rating scale Attitude scale § Likert scale § Semantic differential

Validity Content validity Criterion related validity Construct validity

Types of Validity Description Content validity Does the measure adequately measure the concept? Face validity Do “experts” validate that the instrument measures what is name suggests it measures? Criterion related validity Does the measure differentiate in a manner that helps to predict a criterion variable? Concurrent validity Does the measure differentiate in a manner that helps to predict a criterion variable currently? Predictive validity Does the measure differentiate individuals in a manner as to help predict a future criterion? Construct validity Does the instrument tap the concept as theorized? Convergent validity Do two instruments measuring the concept correlate highly? Discriminant validity Does the measure have a low correlation with a variable that is supposed to be unrelated to this variable?

Reliability Stability of measures § Test retest reliability § Parallel form reliability Internal consistency of measures § Interitem consistency reliability § Split half reliability § Interrater reliability § Goodness of measures

Summary In this chapter, we saw how concepts are operationally defined and what kinds of scales can be used in developing instruments. We also discussed how the goodness of measures can be established in terms of validity and reliability. In the next chapter, we will see the different methods by which data can be collected.

CHAPTER 7 DATA COLLECTION METHODS

Data Collection Methods, Settings, and Sources of Data collection methods include face to face interviews, telephone interviews, computer assisted interviews; questionnaires that are either personally administered, sent through the mail, or electronically administered; observation of individuals and events with or without videotaping or audio recording; and a variety of other motivational techniques such as projective tests. As for the setting, data can be collected in any one of the aforementioned ways in the natural environment in which phenomena occur. Data may also be collected in lab experimental settings where variables are controlled and manipulated, or gathered in the homes of the respondents, on the street, in malls, or in a setting where a LAN (Local Area Network) system is available. Data sources can be primary and/or secondary.

Interviewing Unstructured interviews are also labeled because the interviewer does not enter the interview setting with a planned sequence of questions that he will be asking the respondent. Structured interviews are those conducted by the interviewer when he or she knows exactly what information is needed and has a predetermined list of questions that will be posed to the respondent.

Some tips to follow while interviewing § Establishing credibility and rapport, and motivating individuals to respond § The questioning technique w Funneling w Unbiased questions w Clarifying issues w Helping the respondent to think through issues w Taking notes Face to face and telephone interviews Computer assisted interviewing

Questionnaires Personally administered questionnaires Mail questionnaires

Guidelines for Questionnaire Design Principles of wording § Content and purpose of the question § The language and wording of the questionnaire § Type and form of questions • Open ended versus closed questions • Positively and negatively worded questions • Double barreled questions • Ambiguous questions • Recall dependent questions • Leading questions • Loaded questions • Social desirability • Length of questions § Sequencing of questions § Classification data or personal information

Principles of measurement Just as there are rules or guidelines that have to be followed to ensure that the wording of the questionnaire is appropriate to minimize bias, so also are there some principles of measurement that are to be followed to ensure that the data collected are appropriate to test our hypotheses. These principles of measurement encompass the scales and scaling techniques used in measuring concepts, as well as the assessment of reliability and validity of the measures used.

Electronic Questionnaire Design and Surveys On line questionnaire surveys for respondents are possible when microcomputers are hooked up to computer networks. Data disks can also be mailed to respondents, who can use their own personal computers for responding to the questions.

Other Methods of Data Collection Observational surveys § Nonparticipant observer § Participant observer § Structured versus unstructured observational studies Biases in observational studies Data observed from the researcher’s point of view are likely to be prone to observer biases. Moreover, where several observers are involved, interobserver reliability has to be established before the data can be accepted. Observer fatigue could also be a source of bias.

Some Special Data Sources Focus groups Static and dynamic panels Unobtrusive measures

Summary In this chapter we examined various data collection method and different primary sources of data. We discussed the advantages and disadvantages as well as the biases embedded in each data collection method. We also traced the current level of impact of personal computers in data collection. Because of the inherent biases in each of the data collection methods, obtaining data from multiple sources and through multiple methods was recommended. The choices, of course, will be governed by cost considerations and the extent of rigor desired for a given research goal. In the next chapter we will discuss sampling designs and how data can be collected from samples to make the results generalizable to the population.

CHAPTER 8 SAMPLING

Population, Element, Population Frame, Sample, and Subject Population refers to the entire group of people, events, or things of interest that the researcher wishes to investigate. Element. An element is a single member of the population. Population frame. The population frame is a listing of all the elements in the population from which the sample is to be drawn. Sample. A sample is a subset of the population. It comprises some members selected from the population. In other words, some, but not all, elements of the population would form the sample. Subject. A subject is a single member of the sample, just as an element is a single member of the population.

Sampling is the process of selecting a sufficient number of elements from the population so that by studying the sample, and understanding the properties or the characteristics of the sample subjects, we will be able to generalize the properties or characteristics to the population elements.

Normality of Distribution Many attributes or characteristics in the population are generally normally distributed. If we are to estimate the population characteristics reasonably precisely from the characteristics represented in a sample, the sample has to be chosen such normal distribution of the characteristics of interest follows the same type of normal distribution in the sample as it does in the population. Low μ High

Probability and Nonprobability Sampling Probability sampling – Unrestricted or simple random sampling – Restricted or complex probability sampling Systematic sampling Stratified random sampling Cluster sampling Area sampling Double sampling Non probability sampling – Convenience sampling – Purposive sampling Judgment sampling Quota sampling

Probability and Nonprobability sampling Design Sampling Design Description Advantage/Disadvantage Probability Sampling 1. Simple random sampling 2. Systematic sampling 3. Stratified random sampling (Str. R. S. ) Proportionate Str. R. S. Disproportionate Str. R. S. 4. Cluster sampling All elements in the population are considered and each element has an equal chance of being chosen as the subject. High generalizability of findings. Not as efficient as stratified sampling. Every nth element in the population is chosen starting from a random point in the population frame. Easy to use if population frame is available. Systematic biases are possible. Population is first divided into meaningful segments; thereafter subjects are drawn: Most efficient among the probability designs. in proportion to their original numbers in the population. based on criteria other their original population numbers. Groups that have heterogeneous members are first identified; then some are chosen at random; all the members in each of the randomly chosen groups are studied. Population frame for each stratum is essential. Would adequately represent strata with low numbers. In geographical clusters, costs of data collection are low. The least reliable among all the probability sampling designs

Sampling Design Description Advantage/Disadvantage 5. Area sampling Cluster sampling within a particular area or locality. Cost–effective. Useful for decisions regarding location. 6. Double sampling The sample or a subset of the sample is studied twice. Offers more detailed information on the topic of study. Original biases, if any will be carried over. 7. Convenience sampling The most easily accessible members are chosen as subjects. Quick, convenient, less expensive. Not generalizable at all. 8. Judgment sampling Subjects selected on the basis of their expertise in the subject investigated. Sometimes, the only meaningful way to investigate. 9. Quota sampling Subjects are conveniently chosen from targeted groups according to some predetermined number or quota. Very useful where minority groups participation in a study is critical. Not easily generalizable. Nonprobability sampling

Issues of Precision and Confidence in Determining Sample Size Precision refers to how close our estimate is to the true population characteristics. Confidence Whereas precision denotes how close we are in estimating the population parameter based on the sample statistic, confidence denotes how certain we are that our estimates will really hold true for the population.

Efficiency in Sampling Efficiency in sampling is attained when for a given level of precision (standard error), the sample size could be reduced, or for a given sample size, the level of precision could be increased. The factors affecting decisions on sample size: 1. The extent of precision desired (the confidence interval) 2. The amount of risk allowable in predicting that level of precision (confidence level) 3. The amount of variability in the population itself 4. The cost and time constraints 5. The size of the population itself

Summary Decisions regarding sampling are important aspects of research design. Sampling design decision include both the sampling plan to be used and the sample size that will be needed. Probability sampling plans lend themselves to generalizability and nonprobability sampling designs do not. Some probability plans are more efficient than others. Though nonprobability sampling designs are not readily generalizable, they are often useful for obtaining certain types of information quickly and relatively inexpensively. The sample size is determined by the level of precision and accuracy desired in estimating the population parameters, as well as the variability in the population itself. The generalizability of the findings from a study of the sample to the population is dependent on the sophistication of the sampling designs used, which includes the sample size used in the study. In all research, care should also be taken not to overgeneralize the results of the study to populations that are not represented by the sample. This is a common problem in many research studies. Sample data are used for both estimating population parameters and hypothesis testing. In the next two chapters, we will see how the data that are gathered from a sample of respondents in the population will be analyzed to test the hypotheses generated answer the research questions.

CHAPTER 9 A REFRESHER ON SOME STATISTICAL TERMS AND TESTS

Descriptive Statistics Frequencies simply refer to the number of times various subcategories of a certain phenomenon occur, from which the percentage and the cumulative percentage of the occurrence of the subcategories can be easily calculated. Measures of central tendencies and dispersion – Mean – Median – Mode

Inferential Statistics Correlations Relationship among nominal variables: x 2 test Significant mean differences between two groups: the t test Significant mean differences multiple groups: ANOVA Multiple regression

Summary In this chapter, we briefly revisited some statistical terms and tests. Now that we have refreshed our memory with the necessary background materials, in the next chapter, we will discuss how the data are actually analyzed. The Supplementary Readings List offered for this chapter at the end of the book will be useful if more in depth understanding of the various statistical tests is needed.

CHAPTER 10 DATA ANALYSIS AND INTERPRETATION

Flow Diagram of Data Analysis Process Data analysis DATA COLLECTION Interpretation of result Discussion Feel for data Getting data ready for analysis Editing data Handling blank responses Coding data Categorizing data Creating data file Programming 1. 2. 3. 3. 4. Goodness of data Mean Standard deviation Reliability Validity Correlations Frequency distribution, etc. Hypotheses testing Appropriates statistical manipulations Research question answered?

Getting Data Ready for Analysis Editing data Handling blank responses Coding Categorization Keying data

Basic Objective in Data Analysis Feet for the data Testing goodness of data

Validity Factorial validity can be established by submitting the data for factor analysis. The results of factor analysis (a multivariate technique) will confirm whether or not theorized dimensions emerge.

Reliability The reliability of a measure is established by testing for both consistency and stability. Consistency indicates how the items measuring a concept hang well together as a set.

Hypothesis Testing Once the data are cleaned up (i. e. , out of range/missing responses, etc. , are taken care of) and the goodness of the measures is established, the researcher is ready to test the hypotheses developed for the study.

Data Analysis and Interpretation Data analysis and interpretation of results can be most meaningfully explained by referring to a business research. After a very brief description of the background of the company in which the research was carried out and the sample, we will discuss the data analysis done for testing each hypothesis and how the results were interpreted.

Summary In this chapter we covered the procedure for analyzing data after they have been collected. By means of an example, we saw the steps necessary to get the data ready for analysis editing, coding, and categorizing. Through the example of the research on KRIYA Enterprises, we saw various statistical analysis and tests used to examine the different hypotheses to answer the research question. We also saw how the computer results are interpreted. We also described some microcomputer programs and discussed the SPSS/PC+ Studentware in some detail. Some expert systems for data analysis and managerial decision making and problem solving were also mentioned. In the next chapter we will learn how to write a research report after the data have been analyzed and the results are interpreted.

CHAPTER 11 THE RESEARCH REPORT

The Written Report The written report enables the decision maker to weigh the facts and arguments presented in the document and implement the final solutions(s), with a view to closing the gap between the desired and the present state of affairs in any given problem area.

Integral Parts of the Report The title of the research report Table of contents The synopsis The introductory section – Introduction – A brief literature survey – Research question – Theoretical framework – Hypotheses – Study design

Methods section – Population and sample for the mechanization study – Variables and measures used – Data collection methods – Data analysis techniques Result section Discussion section Recommendations and implementation

Summary Section The summary section highlights the research question and the answers found by doing the study and also briefly recounts the recommendations and the implementation. It should be elaborate than the synopsis.

Acknowledgements An acknowledgement of the help received from others is made in the acknowledgements section. Usually, the people who allowed the researcher(s) access to the organization are thanked, as well as those who assisted in the study by collecting the questionnaires, acting as liaison persons, helping in data analysis, and so on.

References Immediately after the acknowledgements, starting on afresh page, appears a list of the references cited in the literature review and at other places in the report.

Appendix The appendix, which comes last, is the appropriate place for the organization chart, for newspaper clippings or other materials that substantiate what is said in the text of the report, for verbatim narration of interviews with organizational members, and for anything else that is considered useful for following the text.

Oral Presentation Deciding on the content Visual aids Delivery Handling questions

Summary The components of various types of written research reports were discussed in this chapter. It was emphasized that the purpose of the report and the intended audience are critical factors in deciding what aspects of the study will be stressed the most. Tips regarding oral presentation were also offered. Three different styles of report, referred to and discussed in the text, are provided in the appendix to this chapter.

MULTIVARIATE DATA ANALYSIS Oleh: Prof. Dr. Djumilah Zain, SE

Factors Affecting the Performance of Individual Chain Store Units: An Empirical Analysis Richard T. Hise. Myron Gable J. Patrick Kelly lames B. Mc. Donald Retail chain store executives are constantly faced with the problem of achieving success through the choice of a combination of decision variables. Although many executives are increasingly using more sophisticated tools for decision making, there still appear to be instances of decisions by intuition, hunch, or untested rules of thumb. There a number of variables or factors which can have an impact on the success of individual chain store units. They include such variables as product offerings; store location; strength, number, and strategies of competitors; promotional‑efforts; store factors, such as store size, inventory levels, and number of employees; store manager characteristics, including such factors as the store manager's experience, age, marital status, and educational level; and market factors, such as disposable income and population. While non retailing industries have had large scale studies conducted on the factors that affect their performance, especially in the area of return on investment (Schoeffler, Buzzell, and Heany 1974; Buzzell, Gale, and Sultan'1975), few such studies have been done in the retailing sector. Those that have are relatively small‑scale in nature and are now somewhat dated. One early study in retailing focused upon the variety of merchandise offered by a store and its impact on sales, cost, and profits (Baumol and Ide, 1962). This study indicated that inventory level and variety of merchandise had a positive effect on sales and profits. A later study focused upon how profit was influenced by a number of internally controllable operating factors, such as markdowns, rent, publicity, sales volume, stock turnover, and average sale (Dalrymple 1966). In that study, sales volume was found to explain the greatest variance in profit levels.

None of these prior studies has dealt with the importance. 9 f relatively uncontrollable, long‑run, and irreversible variables on store performance. The present study reports the results of an analysis of the impact of 18 independent variables‑controllable/uncontrollable, short‑run/long‑run, and reversible/irreversible‑on three performance factors of 132 retail chain store units. The study's objectives, all relating to individual retail chain store units, were: – To determine the aggregate impact of these 18 independent variables on the three performance measures. – To identify the effect of four major groups of predictor variables on the three performance measures. – To identify those individual independent variables which had the greatest impact on the three measures of performance. – To develop some conclusions as to the value of using these independent variables to predict unit performance. – To formulate some recommendations regarding a marketing strategy that chain store executives should consider in efforts to improve the performance of their retail stores. Methodology The 18 independent variables used to predict the chain stores' performance can be grouped into four major areas. These variables are identified with their mean, minimum, maximum, and stan dard deviations in Table 1, The performance factors (or dependent variables) were sales vol ume, contribution income (gross margin less direct expenses), and return on assets. The chain's top management considered return on assets to be the most important of the three performance variables.

The executives of this large retailing corpora tionagreed to provide data on all its units. These data consisted of the three performance measures and three of the four groups of independent variables: store, competition, and location. The store manager factors were obtained from a questionnaire administered to each store manager. A total of 179 out of 180 units responded the questionnaire. Some of the secondary data the units contained missing or unusable data. Because of missing and unusable data, 37 units were dropped from the analysis; the usable sample was thus 132 units. The corporation sells non clothing items in stores located In malls throughout the United States. (Company executives asked the authors not to reveal the name of the company or the nature of its products. ) Although the product is essentially a shopping good, customers frequently make purchase decisions before visiting the retail outlet, so the product sold tends to take on some of the characteristics of a specialty good. Average annual sales volume for these stores was approximately $565. 000; their average size was slightly over 3, 300 feet, and the average number of employees, including the store manager, was 7. 4. The average number of years the stores had been open was 5. 7, and the average mall size in which the stores operated was 778, 100 square feet. Because the product assortment was similar for each store, and their promotional budgets, copy, and media were also consistent and basically the same, these factors were not included as independent variables.

The nine store manager variables were in cludedas independent variables because of previ ous research on retail employees. While previous studies did focus on store managers, their results are perhaps indicative of what might be found if store managers were analyzed. Weaver (1969) found that older, better‑educated, married, or divorced salespersons were more productive. Cotham (1969) found that retail employees who were old, had prior retailing experience, and had worked for another retail firm were more productive. Paul and Bell (1968) identified older, more experienced, and slightly better educated retail employees as being more productive. They also found retail employees who "worked harder ” were able to generate higher sales. These results suggest that retail store managers who are older, married, and have more children might be more serious about their work and therefore more pro ductive. While some of the other nine variables have been used in previous research (Baumol and Ide 1962; Dalrymple 1966), several were used because of their availability from company records and the intuitive belief that they would likely be an important determinant of chain store units’ performance. For example, because the products sold are basically shopping goods, the presence of secondary, rather than primary, competitive stores should increase sales and profits. Larger malls with more traffic and larger SMSAs would tend to suggest high sales volume and possibly better profit results. Stores with more inventory and more sales help should result in more accept able products and greater sales rates. Stores that were in business longer would have overcome the start‑up problems that initially would adversely affect sales and profits in newer locations.

The above factors can be viewed in terms o their reversibility, the control that management can exert, and their time duration. Store location land store size, especially in mall locations, are examples of somewhat irreversible factors whereas, for example, inventory levels and num ber of employees are relatively reversible. Although such factors as store manager characteris tics and number of employees are factors over which management can exert a good deal of control, market dimensions and various actions by competitors are largely beyond their control.

JOURNAL The big five personality model: comparing male and female entrepreneurs The key success factors, distinctive capabilities, and strategic thrusts of top SMEs in Singapore Market share, profits and business strategy Strategy and management control systems: a study of the design and use of management control systems following takeover The effect of retail store environment on retailer performance

THE BIG‑FIVE PERSONALITY MODEL: COMPARING MALE AND FEMALE ENTREPRENEURS Brooke R. Envick St. Mary's University Margaret Langford, St. Mary's University ABSTRACT This study differentiates female entrepreneurs from male entrepreneurs using the Big‑Five Personality Model. The five factors include adjustment, sociability, conscientiousness, agreeableness, and intellectual openness. Adjustment determines confidence versus instability. Sociability measures extraversion versus introversion. Conscientiousness determines impulsiveness versus cautiousness. Agreeableness measures team‑orientation versus self‑interest. Intellectual openness involves practicality versus originality. Results indicate that female entrepreneurs are significantly more open than male entrepreneurs. They are also more adjusted, social and agreeable, but not to a significant degree. Male entrepreneurs are significantly more conscientious than female entrepreneurs.

INTRODUCTION Today, women are starting businesses at a rate twice that of men (Allen, 1999). The Small Business Administration estimates that by the end of the year 2000, more than 40% of all businesses will be owned by women (Bygrave, 1997). Women‑owned businesses employ more than 15 million workers in the United States and the sales generated amount to approximately $1. 4 trillion (Nelton, 1996). With these demographic trends, interest continues to grow in the personal characteristics of female entrepreneurs, especially those factors that might explain their success. Research has predominantly focused upon the similarities and differences between male and female entrepreneurs demographically as well as psychologically. The purpose of this paper is to continue the study of female versus male entrepreneurs. The Big Five Personality Model (Goldberg, 1990; Goldberg, 1992; Goldberg, Sweeney, Merenda, & Hughes, 1998) that has recently emerged from the field of psychology into business applications is used to analyze both genders. The paper compares female entrepreneurs to male entrepreneurs on each of the five factors. First, we describe the Big‑Five Personality Model and discuss its recent applications to business research and gender differences. Next, we review research regarding similarities among and differences between male and female entrepreneurs and suggest hypotheses regarding the Big‑Five Model. Then, we describe our research methodology. Finally, we discuss results of the study and draw conclusions.

LITERATURE REVIEW With some controversy in the psychological community, the Big‑Five Personality Model emerged in recent years as a "robust model” or "Great Theory" of personality. While a discussion of theoretical arguments pertaining to the Big‑Five is beyond the scope of this paper, its proponents believe that the model is robust in that the personality of every human being, regardless of his or her culture, can be described utilizing the five dimensions (see Costa & Mc. Crae, 1995; Digman, 1990; Goldberg, 1990, 1992; Goldberg, Sweeney, Merenda, & Hughes, 1998, Wideger & Trull, 1997). Disagreement exists regarding the exact vocabulary of the five factors (or superfactors); however, conceptually, the factors are these: (1) adjustment (on a continuum from stable to neurotic), (2) sociability (from extroverted to introverted), (3) intellectual openness (from imaginative and interested in many things to practical and narrowly focused), (4) agreeableness (from benevolent to belligerent), and (5) conscientiousness (from dependable and goal‑oriented to undependable and impulsive). The interest of psychologists is not in describing a universal "right" personality (there is none), but rather in examining a person's "score" on each of the five factors in conjunction with other factors (e. g. , education, age, gender, job). Recently, researchers have reported the Big‑Five results contain implications for the workplace.

The Big‑Five in the Workplace In jobs involving personal interactions, one study reported that the factors of conscientiousness, agreeableness, and adjustment were related to job performance. Not surprisingly, emotional stability and agreeableness wore found to be especially important in jobs involving teamwork (Mount Barrick, & Stewart, 1998). With business franchise owners as subjects, Morrison (1997) examined the relationships between the Five‑Factor Model and other psychological constructs (e. g. , Self‑Monitoring, Type A Behavior, Locus of Control, and Subjective Well‑being). Results indicate that franchise owners tend to be Type A persons who are more sociable and conscientious than not. They are relatively more agreeable than not, slightly less open to new experiences than average. As a group, franchise owners tend to have an internal locus of control, which is also strongly associated with adjustment. The results of a study by Collins and Gleaves (1998) regarding job applicants indicated no significant differences in the Big‑Five Personality Model between African American and Caucasian applicants, although both groups tended to provide socially desirable survey responses regarding the Big‑Five dimensions. Another study reported that applicants who wore more sociable, open to experience, and relatively conscientious tended to employ more effective job search strategies and were more successful in obtaining second interviews than those who did not (Caldwell & Burger, 1998).

Although each factor represents a collection of traits, the link between personality and behavior becomes clearer when only one trait is the focus rather than one factor. There are several common personality traits that render a natural fit into one of the five factors. For example, locus of control is considered to be a part of the conscientiousness factor as it relates to job performance behaviors regarding dependability and responsibility (Lefcourt, 1992; Black, 1990). Another example is self‑esteem. People with high self‑esteem are more likely to take risks and enter difficult and unconventional occupations because they believe in their abilities. This is an important part of the adjustment factor as it relates to stability and confidence (Ellis & Taylor, 1983; Hollenbeck & Brief, 1987). Other workplace‑related studies utilized the Five‑Factor Model include those involving employee absence (Judge, Martocchio, & Thoresen, 1997), expatriate success (Ones & Viswesvaran, l 997), job performance in the European Community (Salgado, 1997), and teamwork (Neuman, Wagner, & Chnstianson, 1999). The Big‑Five has also been applied in gender studies.

The Big‑Five and Gender Pertinent to the current research, Lippa (1995) found that sociability, openness, and low levels of adjustment wore the factors most linked to "masculinity, " while agreeableness and conscientiousness were linked to "femininity" (note: not all males in the study measured as “masculine” and not all females as "feminine"). In a similar study, Marusic and Bratko (1998) stated that sociability was highly associated with "masculinity" and agreeableness with. "femininity”. A low adjustment score was associated with both high "feminine" and low "masculine" subjects. Using data collected in an on‑going longitudinal research study (over twenty years), Pulkinnen (1995) reported adult females who had been identified previously as "conflicted" tended to be less adjusted, more introverted and less conscientious and open to experience than females identified as “adjusted”. “Conflicted” males were less adjusted and conscientious than "adjusted” males. Direct comparisons between females and males were not made. In an extensive study examining many demographic variables (e. g. , age, race) including gender, Goldberg et al. (1998), reported that men tended to be less agreeable than women, but found no significant differences in the other four factors.

Female Venus Male Entrepreneurs Past research reveals both similarities among and differences between male and female entrepreneurs. For example, early studies exploring why females become entrepreneurs found they gave similar responses to their male counterparts such as need to achieve and independence (Cook, 1982; Schwartz, 1976). Contemporary research also supports similarities. For instance, Smith, Smits, and Hoy (1992) report females in traditionally dominated male‑industries gave similar reasons for operating their own businesses such as the desire for independence. Another study reports that no differences exist regarding personal goals such as independence, achievement, and economic necessity (Hisrich, Brush, Good, & De Sotiza, 1996). Fagenson (1993) found that both males and females value self‑respect, freedom, a sense of accomplishment and an exciting life. Cooper and Artz (1995) discovered both males and females held initial optimistic expectations regarding their ventures. Sexton and Bowman‑Upton (1990) found that both males and females were low in their need for conformity with others, need for "succorance" (seeking advice or sympathy), and need for avoiding harm. Male and female entrepreneurs were both high in "interpersonal affect" (they displayed compassion, were not aloof, and related well to others), and "social adroitness" (they were skillful at persuading others, diplomatic but somewhat manipulative).

On the other hand, several studies contend there are differences between male and female entrepreneurs. Envick and Langford (1998) found that female entrepreneurs engage in controlling, internal communication, human resource management, and work‑related task behaviors more often than male entrepreneurs. The National Foundation for Women Business Owners found women define success very differently from men. Women see success as having control over their own destinies, building ongoing relationships with clients, and doing something fulfilling, while males define success in terms of achieving goals (Romano, 1994). Smith et al. (1992) found that female entrepreneurs employ more females than male entrepreneurs in male dominatedindustries and select females with whom they share similar attitudes. Fagenson (1993) reveals that females value equality and world peace more than males. A longitudinal study conducted by Gatewood, Shaver, and Gartner (1994) found female entrepreneurs have higher internal attributions for starting their ventures than males. However, Brandstatter (1997) found that male entrepreneurs made internal attributions regarding either failure or success of a venture and were significantly less likely than women to make external attributions (e. g. , the prevailing economy) for either failure or success. One study investigated entrepreneurs and family‑career conflict (Parasuraman, Purohit, Godshalk, & Bcutell, 1996) and found that females reduce family‑career conflict by spending less time at work, while males increase their time at work. Sexton and Bowmen‑Upton (1990) found male and female entrepreneurs to differ significantly in four traits. Males had higher sustainable energy levels and were more risk‑taking than females. Female entrepreneurs desired autonomy more and were more open to new experiences than males.

HYPOTHESES The hypotheses are generated based upon empirical findings regarding the Big Five Model and previous research regarding similarities among and differences between female and male entrepreneurs. There is one hypothesis for each of the five factors. H 1: Female entrepreneurs will score higher than male entrepreneurs on the sociability factor. Behavioral research reveals that female entrepreneurs engage in communication activities more often than male entrepreneurs (Envick & Langford, 1998). Other findings indicate that women are more likely to encourage participation, share information and have good interpersonal skills (Rosener, 1990; Eagly, Malchijani, & Klonsky, 1992; Offiermann, & Beil, 1992). Showing concern and being relationship‑oriented are more characteristic of females than males (Coppolino & Seath, 1997; Porter, Geis, Cooper & Newman, 1985; Vroom & Jago, 1982). The Big‑Five research suggests that while masculinity is related to extroversion, not all males are considered 'masculine', while not all females are considered 'feminine’ (Lippa, 1995).

H 2: There will be no significant difference between male and female entrepreneurs regarding the adjustment factor. No previous research regarding the adjustment factor suggests a gender difference. Lippa (1995) found that a low adjustment score was related to masculinity, while Marusic and Braiko (1998) found low adjustment relating to 'femininity. Again however, not all male subjects are considered 'masculine, while not all female subjects are considered 'feminine'. Therefore, no rationale exists to hypothesize a significant difference in either direction. H 3: Female entrepreneurs will score higher than male entrepreneurs on the openness factor. Sexton and Bowman (1990) found that female entrepreneurs desired more autonomy and were more open to new experiences than male entrepreneurs. Fagenson (1993) discovered that female entrepreneurs had a much broader vision involving their desires including total equality and world peace. Therefore, it is logical to assume that female entrepreneurs will be more open than male entrepreneurs. H 4: There will be no significant difference between male and female entrepreneurs regarding the conscientiousness factor. While conscientiousness is linked to 'femininity' (Lippa, 1995), no rationale exists to hypothesize a significant difference in either direction. Femininity is used to describe both males and females. No other findings suggest a difference in gender‑related or business research.

H 5: Female entrepreneurs will score higher than male entrepreneurs on the agreeableness factor. Goldberg (1998) found that, in general, females are more agreeable than males. Previous research suggests that female entrepreneurs are more supportive, encourage participation, and adopt a more democratic style than male entrepreneurs (Tannen, 1991; Offermann, & Beil, 1992; and Rosener, 1990). Smith, Smits, & Hoy (1992) found that female entrepreneurs actively seek female employees with whom they share similar attitudes. Fagenson (1993) contends that female entrepreneurs value equality more than their male counterparts.

METHODOLOGY The hypotheses are tested using ANOVA to determine if significant differences exist between entrepreneurs and managers on all five factors. One hundred and nineteen subjects represent the findings, 86 males and 33 females. Although these two comparisons are disproportional in number, it is representative of the entrepreneurs in this geographical area (approximately one third females, two‑thirds male). The Chamber of Commerce in a large Southwestern city generated a list of entrepreneurs, and 650 were randomly selected from this list to receive the survey. With a response rate of over 20%, 130 surveys were returned, and 119 were usable. All subjects received a survey containing background questions regarding their job role and type of business. The Big‑Five Model was tested using the questionnaire developed by Howard, Medina, and Howard (1996), which is commonly used by consultants and trainers and published in textbooks (Hellriegel, Slocum & Woodman, 1998). The survey included twenty‑five sets of descriptive words on opposite ends of a continuum. Respondents were asked to circle the number on the continuum. that most closely describes their personality. Each of the five factors is measured by the sum of scores received on a total of five questions. The highest score possible is a 35, while a 5 is the lowest score possible.

RESULTS ANOVA was used to test all five hypotheses in order to compare entrepreneurs to managers on each personality factor. The first hypothesis tests sociability. The second tests adjustment. The third hypothesis tests openness. The fourth one tests conscientiousness. And the fifth hypothesis tests agreeableness. Table 1 presents the all means, standard deviations, and p‑values. The first hypothesis, regarding sociability, is not supported. However, the general direction of the hypothesis holds true with females scoring higher (M‑‑ 18. 061) than males (M‑‑ 16. 977). The second hypothesis is supported. No significant difference exists between males and females regarding adjustment. The third hypothesis is supported. Females (M=16. 333) are significantly more open [F(1, 118) = 1. 950; p<. 011 than males (M=14. 407). The fourth hypothesis is not supported. Males (M=19. 093) scored significantly higher on the conscientiousness factor [F(1, 118) = 3. 262; p<. 051 than females (M=17. 455). The fifth hypothesis is not supported. However, the general direction appears to hold some merit. Females are more agreeable (M=19. 667) than males (M=18. 884).

DISCUSSION This paper makes a contribution by further identifying psychological traits that illustrate similarities among and differences between female and male entrepreneurs. While several psychological characteristics have been analyzed in order to identify the two groups, the Big‑Five Model provides another avenue to further define and describe each group. Two of the five hypotheses are supported. Neither male nor female entrepreneurs are more adjusted than the other. Female entrepreneurs are significantly more open than their male counterparts. While hypotheses one and five are not supported, the general direction holds true. Female entrepreneurs are wore sociable and agreeable than male entrepreneurs, but not to a significant degree. The fourth hypothesis is not supported suggesting that no differences would be present on the factor of conscientiousness. However, male entrepreneurs scored significantly higher on this factor, meaning that they wore more cautious and less impulsive than females. This is a mystery, since the only explanation in the literature is that both male and female groups that score high on adjustment also obtain high scores on conscientiousness (Pulkinnen, 1995). This is not the case in this study. Females actually scored slightly higher on adjustment. Perhaps, this finding is unique to entrepreneurs. This study does provide more insight into the psychological profile of male and female entrepreneurs. While most of the findings are not surprising, an interesting research question presents itself‑why did male entrepreneurs score significantly higher on the conscientious factor than female entrepreneurs, when females did obtain the higher adjustment scores. Further research is certainly needed regarding the Big‑Five and entrepreneurs.

REFERENCES Allen, K. (1999). Launching new ventures: An entrepreneurial approach. Boston: Houghton Mifflin Co. Black, J. S. (1990). Locus of control, social suppo^ stress, and adjustment in international transfer Asia Pacific Journal of. Management, April, 1‑ 30. Brandstatter, H. (1997). Becoming an entrepreneur: A question of personality structure? Journal of'Economic. Psychologv, 18, 157‑ 177. Bygrave, W. D. (1997). The portable MBA in entrepreneurship. New York: John Wiley & Sons, Inc. Caldwell, D. F. & Burger, J. M. (1998). Personality characteristics of job applicants and success in screening interviews. Personnel Psychology, 51, 119‑ 136. Collins, J, M. & Gleaves, D. H. (1998). Race, job applicants, and the Five Factor model of personality: Implications for black psychology, industriallorganizational psychology, and the Five‑Factor theory. Journal of. Applied Psychology, 83, 531‑ 544. Cook, J. (1982). Women: Ue best entrepreneurs. Canadian Business, June, 68‑ 73. Cooper, A. & Artz, K. (1995). Detenninarits of satisfaction for entrepreneurs. Journal of. Business Venturing, 10, 439‑ 457. Costa, P. T. , Jr. & Mc. Crae, R. R. (1995). Four ways five factors are basic. Personality and Individual Differences, 13, 653‑ 665. Digman, J. M. (1990). Personality structure: Emergence of the five‑factor model. Annual Review of. Psychology, 417‑ 440. Eagly, A. H. , Makhijani, M. G. & Klonsky, B. G. (1992). Gender evaluation of leaders: A meta analysis. Psychological Bulletin, Jan. , 3‑ 22.

JOURNALOF BUSINESS RESEARCH THE KEY SUCCESS FACTORS, DISTINCTIVE CAPABILITIES, AND STRATEGIC THRUSTS OF TOP SMEs IN SINGAPORE B. C. Ghosh*, Tan Wee Liang, Tan Teck Mengl, Ben Chan Abstract The research tries to determine the strategy dynamics and key success factors (KSFs) for excellence in performance of the so‑called "tiger" SMEs in Singapore. In 1995 and 1996, 50 top privately owned and successful enterprises in Singapore were identified. They have shown that they can excel, even in the current highly competitive and high o~on‑cost environment. Their performance can he attributed to their dynamism and a few KSFs that am apparently universal to these successful companies. The strategy dynamics and their specific components (i. e. , the six top SMEs) are found to be: 1) A committed, supportive, and strong management team. 2) A strong, visionary, and capable leadership. 3) Adopting the correct strategic approach. 4) Ability to identify and focus on market. 5) Ability to develop and sustain capability. 6) A good custouier and client relationship‑ Approximately 60*% of the companies surveyed were found to be of Defender type organizations (Miles and Snow typology). As a majority of the companies are from mwufi~g and servicing sectors and from OEMs supporting the MNCs, it is not surprising that the Defender type strategy is predominant. However, such organizations may have to evolve in order to adopt a more superior strategy such as Prospector and Analyzer when environment changes. The importance placed by organizations adopting different sb~ types on their strategic posture are different although KSFs and capabilities are generally universal. Comparing proactive with passive strategy types, the degree of emphasis given various success factors by proactive type companies was generally found to be higher. Specifically, proactive type companies placed higher importance on the following factors for excellent performance: (i) Satisfying customers needs, (ii) Close working relationship between top management and employee, (ifi) Regionalization, (iv) Leadership, (v) Availability of financial and technology resources; and support Further, the research also found that the importance attached to the various strategy‑rdated sucem factors changes with the development. As it becomes more established, the ranking of KSFs changes as the organization faces different challenges when competition becomes tougher. Enterprises need to pay attention to the following dynamics and strategic thrusts: (a) Strong market orientation and relevant capability; (b) Effective management; (1) Strong management commitment and support (2) Strong organizational capability and management cohesiveness; (c) Access to broad base support and resources (i. e. , Networking). 0 2000 El~ Science Inc. All rights reserved. Keywords: Key success factors; Strategic thrust; SMEs

1. Introduction Businesses have been finding it hard to operate in Singapore owing to the high operating cost and increasingly intense competition from the region. Coupled with the fact of small local market size, and relying very much on export market, the business environment has become even more difficult to operate. However, there are companies that excelled. The ‘Top 50’ privately held enterprises for 1995 and 1996 have shown that they can excel even under these adverse conditions. The objective of this research is to identify and examine the strategic key success factors (KSFs) [key success factor is defined for our purpose as factors which are critical for excellent performance of the company, rather than just survival which is the function of critical success factors (CSFs)]. The successful privately held companies are potential candidates for public listing if they need to raise public funds. Some of them have crossed a turnover of S$800 million and others have yet to cross the S$100 million turnover mark. But they are all on the fast track, growing rapidly and regionalizing their operations. Some of those that appeared in the 1996 listing were already publicly listed at the end of that year. The 'Top 50' enterprises were ranked based on their performance over a 4‑year period. Various quantitative factors were used such as gross turnover, profits, return per employee, return on assets, and percentage of business from overseas operations. The evaluation team, from Andersen Consulting, The Business Times, and The Economic Development Board, developed and applied a balanced scorecard of these performance indicators for companies across a wide mix of industries. A key consideration in the scorecard is giving a balanced weight to both absolute financial performance and growth. The aim is to capture how fast the company is growing, not just how much it makes. So, aside from turnover and profit levels, the team looked at growth in turnover and profits. However this information is privy to the group and were not available to the researchers.

In the study conducted here, companies are defined successful if they appeared in the 1995 or 1996 'Top 50' Enterprises list compiled by Andersen Consulting, The Business Times, and The Economic Development Board of Singapore, a statutory board within Singapore. They form our base group whose strategies we study. 1. 1. Objectives of the study The focus of this article is to isolate the top KSFs (not CSFs) that are commonly practiced by these successful privately held companies, so that they may be adopted by any enterprise (even start‑ups). The study also examines what are the most commonly adopted strategy types by these successful companies and to determine if the KSFs correlate well with the strategy type adopted‑ We will also argue that these factors will be positivist and universal in nature. This will be one hypothesis. A second hypothesis at this stage is to identify a discriminating model that will be able to predict the strategy type from its KSFs and distinctive capabilities, such that such a model can be used for a firm's future strategic direction 1. 2. Significance of research Privately held enterprises play an important role in the growth of the Singapore economy. There already 80, 000 SMEs in Singapore employing about 40% of the workforce and contributing around 30% of the total value added in the economy. Their continual survival and success is therefore critical, especially with increasing competition from the region and uncertainties associated with multinationals. By identifying the KSFs of top successful local enterprises, and many of them are SMEs, one will gain insight and understand better how they remain competitive and even excel in the harsh operating environment in Singapore. The findings of the research are likely to be more relevant and applicable to local companies, though there are other generalizable aspects.

2. Some discussion of the relevant literature There were studies done on the local SM[Es to identify CSFs for business success; a significant contribution comes from Ghosh and Kwan (1996) who covered Singapore, Malaysia, Australia, and New Zealand, a cross‑national study. This study found that factors that contributed to the successful SMEs were (among others): having a good customer relationship, effective management and marketing. Constraints to success were found mainly m the high cost of doing business and competition. Similar research has been undertaken elsewhere. De. Hayes and Haeberle (1990) found that the most frequent reason for success among businesses was their ability to identify and focus on one or a few market niches. Further, they identified that most business plateau at US$10 million and that business is difficult to expand without change in management style and building up human resources in terms of quantity and quality. The finding is supported by Evans and Evans (1986) who asserted that a firm's success in competing in a hostile environment was directly related to human resources development, besides top management's involvement with all phases of the operation. De. Hayes and Haeberle (1990) further identified the following CSFs: the ability to develop and sustain technological advantage, the ability to identify and focus on one or several market niches, having strong leadership on top, having significant , people bonding' mechanism in the system, strong management team, strategic alliance with customers and strategic use of information technology. Huck and Mc. Ewen (1991) found technical knowledge and customer relations to be the competencies most important for the success of small business.

Duncan (1991) found the ability to identify a market ruche and develop a standing m that niche market as a CS 17. The findings of Prescott (1986) and Steiner and Solem (1988) were similar to those of Del. Hayes and Haeberle (1990) and Duncan (1991). Campbell (1991) identified 12 keys for a successful business, among which were a clear mission statement and a corporate value system, a customer‑oriented policy, a competitive strategy, and a personal commitment from the top management. Gaskill and Hyland (1989) identified six keys to a successful business, which included people power, a business plan, a study of the competition, performance measurement, and avoiding complacence. Mraz (1989) found that thorough planning was vital to start a successful business while Schilit (1986) found that critical success included: having a formal business plan, continual monitoring of the business environment, retaining a market orientation, developing a common value system, ensuring adequate capitalization, and encouraging entrepreneurial thinking through all levels of the company. Foley (1987) reported that, a written business plan, new product development, a strong sales and marketing team were among the prerequi sites for a successful business. Barkham (1989) and Pol lock (1989) identified skill, attitudes, and the gathering of the marketing information as factors contributing to the success of an enterprise.

Market share, profits and business strategy Keyin J. Laverty University of Washington, Bothell, Washington, USA Keywords Market share Strategy, Competitive advantage Abstract Reward into the Correlation between market share and profitability has led to debate over whether the observed association is direct or spurious. The fundamental question remains whether the pursuit of market share is an appropriate strategy. This paper reports the results of a structural equations model which provides the clearest to data of the competing viewpoints. The direct association between market share. and profitability depends upon restrictive statistical assumptions. When these assumptions are relaxed, the resu. Its show that there is no direct association between market share and profitability. Understanding the causes of profitability is a central issue in disciplines which study business firms. Business policy's distinctive approach is to focus on goals and strategies designed to achieve superior performance (Chandler, 1962; Andrews, 1971), resulting in greater profit.

The relationship between market share and profitability is perhaps the most‑studied single phenomenon in business policy. The correlation between these measures[i] is undeniable. Although it has been more than two decades since the first published studies reporting a positive market share profitability association (Gale, 1972; Shepherd, 1972), the nature of this relationship continues to receive a great deal of attention (Buzzell and Gale, 1967; Jacobson, 1988 a; Cool at al. , 1989; Boulding and Staelin, 1990; Venkatraman and Prescott, 1990; Schwalbach 1991; Montgomery and Wernerfelt, 1991; Szyimanski et al. , 1993; Fraering and Minor, 1994). While we observe that many practitioners hold the view that higher market shareleads to higher profits, research findings indicate that the market share‑profitability association is dependent upon strategic and competitive settings, and spurious effects account for at least a. sizable. component of the measured association. The purpose of this paper is to seek an answer to the most fundamental question: is the pursuit of market share an appropriate strategy? Despite studies indicating that low share businesses can be quite profitable (Woo, 1982; Schwalbach, 1991), the "dominant" finding of prior research is a significant positive relationship between market share and profitability[2].

It has become evident, however, that there are "competing theories" (Jacobson, 1988 a); Rumelt and Wensley (1981), Jacobson and Aaker (1985) and others have argued that the observed market share‑profitability association is spurious[3]. However, little attention has been paid to reconciling conflicting results. In fact, there has been essentially no acknowledgment of the different assumptions imbedded in the work that has produced these results. This paper offers the following contributions to the literature on the market share‑prolitability relationship: – a clarification of the different starting points from which the conflicting research results have arisen; – an approach to testing the competing theories, using a structural equations model with latent varilables; and – results which indicate that the dominant view of market share and profitability is misleading; that market share increases are not associated with improvements in profitability.

Prior research in market share and profitability Origins in industrial organization economics Although the share relationship has become a central concern of business policy, much of the early research was actually a reaction to then‑dominant "concentration‑profit”[4] (Bain, 1951) doctrine in industrial organization economics (l/O). For many years, a "pillar" of I/0 was the notion that more concentrated industries displayed higher profits because of "oligopolistic coordination”. Business policy was enthusiastic about the association between market share and profitability originally reported by Gale (1972) and Shepherd (1972) because an explanation for superior performance based upon market share was free of the "collusion' that tainted the concentration‑profitability doctrine. Firms could legally pursue market share (within limits), whereas they could not collude. The classic Harvard Business Review article by Buzzell et al. (1975) launched policy's treatment of the market share‑profitability association.

The direct effects viewpoint The study by Buzzell et al. (1975) and a subsequent stream of work represent the direct effects viewpoint. Buzzell and Gale (1987) argued that higher market share leads to greater profits because of market power and lower costs resulting from scale effects and learning effects. Phillips et al. (1983) found that market share affects return directly, and also has an indirect effect through the reduction of costs. Prescott et al. (1986) considered different environments (e. g. mature, declining, emerging) and a series of "conduct variables" (e. g. capacity utilization, relative price, relative quality) hypothesized to be associated with both market share and profitability. Although they argue both that the market share‑profitability relationship is contingent on the environment and that significant spurious effects are present, they found significant direct effects. A subsequent study (Venkatraman and Prescott, 1990) examined a different time period with distinct economic conditions, and confirmed the direct effects reported by Prescott et al. (1986), although the size of the direct effects changed in several environments.

The spurious effects viewpoint Rumelt and Wensley (1981) hypothesized that random "shocks" (e. g. luck in uncertain ventures) will increase both market share and profit and found that change in market share is not a significant predictor of change in return. Jacobson and Aaker (1985) argued that the "dominant" explanation of the market share‑return relationship ignores factors (such as management skill, firm culture, access to scarce resources, and luck) which may Influence both variables. They found a strong association between current year return and return from the prior year, and that the market share‑return relationship is largely the result of not controlling for factors which influence both variables. Contrasting the two viewpoints The most striking contrast between the direct effects viewpoints is the strength of the market share profitability association. Buzzell and Gale (1987) and Buzzell et al. (1975) estimated that a 1 per cent change in market share is associated with a 0. 50 per cent change in return on investment. Buzzell and Gale (1987), allowing for the existence of spurious effects, still found that at 1 per cent change in market share is associated with a 0. 34 per cent change in return on investment.

In contrast, Jacobson and Aaker (1985) estimated that a 1 per cent change in market share is associated with a 0. 1 per cent change in return on investment, and Jacobson (1988 a) reported that a 1 per cent change in market share is associated with a 0. 03 per cent change in return on investment. The two viewpoints have not only sharply differing conclusions, but also disagreement as to the proper approach for studying the observed market share‑profitability association. Figure 1 depicts the approaches implicit in the viewpoints discussed above[5]. Figure 1(a) represents the direct effects viewpoint (Buzzell et al. 1975; Buzzell and Gale, 1987); market share has a direct causal effect upon profitability. Figures 1(b). and 1(c) depict two distinct versions of the primarily spurious effects viewpoint Figure 1(b) represents the relationships investigated by Rumelt and Wensley (1981). It indicates that the association between change in market share and change in profitability[6] becomes insignificant when both are affected by a variable representing a “shock". The models used by Jacobson and Aaker (1985) and Jacobson (1988 a) are displayed in Figure 1(c). Here, when profitability from a prior period is included in the model (to control for unobserved variables), the relationship between market share and profitability becomes extremely weak.

Each study reviewed has tested different relationships, and none has tried to replicate directly the findings of the others. Therefore, it is, unclear which of the viewpoints provides the best explanation for the observed market share‑profitability association. Without a test of the competing viewpoints, it Is not clear whether the pursuit of market share is an appropriate strategy. Hypotheses The hypotheses seek to test the relationships between variables represented in Figure 1. In addition to testing the following hypotheses, the study will attempt to determine if any of the viewpoints discussed offers a preferred way to understand the market share profitability association. H 1. A larger market share is associated with a higher profitability. H 2. A larger growth in market share is associated with a larger growth in profitability. H 3. Market share and growth in market share will not be predictors of profitability and growth in profitability, respectively, in a mode, which allows for the existence of “shocks" which affect both market share and profitability.

Model, data and methods A model which incorporates all of the variables displayed in Figure 1 and allows for the testing of the competing viewpoints was developed as follows. Return on investment is a common measure of profitability and has been the dependent variable in most of the cited studies. Market share, growth in market share and return on investment from a prior year, as well as a method for representing the possibility of "shocks", will be tested as predictors of return on investment for comparison with earlier studies. In establishing a, broader context for the share‑return relationship, we realize that market share is well accepted as a measure of successful performance and as a "springboard" for future success, even among those who dispute the "dominant paradigm" (e. g. Rumelt and Wensley, 1981). If we use a measure of market share from a prior period (following Rumelt and Wensley, 1981; Jacobson, 1988 a), then this "prior" market share and return on investment from prior years are all indicators of prior performance. Similarly, growth in market share is one indicator of short‑term firm growth[7]. The influence of industry‑wide factors on a firm's profit is a centerpiece of industrial organization economics, and has become a main tenet of business policy (Porter, 1980).

The notion that a firm's growth and profitability are related to its industry has been called "shared asset" theory (Porter, 1979) and suggests that a measure of industry growth is appropriate as a predictor of firm growth and profitability[81. Similar logic suggests that a final influence on firm growth and performance is general economic activity. Cubbin and Geroski (1987), for example, model the relationships among firm‑specific, industry‑specific, and economy‑wide average returns. Figure 2 represents the context of the share‑return relationship that is hypothesized. Two independent factors, general economic activity and industry growth, influence return. These two factors each also influence two other factors, prior firm performance and short‑term firm growth, which in turn also influence return. Finally, prior firm performance influences short‑term firm growth. 'Recall from the above discussion that prior firm performance incorporates the market share previously achieved by a firm; short‑term firm growth incorporates the change in market share. ) Figure 3 summarizes the key relationships. Each of the is represented here.

The data used for the 3 tudy are drawn directly from the sample generated by Rumelt and Wensley (1981). That studied 976 business units selected from the profit impact of market strategies (PIMS) data base oi the Strategic Planning Institute[10]. This is a multi‑year data base which includes a wide variety of information on the firms participating in the PIMS project. The measure of profitability/return chosen for this study was return on investment (RGI). This measure and the other variables used in the study are displayed and defined where necessary in Tabie 1. Structural equations model The structural equations program EQS[11] was chosen for the analysis because of its ability to model the system of equations suggested by the relationships shown in Figure 3. It is felt that the use of such an approach allows for testing the adequacy of the competing hypotheses that have been established based upon earlier work. Second, EQS will allow us to model the share‑return relationship to be included in a broader context that includes industry growth and general economic activity. The correlation between the "epsilons" (residuals) represents the "shock" modeled by Rumelt and Wensley (1981)[9] and allows for the existence of unobserved effects (Jacobson, 1988 b). (This technique is discussed in more detail in subsequent sections. )

Latent factors The EQS program allows for the estimation of latent (i. e. unmeasured) factors which underlie measured (i. e. observable) variables[12]. The latent factors for the data shown in Table II are postulated based upon the use of an exploratory factor analysis model. Estimation The procedure chosen was to first run OLS regressions to replicate the results of other researchers. The next step was to estimate a structural equations model based upon the relationships displayed in Figure 2 and the variables and latent factors identified above. Then, consistent with the hypothesis concerning the effects of allowing for "shocks", the residuals of market share and return on investment were allowed to be correlated. Finally, the "best fitting" model was determined.

Strategy and management control systems: A study of the design and use of management control systems following takeover Fredrik Ni. Isson Deloitte Consulting and Linkoping University and Institute of Technology, Sweden Abstract This paper describes and analyses the approach taken by four well‑known Swedish companies to management control following takeover. The findings suggest two factors which can explain how the management control systems were designed and used after an acquisition: the corporate strategy of the acquirer and the business strategy of the acquired company. The case studies show these forces could impose mutually inconsistent requirement's on the management control system of the acquired firm, and also how these inconsistencies were resolved. Key words: Acquisitions; Business strategy; Corporate strategy; Contingency theory; Management control systems

1. Introduction Management control systems have been recognised as important in the formulation and implementation of strategies (Dent, 1990; Bromwich and Bhimani, 1994). The orientation of corporate and business strategy should, therefore, be reflected in the design and use of the management control systems (MCSs) at the respective organisational levels (Shank and Govindarajan, 1993; Langfield‑Smith, 1997). Particularly after takeovers, when a clear strategic profile is important for the enlarged company (Jemison and Sitkin, 1986), it is essential that the systems of control be designed and used in accordance with the chosen strategies (Goold and Campbell, 1987). One explanation for the high proportion of failed acquisitions is the difficulty of establishing a Management control system (MCS) that will suit both the acquiring and acquired companies (Jones, 1983; Harrison et al. , 1988). This study examines the implications for the design and use of MCSs after takeover. While there are many components that make up a MCS, we focus on the following key elements: the strategic planning and budgeting system, the performance measurement system, including budget variance information and other financial and non‑financial data (i. e. RO 1, level of orders, quality statistics, stock levels, lead time), capital expenditure procedures and transfer pricing systems.

The study has three purposes. First, we examine how the MCSs of acquiring firms are designed and used for the purpose of implementing the chosen corporate strategy. Acquiring companies tend to impose their own MCS oil the acquired company (Jones, 1985 a, 1985 b). It is argued that this is necessary to achieve conformity in administrative practices, and to ensure effective monitoring and control of the acquisition (Caulkin, 1975). It is also possible that the acquiring company's control system is used for the purpose of exploiting synergies and implementing the chosen corporate strategy. While there has been limited research attention examining this rationale, a study by Goold and Campbell (1987), suggests that implementation of corporate strategy is facilitated in cases where the corporate strategic planning system is consist ent with the system implemented at the business‑unit level. If this is the case, it is likely that corporate strategy will be ail important driving force influencing the way in which MCSs are designed and used following takeover. The second purpose of the study is to explore the problems created in the acquired company when the requirement of conformity is imposed. One of the more obvious prob lems is the danger of declining morale if the acquired company must abandon a control system that is well‑suited to its needs. The acquired company, in effect, becomes a business unit of the larger entity. However, it faces its own environment, technological and strategic imperatives and will have designed its MCS to suit these conditions (Jones, 1983).

While the literature has recognised the importance of these conditions on the design of MCSs at the business unit level, there is no empirical research, of which the author is aware, that has examined the role of the acquired firm's control system following the takeover. We explore how the business strategy of the acquired firm influences the design and use of the MCS after takeover. In sum, we are interested in examining the corporate strategy of the acquirer and the business strategy of the acquired company as the two major driving forces influencing the design and use of the MCSs. Prior research has concen trated on studying these two driving forces separately (Gupta and Govindarajan, 1991; Bruggeman and Van der Stede, 1993; Langfield‑Smith, 1997). There is no empirical research examining when these two driving forces should lead to the same requirements in regard to the acquired company's control systems, and when they should lead to different requirements. The third purpose elf this paper is to examine the combination of corporate and business strategy that leads to a "fit" between the control systems of the acquirer and the acquired firms, and what combinations create a "misfit". We also use our data to provide some guidance for the management of a potential misfit in control system design. The remainder of the paper is structured as follows: Section 2 presents theoretical framework for the study. Section 3 is devoted to the method and research design. Section 4 discusses the empirical aspects of the study, begin ning with a description of the case‑study companies and their strategies. Finally, Section 5 presents some general conclusions.

2. Theoretical starting points 2. 1 The role management control systems in implementing corporate strategy Based on the concepts of synergy potential and the need for integration, Porter (1987) identified two corporate strategies 2 that may be considered diametrically opposed: portfolio management and activity sharing. With a portfolio‑management strategy there is a high degree of diversification since the corporation is involved in unrelated industries. Given the limited synergy potential, there is little need for co‑ordination, and the business units can oper atevirtually as independent companies. On the other hand, a strategy of activity sharing is characterised by limited diversification and closely related business units. In this case there should be a potential for exploiting synergies by sharing activities in areas such as manufacturing and sales. In view of the above, the corporate strategy of the acquirer is clearly a major factor in determining how far to go in co‑ordinating and integrating the business of the acquired company (Jones, 1983; Lorange et. al. , 1987). Management control systems fulfil an important function in this regard, since they provide corporate management with information for overall planning and coordination (Chandler, 1977; Goold et. al, 1994). They also play an important role at lower levels in an organisation as they can facilitate synergies among financing, production and marketing (Galbraith, 1973). An acquirer company with a portfolio‑management strategy is expected to let the acquired company operate on a more autonomous basis (Porter, 1987).

Since the synergies are primarily financial in nature, the need for co‑ordination is limited, and the control system of the acquirer will be directed at optimising the profit of each business unit (Espeland Hirsch, 1990; Hitt et al. , 1996). The strategic. planning and budgeting procedures can, therefore, be expected to be less extensive and to have a strong monetary focus. Goold and Canipbell (1987) have shown that when the acquired company is given substantial freedom to prepare its business plan and budget, this is usually combined with financially‑focused reporting procedures. The emphasis on short‑term financial performance also means that the use of monetary information is strongly emphasised in the capital‑expenditure control procedures of the acquiring company and in the key ratios used. Since the acquired company continues to operate as a largely autonomous business unit, no formalised guidelines for product transfers and transfer pricing are necessary (Anthony et al. , 1992). With this kind of relatively uncomplicated corporate‑wide co‑ordination, there is no need for far‑reaching conformity between the MCSs of the acquiring and acquired companies (Jones, 1986). In contrast, with an activity‑sharing strategy, there is a substantial need for co‑ordination and integration of the acquired company, and thus a greater need for conformity (Jones, 1986). Transfer‑pricing systems will be designed to encourage internal transfers between business units within the corporation (Anthony et al. , 1992).

Further, it is likely that the management of the acquiring company will participate in the preparation of the acquired company's strategic plan and budget in order to ensure co‑ordination occurs with major projects (Haspeslagh and Jemison, 1994). With this kind of active management, it is likely that the emphasis will shift from financial information to non‑monetary information for monitoring, controlling and co‑ordinating activities (e. g. ratios based on quantit ative, but non‑financial information) (Goold et al. , 1994). Other information will. relate to key performance targets and strategic milestones that are difficult to quantify. Non‑monetary information will also play a major role. in the evaluation of capital expenditure proposals submitted by the acquired company. For example, assessment of new capital acquisitions will consider the benefits associated with increased activity sharing across the business units. Moreover, de‑emphasis of financial performance is considered typical of control systems in corporations where business units are mutually dependent and operations are viewed frorn a long‑term perspective (Espeland Hirsch, 1990; Hitt et al. , 1996).

2. 2 The role of management control systems in implementing business strategy According to Porter (1980, 1985), a company has a choice of two business strategies for building a sustainable competitive advantage in its industry: a differentiation strategy and a cost‑leadership strategy. With a differentiation strategy, the company endeavours to establish a strong market position by prod uct offerings that are perceived throughout the industry as being unique. With a cost‑leadership strategy, the company seeks to achieve the lowest costs in its industry by standardising products and manufacturing them in long production runs. A cost‑leadership strategy, however, does not mean that quality, product image, and bases for differentiation can be ignored. Similarly, a differentiation strategy must be founded on effective cost control ‑ even though the latter is not the primary strategic objective. It has been well established in the literature that the choice of business strategy impacts on the level of uncertainty faced by managers (Hambrick, 1983, Dess and Davis, 1984; Prigogine and Stengers, 1984; Shank and Govindarajan, 1993), Uncertainty, in turn, directly influences the design and use of MCSs (Thompson, 1967; Galbralth, 1973). With a differentiation strategy, and the combination of many newly launched products and rapid changes in customer preferences, future revenues and expenses are difficult to estimate (Kald et al. , 2000), thus reducing the importance of the budget, capital

expenditure calculations based on discounted cash flows, and other formalised procedures for planning and con trol (Govindarajan, 1988). Instead, non‑monetary information is likely to be more appropriate for evaluating and monitoring strategic plans, capital expenditure proposals, and operations of the acquired company (Shank and Govindarajan, 1993). Furthermore, it is expected that a turbulent environment will lead to greater use of small‑scale, flexible production technology to permit rapid response to changes in the business environment (Vollmann et al. , 1992). Smaller scale operations reduce the need formal co‑ordination mechanisms and the implementation of formalised procedures (Bruns and Waterhouse, 1975; Anthony et al. , 1992). A cost‑leadership strategy, with a stable business environment and a low degree of uncertainty, calls for a very different kind of MCS. Since revenues and expenses can be estimated without great difficulty, the use of budgeting for monitoring performance of the acquired company is likely to be effective (Kald et al. , 2000). Capital‑expenditure calculations based on discounted cash flows are also likely to be appropriate when the business faces a low degree of uncer tainty (Shank and Govindarajan, 1993). A stable environment and the focus on low. costs, encourages the use of large‑scale manufacturing technology (Vollmann et al. , 1992). This will in turn require comprehensive operational co‑ordination of the acquired company's activities. In this situation, budgeting can serve a useful role for planning, coordinating and integrating different functions in the acquired company (Alaluusua, 1982; (Govindarajan, 1988). In addition, key financial ratios are likely to be used extensively in monitoring cost effectiveness at the business, functional, and activity levels (Hambrick and Schecter, 1983; Robbins and Pearce, 1991). It is also likely that the firm, will implement a formal transfer pricing system to encourage internal transfers and to guide the pricing decisions of the acquired company (Anthony et al, 1992).

The Effect of Retail Store Environment on Retailer Performance V. Kumar UNIVERSITY OF HOUSTON Kiran Karande OLD DOMINION UNIVERSITY Retail stores are segmented using socioeconomic characteristics of the trade area, and it is shown that the effects of store environment on store performance vary across segments. Store performance is measured by a market‑based measure‑sales and a productivity‑based measure‑sales per square feet. The internal store environment includes the number of checkout counters per square foot of selling area, the number of nongrocery products sold (extent of scrambled merchandising), whether the store at least doubles manufacturers coupons, whethere is a banking facility, and whether the store is open for 24 hours. The external store environment includes the type of neighborhood it is located in. A methodo. Iogy for predicting store performance or existing and new stores) based on the type of environment and store location by using aggregate secondary data is demonstrated. The proposed models are estimated and validated using Market Metrics geodemographic data for 646 grocery stores provided by A. C Nielsen. It is shown how the findings of this retail environment study can be used to of guidelines to retailers for attaining desired levels of sales and sales per square feet by using readily available data.

Retailing atmospherics have been used to create differences across stores in order to exploit consumer characteristics and the competitive environment (Kotler, 1973; Hoch, Byung‑Do, Montgomery, and Rossi, 1995). In the past, marketers' ability to test the variations in retail atmospherics was constrained because of nonavailability of data covering a wide range of variables across a cross‑section of retailers. However, with the advent of technology, marketers now have access to such data on a wide range of variables (for example, the Market Metrics data in the United States), In this study, we use a broad definition of retail atmospherics, which represents the retail environment. We use a unique data set covering a wide range of variables related to retail environmental variables both within and outside the store. The internal retail environmental variables include retailers' micro‑marketing strategies and the external retail environment that pertains to the neighborhood location and trade area demographics. The effect of the influence of retail environment on retail performance is studied using data for 646 grocery stores across the United States. A. C. Nielsen provided this Market Metrics data. We add to the past literature on retail atmospherics and retail performance (see Ghosh and Mc. Lafferty 119871; Wrigley 119901, and Rogers [19921 for excellent reviews) by showing how such information can be used by retailers to improve store performance.

Past research has shown that store performance' is influenced by variables such as trade area demographics including population and socioeconomic characteristics (Craig, Ghosh, and Mc. Lafferty, 1984), level of competition (Ghosh, 1984), retail atmospherics including location on intersection, sales area, credit card service, number of checkout counters, number of aisles Jain and Mahajan. 1979), and promotions (Walters and Rinne, 1986; Wa. Iters and Mac. Kenzie, 1988). In recent years, the grocery store industry has become increasingly competitive, and therefore additional variables, particularly those related to the effects of actions taken by the retailer, need to be studied. For example, retailers are looking at ways to differentiate themselves from competition by increasing the level of service at checkout and adding services, such as keeping the store open for 24 hours. They are using promotions such as double or triple couponing selling a variety nongrocery products (scrambled merchandising), and locating stores in smaller markets with an intent to improve store performance. To the best of our knowledge, no past study has researched the effect of internal retail environment including level of service, double couponing extended store hours. ' the extent of scrambled merchandising, and external retail envi ronment, such as store neighborhood characteristics on gro cery store performance.

An interesting question to address is whether the effect of retail environment on retail store performance is similar across all neighborhoods. Past research has studied the effects of different variables on store performance without attempting to segment stores to assess the differential impact of variables across stores, Researchers studying store choice have acknowl edged(Ghosh. 1984) and shown the effect of factors, such as income, occupation, and ethnicity on the relative importance given to variables by shoppers (Craig, Ghosh, and Mc. Lafferty, 1984). However, such modeling requires the collection of primary data. Our study deals with store level secondary data that is cross‑sectional in nature. As a result, segmenting stores by using this data offers useful insights into the determinants of grocery store performance. We add to the knowledge generated by past research by integrating a wide range of variables to model grocery store performance. Specifically, we develop two models explaining the effect of a store's internal and external environment on the variation in (1) a market‑based measure of performance‑sales, and (2) a productivity‑based measure of performance sales per square foot. Furthermore, we segment the stores on the basis of their socioeconomic characteristics and empirically demonstrate the differential impact of a store's internal and external environment on store performance across segments. The proposed models` are estimated and the consistency of results assessed using the Market Metrics data.

Our findings should provide useful insights into issues that retailers regularly encounter within a store, such as the impact of the number of checkout counters, the kind of merchandise sold, double couponing, keeping the store open for 24 hours, and neighborhood characteristics on sales and sales per square foot. in addition, it should help them understand whether the socioeconomic characteristics of the trade area influence the effect of a store's internal and external environment on sales and sales per square foot. These insights should enable managers to plan for desired levels of performance by using readily available secondary data. To summarize, this study contributes to the literature in several ways. This study focuses on empirically demonstrating the effect of a stores internal and external environment on store performance. It is, also, one of the first studies to empiri cally show that using the same retail environment strategies across all stores might not be an appropriate strategy. Strate giesshould in fact, depend upon the socioeconomic characteristics of the trade area. Most studies in the past have focused on forecasting market‑based performance measures, such as sales (Ingene and Lusch, 1980; Kumar and Leone, 1988) or market share (Durvasula, Sharma, and Andrews, 1992). Here, in a single study, we address two dimensions of perfor mance‑market‑based and productivity‑based performance. In other words, our study integrates a broad spectrum of variables in modeling the effect of store atmospherics on store performance.

In this study, we show that socioeconomic characteristics of the trade area moderate the effect of retail atmospherics on store performance. We demonstrate a methodology for predicting store performance (for existing and new stores) based on the internal and external environment of grocery stores by using aggregate secondary data. Specifically, the findings are used to offer guidelines to retailers for attaining desired levels of sales and sales per square foot by using readily available data. This research is organized as follows: first, past research is reviewed in light of the current study. Hypotheses are pre sented next. Then, data, methodology, and results are dis cussed. Implications of the results to grocery store managers are drawn. Finally, limitations of this study and suggestions for future research are discussed. Literature Review and Hypotheses We position this research in two ways. First, we discuss the literature on why retail atmospherics is critical for a store's performance. Next, we discuss our study in the light of past research on measuring store performance and then other rele vantstudies. In doing so, we provide a brief review of the literature on methods used for assessing/forecasting sales per formance in retail outlets. This is important since our frame work also enables retailers to generate an estimate of performance measures.

Retail Atmospherics are often designed to create a buying environ mentthat produces specific emotional effects that will enhance a consumer's likelihood of purchase. Both the functional attri butes in the store (e. g. , merchandise type) and the emotional attributes (e. g. , pleasantness) that a consumer elicits in his/ her mind determine a store's personality. Darden and Babin (1994) indicated that the emotional and functional aspects are strongly correlated with each other. Therefore, this study focuses on the retailing personality as measured by the functional aspects of a store's environment. In fact, Titus and Everett (1995) argued that the design of the shopping environ mentis an important element in the consumer retail search process. According to Ward, Bitner, and Barnes (1992), The physical environment that a retail store presents to potential customers can be divided into two parts: the external environment, that part of the store visible prior to entry into the retail sales or service area (parking lot, facade, entrance, etc. ), and the internal environment, that part visible from the retail selling space. Although no re search that we know of addresses the issue, it is reasonable to suppose that the relative importance of the external versus the internal environment in determining a store's categorization may differ across retailers.

Darley and Gilbert (1985) and Holahan (1. 986) suggested that the built environment has many significant influences on human psychology and behavior. Donovan and Rossiter (1982) and Gardner and Siornkos (1986) have suggested the importance of physical environ ment in retailing. Based on the literature on typicality and attitude (Loken and Ward, 1990), it is proposed that a con sumer's attitude be positively related to the internal and exter nal environment of the retail setting. Recently K‑Mart has created three different types of stores‑K‑Mart, Big K, and Super K. Similarly, Kroger has developed a new concept called the Signature store (in addition to the regular store). The differences across these types of stores being the assortment of merchandise sold as well as various services including in store bakery, banking facilities, open for 24 hours, etc. More quality services are offered in larger stores, which are located in many neighborhoods. In other words, the store's environment should be attractive enough (in case of grocery stores, it could be the advertising of the availability of banking services, travel agency. restaurant, open for 24 hours, offering of double or triple couponing, etc. ) for a consumer to enter. In order to complete a sale, the other environmental variables, such as assortment of merchandise and in‑store bakery. should be acceptable to the consumer. Thus, retail atmospherics play a critical role in consumer shopping behavior.

Store Performance An aggregate measure of the effect of consumer shopping behavior is store performance. Ghosh and Mc. Lafferty (1987), Wrigley (1990), and Rogers (1992) provide excellent reviews on forecasting retail performance. Table 1 summarizes the literature on forecasting retail performance along four dimen sions‑the method used, the type of' performance measure used (sales. market share), the type of explanatory variables used (for example, managerial judgements. distance, etc. ), and type of data collection required (individual level vs. aggregate). Our study uses linear models and requires aggregate level secondary data. It adds to the knowledge generated by past research by proposing a method forecasting weekly dollar sales (a measure of market‑based performance) and sales per square foot (a measure of productivity‑based performance). While earlier research has predominantly explained market based performance (sales, market share) by using variables including distance, size of the store, and trade area demo graphics. this study focuses on the effect of a store's internal and external environment on sales and sales per square foot. In addition, the moderating effects of the socioeconomic char acteristics of the trade area also are demonstrated.

Other related research on retail store choice has dealt with helping managers select neighborboods under different cir cumstances. These studies include selecting store locations by using managerial judgement (Durvasula. Sharma, and An drews, 1992), under a changing environment (Ghosh and Mc. Lafferty, 1982), selecting multiple store locations (Achabal, Gorr, and Mahajan, 1982), and a portfolio of stores (Mahajan, Sharma, and Srinivas. 1985). Like forecasting performance studies, these studies primarily focus on achieving desirable levels of sales or market share by using external environment characteristics. It is also important to compare this study with other rele vantstudies. Walters and Mackenzie (1988) and Walters and Rinne (1986) developed hypotheses on the effects of loss leaders, in‑store price specials, and double coupon promo tionson overall store sales, profits, and traffic. Thus, in addi tion to market‑based factors (sales and traffic), they also ex plained profitability based performance (profits). Their study primarily focused on studying the impact of promotional policies on store performance. Like Walters and Mackenzie's (Wal ters and Mackenzie, 1988) and Walters and Rinne's (Walters and Rinne, 1986) study, the dependent measure of interest is store performance. However, our study differs from theirs in scope. While they considered only promotional policies, we include a number of a store's internal and external environ mentvariables including promotional policy variables, such as double couponing. Also, we explain market‑based perfor mance (dollar sales) and productivity ‑based performance (sales per square foot).

Hoch et al. 's (Hoch, Byung‑Do, Montgomery, and Rossi, 1995) study used geodemographic data to predict store level price elasticities. Their study, too, was conducted with promo tional policy as the focus (the sales impact of a reduction in price due to retail price promotion). They use Becker's (Becker, 1965) theory of allocation of time to develop the hypotheses, which also forms theoretical basis for our hypotheses. Our study differs from Hoch et al. 's (Hoch, Byung‑Do, Montgom ery, and Rossi, 1995) study in focus. They focused on promo tionalpolicy, and their analysis was done at the product category level. We include a wide range of environmental variables, and our analysis is done at the store level. Further more, in addition to explaining dollar sales, we also explain sales per square foot. Ghosh (1984) focused on the nonstationarity of parameter estimates. He proposed that parameter nonstationarity arises due to two factors‑the socioeconomic characteristics of the shopper and the spatial characteristics of the alternative con sidered (competition) by the individuals. His study tested nonstationarity of parameters in a "gravity" model due to spatial characteristics of competition. Like Ghosh (1984), we propose that the effects of environmental variables vary de pending on socioeconomic characteristics of the trade area. While Ghosh's (Ghosh, 1984) study focused on testing of nonstationarity, our study focuses on explaining store perfor mance. Also, while his study was done at the individual level (involving primary data collection), our study uses aggregate secondary data (which is readily available).

Cottrell's (Cottrell 1973) and Davics's (Davics, 1973) stud ies also addressed store performance. However, they were carried out at a time when geodemographic data were difficult to obtain. Therefore, they are not discussed in detail. Thus, the present study builds up on these past studies and attempts to provide an integrated view for the retailer to improve their performance based on the store's internal and external envi ronment.