meets got/0L"; 5‘“ This is to certify that the thesis entitled THE VALIDATION OF A RISK ASSESSMENT INSTRUMENT FOR COMPETITIVE INTELLIGENCE presented by Sung Soo Chung has been accepted towards fulfillment of the requirements for M.S. degree in Criminal Justice Lea/024W Major professor Date May 1, 2000 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY l Michigan State University PLACE IN RETURN Box to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 cJCIRC/DateOuopss-p. 1 5 THE VALIoi’TION OF A RISK ASSESSMENT INSTRUMENT FOR COMPETITIVE INTELLIGENCE By Sung Soo Chung A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE School of Criminal Justice May, 2000 ABSTRACT THE VALIDATION OF A RISK MANAGEMENT INSTRUMENT FOR COMPETITIVE INTELLIGENCE By Sung Soo Chung The purposes of this thesis are to (1) theoretically integrate the concept of competitive intelligence (Cl) with the security management practice of risk assessment, and (2) to estimate the validity and reliability of a measure of ethical boundaries for CI gathering. In addition, this study tests the prediction of perceptions of ethics from several organizational variables, including tenure in the (a) Cl field, (b) job position, (c) organization, and organization size and region. The results revealed the validity and reliability of the Cl measure of ethical boundaries, and ethical perceptions were predicted from five variables: tenure (in the organization, the job position, and the CI field), organization size, and region. DEDICATION I dedicate this thesis to my grand mother, Duk-Rye Cha. ACKNOWLEDGEMENTS I thank my family for their continued support and encouragement, and I also thank Dr. Dae H. Chang and Dr. Yoon-Ho Lee for their help. I acknowledge my committee members, Dr. Vincent Hoffman and Dr. Mahesh K. Nalla, for their comments and recommendations. And I especially thank my advisor, Dr. Judith M. Collins, for her direction and guidance, and for the many hours we worked together on this thesis and on other projects. TABLE OF CONTENTS ABSTRACT ................................................................................................................... n DEDICATION ................................................................................................................ iii ACKNOWLEDGEMENTS ........................................................................................... i v TABLE OF CONTENTS .............................................................................................. v LIST OF TABLES ......................................................................................................... ix PURPOSE ..................................................................................................................... 1 INTRODUCTION ......................................................................................................... 3 THE PRACTICE OF COMPETITIVE INTELLIGENCE ......................................... 5 Definition of Competitive Intelligence .......................................................... 5 Theoretical Rationale ..................................................................................... 5 The Role of Competitive Intelligence in Security Management ................................................................................................... 9 Competitive |nte||igenoe Is Not Espionage ............................................... 1 0 Problems Involving Competitive Intelligence Gathering ------------------------- 10 The Importance of Competitive Intelligence as a Security Management Thesis Topic .................................................................................................. 1 2 RESEARCH OBJECTIVE ........................................................................................ 14 Research Hypotheses .................................................................................. 14 RESEARCH METHOD ............................................................................................. 17 Study Variables ............................................................................................. 1 7 Database ........................................................................................................ 17 Samples & Human Subjects ....................................................................... 19 Statistical Analyses ...................................................................................... 20 RESULTS .................................................................................................................. 21 Development of the Dependent Variables: The Competitive Information and Country Acceptance of CI Subscales ........................................................................................... 21 Factor Analysis of Competitive Intelligence Items ................................................................................................. 21 Descriptive Statistics for the Competitive Intelligence (Dependent Variable) Subscales --------------------------- 24 Means and Standard Deviations ................................................. 25 Coefficient Alphas .......................................................................... 27 Correlations Among the Competitive Intelligence (Dependent Variab|es) Subsoa|es .............................................. 28 Factor Analysis of Country Acceptance [terns ................................................................................................ 30 vi Descriptive Statistics for the Competitive Intelligence (Dependent Variable) Subscales ----------------------- 31 Means and Standard Deviations ............................................. 31 Coefficient Aiphas ...................................................................... 32 Correlations Among CI Training and Cooperation (Dependent Variables) Subscales ........................................... 32 Descriptive Analysis for Independent Variables ................................................................................................. 33 Frequencies ................................................................................ 33 Results of One-Way Analysis of Variance (ANOVA) .................................................................................................. 35 Tenure in the CI Field ................................................................ 35 Tenure in the Job Position ........................................................ 35 Tenure in the Organization ....................................................... 38 Organization Size ....................................................................... 39 Regions ........................................................................................ 41 Results of Simple Regression Analysis .............................................. 43 Results of Multiple Regression Analysis ............................................ 45 DISCUSSION ..................................................................................................... 47 Limitation ................................................................................... I .............. 53 CONCLUSION ................................................................................................... 5 5 vii REFERENCES .................................................................................................. 55 APPENDIX .......................................................................................................... 5 1 Appendix A .............................................................................................. 51 Appendix B .............................................................................................. 54 Appendix C .............................................................................................. 55 viii LIST OF TABLES Table 1. Factor Analysis Results of Competitive Intelligence Items ---------------------- 22 2. Means, Standard Deviations, and Coefficient Alphas for the Competitive Information (Dependent Variable) Subscales ............................................ 25 3. Correlations among the Competitive Intelligence (Dependent Variable) Subscales ........................................................................................................ 29 4. Factor Analysis of Items for Country Acceptance of CI --------------------------- 3O 5. Means, Standard Deviations, and Coefficient Alphas for the Country Acceptance (Dependent Variable) Subscales ......................................... 3 1 5, Frequencies for Independent Variables ..................................................... 34 7. Analysis of Variance by Tenure in the CI Field ......................................... 35 8. Analysis of Variance by Tenure in the Job Position -------------------------------- 37 9. Analysis of Variance by Tenure in the Organization -------------------------------- 38 10, Analysis of Variance by Organization Size ................................................ 4o 11_ Analysis of Variance by Regions ................................................................. 4 2 12. Results of Simple Regression Analysis ...................................................... 43 13. Multiple Regression Analysis Results for Subscales ------------------------------- 46 PURPOSE In this thesis, there are four interrelated goals involving risk assessment and the security of business information. The first goal is to develop a risk assessment instrument for measuring ethical and unethical activities involved in obtaining information on ones’ corporate competitors. Competitive information gathering is practiced by many organizations, but there are no ethical guidelines for this business activity. A risk assessment instrument for estimating the risk of violating ethical boundaries can be used to guide one's own employees in concluding business in an ethical manner and also as a measure for protecting one's own business information from being violated by others. As part of the first goal, factor analysis will be used to create ethical boundary subscales—also referred to as 'competitive information' or ‘competitive intelligence' subscales-and statistical tests will be performed to estimate the validity and reliability of the subscales. A risk assessment measure with established validity and reliability can be used in future research and for practical application. The second goal is to quantitatively examine how organizational variables might influence ethical or unethical information gathering activities. The demographic variables include tenure in the field of competitive intelligence; tenure in the job position; tenure in the organization; and organizational size. Also, the extent to which competitive intelligence activities are considered ethical or unethical may differ by country. Therefore, I will also estimate the influence of different regions in the world on competitive intelligence activities. A third goal is a cross-cultural comparison of employee perceptions of whether or not various competitive intelligence activities are either ethical or unethical. Using the scale or scales developed in part one, I will statistically test whether perceptions differ depending on the country in which the competitive intelligence activities are conducted. The fourth goal, which subsumes all of the above, is to introduce and incorporate into the security management literature the concept of “competitive intelligence.” Competitive intelligence activities are used to secure the safety of proprietary corporate information, but competitive information or intelligence is not part of the security management vocabulary. However, competitive intelligence activities are important for the protection of an important organizational asset: its' business information. Currently, competitive intelligence is considered a marketing function and the scant literature on this topic is found in the management sciences. However, competitive intelligence is a management function that involves risk assessment of assets, which is a fundamental practice in security management. Statistical theory and the theory of socialization will guide the conduct and hypotheses of this thesis. INTRODUCTION Competitive intelligence is information about a competing business, information that is often available in the public domain. However, across countries and cultures, the boundaries of public domain are not clearly defined. Violation of these ethical boundaries for obtaining information on ones' competitor can result in legal sanctions, loss of reputation and subsequent business and, in extreme cases, even death, e.g., recent media reports of business-related homicides in Russia. These kinds of risks to today's global organizations and to their employees can be reduced with knowledge of the acceptable boundaries of practice in other countries. Furthermore, awareness of organizational variables that predict certain Cl practices can be used by an organization to secure it's own competitive information. In other words, there exists a two-fold risk in practicing competitive intelligence: (1) ones' own organization may violate the ethical and cultural norms of another country, and (2) there is the potential for being violated by other organizations whose cultural ethics differ from one's own. An organization's CI employees could be accused of industrial espionage by unknowingly crossing the boundaries of acceptable Cl practices for some countries. At the same time, an organization may be susceptible to industrial spying by others who do not perceive their particular CI activity as being illegal. In this thesis, which is based on theories of socialization (discussed below), I will develop and test a CI activity risk assessment measure and examine several organizational variables that might help to explain and predict perceptions of the extent to which various CI activities are considered ethical across different cultures. THE PRACTICE OF COMPETITIVE INTELLIGENCE: LITERATURE REVIEW Definition of Competitive Intelligence Competitive Intelligence (CI) is the art and science of preparing companies for the future by way of a systematic process of awareness management (Calof & Skinner, 1999). Specifically, CI is the gathering of openly available information by using a process that involves planning, collection, analysis, communication, and the management of information available in the public domain (Calof & Skinner, 1999). Using this Competitive Intelligence process, data on other companies can be collected and then analyzed to help solve organizational problems and for use in organizational decision-making. There are three fundamental elements of the Cl process: (1 ) understanding and defining the intelligence problem, (2) interpreting the data that are gathered to generate judgments and conclusions, and (3) communicating the findings to decision makers (Sawka, 1999). Competitive intelligence professionals use public sources to find and develop information databases on competitors (Vella & McGonagle, 1987). Unlike business espionage, through which information is illegally obtained, competitive intelligence involves the use of 993m informationninformation that is legally and ethically available to the public. For instance, competitive information can be obtained from published annual reports of a company or through the Internet, using a coordinated and systematic approach (Stuart, 1996). Although often used interchangeably in the literature and therefore also in this thesis, competitive "intelligence" differs somewhat from competitive "information." Information is factual: numbers, statistics, and data. However, “intelligence” is a collection of pieces of information that have been sorted, distilled, and analyzed for use by managers in decision-making" (Kahaner, 1996: p.96). Therefore, competitive intelligence can formally be defined as a process that begins with the initial stage of defining problems through the final action stage (Kahaner, 1996), all based on information collected on one's competitor. Unfortunately, as many as 90% of all CI professionals sacrifice their integrity by violating ethical rules (Berger, 1998). Ironically, there is no reported literature outlining what is and what is not considered an ethical boundary to cross when collecting CI information. The reason may be that there is also no reported literature showing the development of any tests to explain or predict what is or is not perceived as ethical Cl activity. In this thesis, I attempt to extend the current and limited research on competitive intelligence by developing a risk assessment instrument and by using it to examine differences in CI activities across cultures. Theoretical Rationale Socialization theory (Schein, 1968) will be used to guide the hypotheses of this thesis. Socialization is the process by which each member of an organization or society learns and adapts to a norm, value system, and required behavior patterns (Schein, 1968). Members of an organization have their own subculture and customs in which organizational members, including newcomers, learn to be effective. In this thesis, socialization is discussed in the context of organizations and the security management of competitive information and activities for obtaining information. Socialization, the process through which individuals internalize the norms and values of an organization and a society, has guided the research in many domains. For example, socialization theory has been applied to the understanding of job attitudes of police officers (Van Maanen, 1975). Another study based on socialization theory examined the relationship between peer- group interactions and adult education in large industrial organization (Evan, 1963); and other studies on education and training for judges, prosecutors (WInfree, Kielich, & Clark, 1984) and policemen (Burgin, 1975) have been guided by socialization theory. These studies all have in common three factors underlying socialization theory: realism, role transition, and subcultures. Realism relates to how the organization lets members know what members should do and should not do in the organization and their specific required tasks (Feldman, 1981 ). A role transition is related to the learning of basic skills of a job, building relationships with co-workers and others, and learning the values and norms of relevant categories (Louis, 1990). Realism and role transition, therefore, are two processes, in which employees learn ethical or unethical work behaviors. In addition, individuals learn ethical and unethical behaviors in their subcultures. Those subcultures could refer to departments within an organization, units within departments, or even organizational sub-units located in other countries. In today's organizations, competitive intelligence gathering is a cross-cultural activity, because many companies are multinational. In these cases, socialization of ethical boundaries for competitive intelligence is a multifaceted process: different organizations have different cultures, and the standards of operation for one organization may differ in different countries, due to differences in cultural practices in those countries. I Furthermore, realism, role transition, and the understanding of subcultural differences all depend on organizational socialization, whether formal through training or informal through word-of-mouth. Similarly, organizational variables such as size of the company, length of time an individual is employed with the company and in a specific job, can all influence the perceptions of how certain job-related activities should be performed. In many organizations today, competitive intelligence gathering is an ongoing job activity. As will be discussed further below, knowledge of the influence of these and other organization factors on employee CI behavior can help to understand and perhaps explain risk factors involved in CI activities. In sum, socialization theory can guide this thesis by pointing to variables for analyses and by helping to explain why perceptions of ethical Cl activities may differ across countries. The Role of Competitive Intelligence in Security Management Security management is an organizational function that focuses on protecting and safeguarding an organization's resources. One of an organization's most important resources is its' competitive information which is the basis for business profits, losses, and existence. However, an organization’s information can be at risk for theft from competitors who may use unethical methods to obtain proprietary information. Furthermore, an organization itself may be at risk if its' own employees cross ethical boundaries in obtaining information on their competitors. For organizations with subunits in other countries and other cultures, the risks can be great: information on ones' company can be lost; business can be lost if employees violate the ethics of other countries; and employees could even personally be harmed if they violate a country’s ethical customs. Fortunately, risks involved in intelligence gathering practices can be reduced with knowledge of acceptable boundaries. An important role of the criminal justice system is to protect intangible property (Burrows, 1997), such as proprietary information. A risk assessment instrument for measuring ethical boundaries of CI activities can therefore be an important tool from a criminal justice perspective. Risk assessment of competitive intelligence practices can be expected to be very important for the next generation of security managers, because in recent years competitive information gathering has been increasing (Lewandowski, 1999) and so has information espionage (CERT Coordination Center, 1997). Competitive Intelligence Is Not Espionage Professionally conducted competitive intelligence uses public sources to locate information on competitors and on the market environment—information that can be legally and ethically identified and obtained is acceptable. In contrast, business information that is illegally obtained is called business "espionage" (Malhota, 1996), which is on the increase. For example, it has recently been 1' revealed that business firms in the US spent $2 billion dollars each month for *- spying on their competitors in attempts to gain competitive intelligence (Jones, 1999) f At the same time, competitive intelligence has become a professional practice in which competitive information is obtained using legal, but not necessarily ethical, methods. The Society of Competitive Intelligence Professionals is a relatively new organization with nearly 5,000 members in 43 different countries. Although the laws in those countries guide the intelligence gathering practices or activities, there are as yet no ethical standards across countries because there are cultural differences as to what is acceptable. Organizations that have Cl functions consider themselves professional and not criminal. Problems Involving Competitive Intelligence Gathering There are three primary problems involving the knowledge of ethical boundaries for CI gathering. First, there is no reported literature on the ethical boundaries of competitive intelligence gathering, but the importance of 10 competitive intelligence is great. Even though competitive intelligence involves the securing of corporate assets and information fraud, competitive intelligence is not discussed in any security management textbook or journal or even in the criminal justice literature. The lack of recognition by these disciplines is likely due to the fact that competitive intelligence is a relatively recent management practice that evolved out of the marketing research literature and that now is proliferating along with the global expansions of organizations. The second problem, referred to previously above, is that each country has a different culture with different moral concepts for appropriate competitive intelligence practices. For example, in some countries bribery is an acceptable practice to obtain information on another company, but in the US. bribery is a violation of statutory law. These cross-cultural differences may explain why attempts have not yet been made to develop a standardized risk assessment instrument for measuring what may be viewed as ethical (or unethical) Cl practices. A third fundamental problem is that even professional CI employees may not know the limitations or where the acceptable boundaries are drawn for collecting Cl information (Berger, 1998). These three fundamental Cl problems-- lack of literature, cross-cultural differences, and lack of defining boundaries-are all addressed in this thesis. 11 The Importance of Competitive Intelligence as a Security Management Thesis Topic Although 'professional' competitive intelligence, as it is known today, is a relatively recent phenomenon, the concept of "Competitive intelligence has been around almost as long as capitalism itself' (Stuart, 1996, p.1). In fact, the history of competitive intelligence can be found from espionage activities, which are as old as human beings (Sarbin, Carney, and Eoyang, 1994). Espionage activities originated from treason in the ancient times, and treason started in early Roman times with the enactment of the Great Statute of Treason in 1352, which defined treason as a crime (Sarbin et al., 1994). In the time of the Kings Henry I, II, and III of England, acts of treason were used for controlling enemies politically. Later, the English concept of treason was moved into Americauand this was the basis of the concept of espionage activities. After the World War II, spying by security agencies in each country was developed with various new devices to obtain intelligence on their competitive countries, such as laser eavesdropping to hear conversations through walls (Inbau, Farber, & Arnold, 1996). In the past decades, gathering intelligence often involved lots of legwork and reams of paper (Stuart, 1996). After the fall of communism and by using the Internet, espionage activities grew. Thus, the practice of competitive intelligence evolved out of illegal activities; however, U.S. organizations do not want to violate laws. Therefore, the field of competitive intelligence has become known as a practice whereby Cl professionals collect information ethically and legally. Organizations use 12 competitive information to make decisions about their own products or services, and competitive information is also used to prevent others from obtaining their own proprietary information. Asset protection management is the primary activity in the security management discipline; investigations involving competitive intelligence are therefore an important research topic in the field of security management. 13 SPECIFIC RESEARCH OBJECTIVES Where are the ethical boundaries for CI gathering in different countries? Currently, many business people do not know the limitations for the ethical boundaries for collecting competitive intelligence (Berger, 1998). Therefore, this thesis proposes to (a) develop and (b) test for validity and reliability a measure (scale or scales) of competitive intelligence gathering using data collected from CI professionals in 30 countries around the world. In addition, using organizational variables, mathematical models will be developed to predict intelligence practices, and cross-cultural comparisons will be made of intelligence gathering activities. The overarching goal is to provide security managers with a new tool for competitive intelligence risk assessment in the 21st century. Research Hypotheses There is no reported literature on the ethical boundaries for competitive intelligence practices, and there are no reported risk assessment measures to determine ethical versus unethical boundaries in different countries. Therefore, this research uses socialization theory to guide the following informal and exploratory hypotheses. Due to cross-cultural differences, I hypothesize that competitive intelligence practices differ for different countries. Also, I hypothesize that tenure in the CI field, the job position, and the organization can explain individual and cultural differences in perceptions of what is considered an ethical Cl practice, 14 i.e., activity. Tenure is hypothesized to be important because, according to socialization theory, employees who have a long time in an organization and in a job position would have more time to be socialized to whatever ethical boundaries are subscribed to by an organization. Further, organization size may be another important variable that influences employees’ learning of ethical Cl boundaries. According to socialization theory, organization size is positively related to role transition. For example, larger organizations may be more likely than smaller companies to have formalized training programs whereby newcomers become formally socialized to norms, value systems, and required behaviors. If training involves socialization to the organization, behaviors are likely to be reinforced through reward and punishment systems. Employees of larger organizations may therefore be exposed to formal training and become socialized more so than employees of smaller companies where there are other priorities or resource limitations. For all of the above reasons, I hypothesize that perceptions of what is and what is not considered an ethical Cl activity can be explained by and predicted from employees' tenure in the Cl field and in their jobs, and from organizational size and by differences across cultures. A formal prediction model is, 15 Y = a + (bi)X1 + (b2)X2 + (b3)X3 + (b4)X4 + (b5)X5 + e, where Y = Perceptions of ethical behavior, and acceptance by a country of a CI activity. X1 = Tenure in the CI field X2 = Tenure in the job position X3 = Tenure in the organization X4 = Organizational size X5 = Country a = Constant (intersect) bis = Beta Weights 6 = error. 16 RESEARCH METHOD Study Variables The dependent variables are measures of (a) employee perceptions of ethics for a Cl activity, and (b) the extent to which a Cl activity is considered acceptable in the employee's country of operation. These variables are described further below. There are five independent variables: tenure in the Cl field, tenure in the job position, tenure in the organization, organizational size, operationalized as the number of employees in an organization, and country. Database Collins & Chung (1999) developed and administered to expatriates, i.e., US. citizens employed by their organizations in other countries, items that measured perceptions of ethical intelligence gathering activities. The items were generated from a review of the literature on business ethics and competitive information, and the items were written to reflect past behavior and experiences (Mumford & Owens, 1987). These kinds of items are also known as biographical data, or simply biodata. The item values were ordered on a Likert scale ranging from 1 to 5, where 1 = strongly unethical; 2 = unethical; 3 = questionable; 4 = ethical; 5 = strongly ethical. A very low score therefore indicates clearly unethical behavior and a very high score suggests a clearly ethical behavior, for any given CI activity. 17 Item examples are, how ethical is it to...“survey a competitor’s employees on their satisfaction with that company"; . . . “fix prices with another company on a similarly manufactured product”; and...“conduct aerial surveillance of competitor.” Other items measure perceptions of a country's acceptance of Cl activities. These items were intended to capture the cross-cultural differences. The responses for these items, also ordered along a Likert scale, were 1 = strongly disagree; 2 = somewhat disagree; 3 = do not know; 4 = most likely agree; and 5 = strongly agree. Examples of these items are as follows: in the country in which you are employed, to what extent do “...most companies have CI training”; “...share competitive information with other CI professionals”; and. “...are targets of CI gathering.” The entire inventory of items, which altogether represent the dependent variables in this thesis, is presented in Appendix A. There are five independent variables. Three independent variables-- tenure in the Cl field, tenure in the job position, and tenure in the—were measured using the following categorical scale: less than two years; three-to-ten years; and more than 10 years. The fourth independent variable, organizational size, had five categories: fewer than 25 employees; 26 to 100 employees; 101 to 500 employees; 501 to 1000 employees; and more than 1001 employees. For the variable, 'country,' there were too few subjects in the database and therefore too few corresponding responses to conduct the statistical analysis. Therefore, I conducted the analysis by the following ten regions: North America, Central America, South America, Eastern Asia, Southern Asia, Western 18 Asia, Africa, Australia, Eastern Europe, and Western Europe. Appendix C lists each country and georgaphic region. Samples & Human Subjects The data (N = 200) were collected by Judith M. Collins from member of the Society of Competitive Intelligence Professionals (SCIP) who were located in 43 countries. The data are anonymous: there are no names or identifying numbers associated with the data. For collecting these data, human subjects approval was obtained by Judith M. Collins, March 26, 1999, IRB #99145, Category 1-C; Title, “Competitive Intelligence: Ethical and Legal Boundaries for the Strategic Management of Corporate Security and Success. For the present thesis, I divided the data into ten regions. North America included the US. and Canada. Mexico was included in the Central America region because the culture in Mexico is more similar to countries in Central America than in North America. Brazil is a part of South America. Eastern Asia included China, Japan, and Taiwan. Southern Asia included Indonesia, Malaysia, India, Singapore, and the Philippines; Western Asia included Israel and Saudi Arabia; Africa included South Africa; the continent of Australia included Australia and New Zealand; Eastern Europe included Croatia, Poland, Russia, Slovenia, Turkey, and Ukraine; Western Europe included Germany, Italy, Spain, Sweden, Switzerland, The Netherlands, and the United Kingdom (Appendix B). 19 Statistical Analyses The following statistical analyses were conducted: (1) principal factor analysis of item responses that measure (a) perceptions of Cl activities, and (b) acceptance of a CI activity, by region which are used as the dependent variables; (2) descriptive statistics for the dependent variables; (3) descriptive statistics for the independent variables; (4) one-way analysis of variance (ANOVA), to estimate percent variance accounted for in the dependent variable by the respective independent variables; (5) simple regressions to test the study hypotheses, and (6) multiple regression analyses to estimate the prediction of Cl activities from the organizational variables and from region. 20 RESULTS Development of the Dependent Variables: The Competitive Information Subscales and Country Acceptance of CI The CI inventory was composed of two sections: part I, in which the items measure ghi_cal persceptions, hereafter referred to as 'competitive intelligence boundaries'; and part II, in which the items measure a countDIfs acceptance of various competitive intelligence gathering activities. here_after referred to Q 'countDI acceptance.’ Factor analyses were performed on the item responses for each of the two parts. Principal component analysis with varimax rotation was used, because this procedure is useful for evaluating simple structures (Mumford 8. Owens, 1987). factor Ana_lvsis of Competitive Intelligence Items Table 1 and Appendix C report the results of the factor analysis. Of 55 items measuring perceptions competitive intelligence activities, the factor analysis identified fifteen subscales. However, three subscales did not meet the eigenvalue criterion of 1.0 or greater (Nunnally, 1972). Those subscales were therefore not included in subsequent analyses, leaving twelve remaining subscales. 21 Table 1. actor Analysis Refllts of Competitive Intelligence Items Item Factors Item Factors Numb“ 1 2 3 4 5 6 Numb“ 7 8 9 10 11 12 42 .82 41 .75 43 .65 9 .64 40 .82 25 .67 44 .66 45 .56 50 .45 55 .64 1 .88 11 .73 52 .69 22 .53 53 .90 4 .61 6 .79 12 .60 19 .66 17 .57 37 .79 23 .74 21 .71 29 .66 16 .52 26 .59 51 .72 15 .86 46 .75 2 .62 47 .63 14 .61 34 .67 46 .64 3 .75 26 .79 54 .65 27 .74 36 .40 7 .61 5 .80 10 .70 24 .57 13 .53 6 .56 32 .55 Note: 1 = Public Domain; 2 = Deceit; 3 = Soliciting; 4 = Price-fixing; 5 = Illicit; 6 = Accidental; 7 = Misleading; 8 = Surveying; 9 = Underhanded; 10 = Observing; 11 = Paid Cl; 12 = Miscellaneous. 22 In Table 1 items for the first factor pertain to obtaining information from the public domain; hereafter, these items are referred to collectively as simply 'Public Domain.‘ Factor two items refer to obtaining information in deceitful ways, such as bribing a competitor’s supplier or employee, or planting an agent on the competitor's payroll, hereafter referred to as 'Deceit.‘ Factor three items refer to soliciting competitors’ information from common suppliers or from salespeople, about competitors; hereafter called 'Soliciting.‘ Items for factor four are related to price-fixing, hereafter called 'Price- Fixing.‘ For factor five, two items pertain specifically to taping private conversations of competitors and one item pertains to hiring an executive search firm to interview a competitor's executive with no intention of actually hiring. Hereafter, factor five is called simply 'Illicit.‘ The items in factor six reflect the use of competitor information that is accidentally obtained, such as from a misdirected fax or a trade secret that was accidentally obtained, or by obtaining inside information on a competitor in other ways. Hereafter, these items are referred to collectively as 'Accidental.‘ Factor seven items all reflect obtaining information by misleading the competitor, such as "altering figures on a financial report that goes to the competitor" (item # 9). Factor seven is hereafter called 'Misleading.‘ Factor eight is called 'Surveying,’ because these items refer to surveying a competitor's employees on their organizational satisfaction or commitment. 23 Factor nine measures underhanded interactions with a competitor such as paying its own employee to hire into a competitors’ company, or agreeing with a competitor to bid-fix, but then underbidding the competitor. Factor nine therefore is called 'Underhanded.‘ Items for factor ten refer to obtaining information by observing competitors property or through common customers; factor ten is called 'Observing.’ Factor eleven measures the obtaining of information by paying people, such as by paying a private investigator or paying a refuse company for information; this item is called 'Paid Cl.’ Last, factor twelve refers to aerial surveillance, eavesdropping, and using handwriting samples of competitor managers, publicly obtained. These items are not conceptually similar, so this scale is called 'Miscellaneous.’ Descriptive Statistics for the Competitive Intelligence (Degndent Variable) Subscales Table 2 presents the means, standard deviations, and coefficient alpha estimates of reliabilities for the twelve subscales for all respondents across all 30 countries. In a following section, descriptive statistics will be presented by region. 24 Table 2. Means Standard Deviations and Coefficient Alphas for the Twelve Competitive Information (Dependent Variable) Subscales Dependent Variable Subscale M SD Aim 1. Public Domain (6 items) 23.17 2.67 .83 2. Deceit (4 items) 4.78 1.88 .74 3. Soliciting (4 items) 11.91 3.64 .74 4. Price-Fixing (4 items) 10.25 3.28 .44 5. Illicit (3 items) 4.35 2.17 .63 6. Accidental (6 items) 12.09 4.13 .75 7. Misleading (4 items) 11.64 3.43 .68 8. Surveying (3 items) 5.83 2.20 .78 9. Underhanded (3 items) 4.63 1.97 .71 10. Observing (3 items) 10.03 1.60 .06 11. Paid Cl (3 items) 7.83 2.62 .64 12. Miscellaneous (3 items) 8.45 2.03 .41 Note: N = 66. Means and Standard Deviations. Mean scores on the ethical items range from 4.35 to 23.17. On Table 3, most respondents scored high on Public domain gathering (M = 23.17; E = 2.67). Since this is a six-item, the mean range is from 6.00 (low end) to 30.00 (high end), and the mean of the continuum is 18.00. A mean score of 23.17, therefore, indicates that obtaining competitive information in the public domain is perceived as ethical. 25 Scores were lower on Depej (M = 4.78; E = 1.88). The range for this four items scale 4.00 (low) to 20.00 (high) and 12.00 (mean of range), suggesting that most respondents considered bribery and other forms of deceit as “clearly unacceptable.” For Soliciting, the mean score was 11.91 (SD = 3.64), indicating that soliciting information from competitor's suppliers and other was perceived as a ‘questionable’ activity (four items with a range of 4.00 to 20.00). The mean of this continuum is 12.00; therefore, the respondents did not consider soliciting as either clearly unethical or ethical. The Price-Fixing responses (range = 4.00-20.00) indicate that this activity is considered "somewhat unethical” (M = 10.25; _S_D = 3.28). The low mean score (range = 3.00 to 15.00) for the m subscale (M = 4.35; _S_D = 2.17), indicates that most respondents view obtaining competitive intelligence illicitly as “unethical behavior.” For the Accidental items (range = 6.00 to 30.00 with mean = 18), respondents viewed information obtained accidentally as ”somewhat unethical” behavior (M = 12.09; _S_D = 4.13). Similarly, Misleading competitors by giving them wrong information is considered "questionable" (M = 11.64; SD = 3.43); again, this score reflects neither ethical or unethical behavior—at the low end of the continuum for four items, the score would be 4.00, and at the high end the score would be 20.00. 26 The mean score for Surveying (M = 5.83; S_D = 2.20) also suggests “somewhat unethical” perceptions: respondents scored in the middle on this three-item subscale (range = 3.00 to 15.00). The Underhanded subscale refers to obtaining information by spying and by breaking agreements. Respondents considered these activities as “unethical" (M = 4.63; g = 1.97; Range = 3.00 to 15.00; mean of range = 9.00). For Observing, M = 10.03 (SD = 1.60); since the low score would be three and the high score would be fifteen, the mean score for Observing suggests only “questionable”-not necessarily either ethical or unethical. The mean score for the Paid Cl (M =7.83; _S_D = 2.62)-paying others for information on a competitor—is considered “unethical.” The range for the three items was 3.00 to 15.00. For Miscellaneou_s, M = 8.45 (SD = 2.03), the range was 3.00 to 15.00. again suggesting these activities are ”questionable” but not necessarily either “ethical” or “unethical.” Coefficient AIMS Table 2 also shows the coefficient alphas for the subscales: Public domain, .83; Deceit, .74; Soliciting, .74; Price-Fixing, .44; Illicit, .63; Accidental, .75; Misleading, .68; Surveying, .78; Underhanded, .71; Observing, .06; Paid Cl, .64; and Miscellaneous, .41. Of the above 12 subscales, the alpha estimates of reliability for three subscales— Price-Fixing; Observing; and Miscellaneous did not meet the acceptable criterion of .60 or greater (Nunnally, 1972). These three scales were therefore dropped from further analysis. 27 The nine remaining subscales, i.e., the dependent variables, now rearranged in alphabetic order, are as follow: 1 = Accidental; 2 = Deceit; 3 = Illicit; 4 = Misleading; 5 = Paid Cl; 6 = Public Domain; 7 = Soliciting; 8 = Surveying; and 9 = Underhanded. For these nine subscales, the coefficient alphas ranged from .63 to .83. Correlations Among the Competitive Intelligence (Dependent Variable) Subscales. Table 3 presents the correlations among the nine subscales. Overall, the correlations ranged from .00 (Accidental and Public Domain) to .52 (Accidental and Soliciting). The correlations were low (.00 to .24) between Public Domain and the remaining eight subscales. These low correlations reveal the divergent validity of the subscales in the direction that would be expected, i.e., low correlations for scales that measure opposite activities, i.e., Public Domain items were scored in the 'ethical' direction and the remaining items were scored in the 'unethical' direction. For the subscales two through nine on Table 3, the correlations ranged from .14 (Illicit and Deceit) to .52 (Accidental and Soliciting). Except for Illicit and Deceit, the correlations ranged from moderately low (.20 - .40) to moderate (around .50). These values are acceptable evidence for convergent validity, which would be expected for scale items that ask respondents for views on unethical activities. The .14 correlation between Illicit and Deceit indicates a lesser degree of convergence. 28 Table 3. Correlations among the CompetitiLe Intelligence (Dependent Variable) Subscales Independent Comgtitive Intelligence Boundaries Variables 1 2 3 4 5 6 7 8 9 1 .Accidental -- .33“ .38” .38“ .44“ .00 .52“ .37“ .38“ 2.Deceit -- .14 .28 .40“ .10 .27“ .26* .23 3.Illicit -- .40“ .37“ .21 .36" 25" .26“ 4.Misleading -- .30" .24 .50“ .20 .30" 5.Paid Cl -- .04 .36“ .30" .29* 6.Pub. Domain - .22 .22 .03 7.Soliciting -- .43“ .27* 8.Surveying -- .38" 9.Underhanded -- Note: 1 = Accidental; 2 = Deceit; 3 = Illicit; 4 = Misleading; 5 = Paid CI; 6 = Public Domain; 7 = Soliciting; 8 = Surveying; 9 = Underhanded. *p <.05, **p <.01; The specific correlations were as follows: Accidental and Deceit ([ = .33); Accidental and Illicit (r = .38); Accidental and Misleading ([ = .38); Accidental and Paid Cl (1 = .44); Accidental and Soliciting ([ = .52); Accidental and Underhanded (r = .38); Deceit and Paid Cl (r = . 40); Deceit and Soliciting (r = .27); Deceit and Surveying (r = .26); Illicit and Misleading ([ = .40); Illicit and Paid Cl (5 = .37); Illicit and Surveying (r = .25); Illicit and Underhanded ([ = .26); Misleading and Paid Cl ([ = .30); Misleading and Soliciting (r = .50); Misleading and Underhanded 29 (g = .30); Paid Cl and Underhanded (r = .29); Soliciting and Surveying (_r_ = .43); Soliciting and Underhanded (I = .27); Surveying and Underhanded ([ = .38). F_actor Ana_lvsis of 'Count_r_’y Acceptance of CI’ Items Table 4 and Appendix C report the results of the factor analysis on the 15 items in Part II in which the items are related to “country acceptance” of various Cl activities. The factor analysis revealed three factors. Table 4. Factor Analysis of Items for County Acceptance of Cl Item Factors Number 1 2 3 64 .82 63 .80 56 .77 62 .50 65 .65 66 .53 59 .78 61 .70 57 .63 68 .80 58 .64 69 .75 Note: 1 = Cl training; 2 = Cooperation; 3 = Miscellaneous. Factor one in Table 4 contains items related to CI training and education; hereafter, this factor is called as ‘Cl Training.’ Items in factor two refer to a cooperative working environment in which competitors share information; 30 hereafter, called ‘Cooperation.’ Last, factor three refers to targets of CI activities and also government support for CI activities. Conceptually, these items are dissimilar, so this scale is named ‘Miscellaneous.’ Descriptive Statistics for the County Acceptance (Dependent Variable) Subscales Table 5 presents the means, standard deviations, and coefficient alpha estimates of reliability for the three Country Acceptance factors. Table 5. Means Standard Devia_tions. and Coefficient Alphas for the Cogntrv Acceptance (Dependent Variable) Subscales Subscale M S_D Alpha 1. CI training (6 items) 11.36 4.14 .78 2. Cooperation (3 items) 6.41 2.45 .63 3. Miscellaneous (3 items) 5.94 1.90 .46 Note: M = 66. Means Standard Deviations. The possible values for the range of means in Table 5 are from 6.00 (low end) to 30.00 (high end) for CI training. The mean for CI Training, which refers to the extent that organizations in different countries promote Cl training and education, is 11.36 (_S_D = 4.14)-a value considerably 31 lower than the mean. This suggests that, across countries, respondents perceive that organizational training in CI is of less than average value. For Cooperation, the range of possible scores is from 3.00 to 15.00. Cooperation items refer to cooperation with other Cl professionals, such as sharing competitive information with one another. The mean score for Cooperation is M = 6.41 (SD = 2.45), indicating that most respondents do not cooperate very much with one another in CI activities. The range of items for Miscellaneous is also from 3.00 to 15.00. The Miscellaneous scale is composed of a number of items that do not seem to relate to a single construct. Nonetheless, empirically (versus rationally) developed items can predict future behavior or events (Nunnally, 1972). The mean score for the Miscellaneous subscale was M = 5.94 (_S_D = 1.90), a value again considerably less than average. However, the coefficient alpha for this scale was very low, prohibiting further use of this scale for further analyses, according to the a priori established criteria. Coefficient Alphas. The coefficient alpha estimates of reliability for the above scales were as follows: .78 (CI Training); .63 (Cooperation); and .43 (Miscellaneous). Only the Cl Training and the Cooperation (dependent variables) subscales will be used in the subsequent analysis. Correlations Among CI Training and Cooperation (Dependent Variable) Subscales. The correlation between the Cl Training and Cooperation is .44. This moderate correlation can be expected for variables that relate to one another; a 32 lo_w correlation would mean little or no relationship between these variables, and a Mgr; correlation would imply construct sameness. To sum thus far, there are nine dependent variables that each pertains to perceptions involving the ethics of gathering or collection of information from competitors. In addition, there are two other dependent variables that pertain to the acceptance by country of competitive intelligence activities. Descriptive Analysis for the Independent Variables Freguencies Table 6 presents the frequencies for the independent variables. For tenyg in the Cl field, the number of subjects with ‘Iess than two years’ was 21 (31.8%), ‘three to ten years,’ 38 (57.6%), and ‘more than ten years,’ 7 (10.6%). For tenure in the CI field, the sampling distribution was positively skewed. The possible reason for the skewed distribution might be that competitive intelligence is a relatively new field or specialty. For tenure in the job position, the number of subjects with ‘less than two years’ was 32 (48.5%), ‘three to ten years,’ 32 (48.5%), and “more than ten years,’ 2 (3.0%). For tenure in the job position, the sampling distribution was also positively skewed, again perhaps because Cl might be a relatively new field. For tenure in the organization, the number of subjects with ‘less than two years’ was 17 (25.8%); ‘three to ten years,’ 30 (45.5%); and ‘more than ten years,’ 19 (28.8%). For tenure in the organization, the sampling distribution was approximately normal. 33 Table 6 Fr uencies for Inde endent Variables Independent Variables N (%) Tengre in CI Field 66 (100.0) less than two years 21 (31.8) three to ten years 38 (57.6) more than ten years 7 (10.6) Tenure in the Job Position 66 (100.0) less than two years 32 (48.5) three to ten years 32 (48.5) more than ten years 2 (3.0) Tenure in the Organization 66 (100.0) less than two years 17 (25.8) three to ten years 30 (45.5) more than ten years 19 (28.8) Qmanigatiop Size 66 (100.0) Less than 26 people 14 (21.2) 26 to 100 people 5 (7.6) 101 to 500 people 12 (18.2) 501 to 1000 people 4 (6.1) more than 1000 31 (47.0) Region 66 (100.0) North America (NA) 19 (28.8) Central America (CA) 2 (3.0) South America (SA) 1 (1.5) Eastern Asia (EA) 9 (13.6) Southern Asia (SA) 9 (13.6) Western Asia (WA) 3 (4.5) Afn'ca (AFR) 4 (6.1) Australia (AUS) 3 (4.5) Eastern Europe (EE) 4 (6.1) Western Europe (WE) 10 (15.2) Not Available 2 (3.0) Note: I! = Number of respondents in a category; % = Percent of respondents in a category. 34 For organization size, the number of subjects with ‘Iess than 26 employees’ was 14 (21.2%); ’26 to 100,’ 5 (7.6%); ‘101 to 500,’ 12 (18.2%); ‘501 to 1000,’ 4 (6.1%); and ‘more than 1001 employees,‘ 31 (47.0%). The distribution of organization size was negatively skewed—most respondents working in the Cl field belong to a large company. The Results of the One-Way Analysis of Variance (ANOVA) The Statistic Analysis System (SAS; SAS Institute, Inc, 1995) software was used to analyze the data. Tables 7 through 11 describe the results. In each of these tables, the independent variables are those from Table 6 above; they are also listed in the left side column of each Table 7 through 11. On each table, the dependent variable is listed at the top, beginning with ‘tenure in the field.’ Tenure in the Cl Field Table 7 presents the effect sizes (12) for ‘tenure in the CI field,’ for each of the Cl subscales. The ANOVA results are across all countries. Except for CI training (E = 2.83, p < .10), there were no significant effects for any measures. And for CI training, the variance explained (r2 = .08) in ‘tenure in the Cl field’ was negligible. That is, other factors besides Cl Training explain 92% of the variance in length of time an individual is employed in the CI field. In other words, there were minimal differences in number of years of employment in the Cl field with respect to Cl training. The sample means displayed in Table 7 show similar scores for the three groups. 35 Table 7 Anal sis of Variance b Tenure in the I Field Tenure in the CI Field Independent leg than two Three to ten More than ten Variables years years years E r2 Cl Subscales Accidental 12.0 (4.32) 12.3 (4.23) 11.2 (3.19) 0.19 .01 Deceit 4.65 (1.35) 4.84 (2.12) 4.83 (1.17) 0.07 .00 Illicit 4.30 (2.36) 4.26 (2.01) 5.17 (2.17) 0.43 .01 Misleading 11.5 (3.19) 11.7 (3.46) 11.7 (4.55) 0.02 .00 Paid CI 7.75 (2.73) 7.89 (2.34) 8.33 (4.18) 0.12 .00 Public domain 27.9 (3.65) 28.2 (2.12) 25.7 (5.68) 1.65 .05 Soliciting 11.9 (3.42) 11.6 (3.56) 13.5 (4.97) 0.64 .02 Surveying 5.70 (2.43) 5.79 (2.07) 6.50 (2.51) 0.42 .01 Underhanded 4.60 (1.88) 4.53 (1.96) 5.33 (5.58) " 0.43 .01 Country Acceptance Cl Training 10.7 (3.69) 11.1 (4.23) 15.0 (3.74) 2.82 .08" Cooperation 6.20 (2.63) 6.32 (2.41) 7.67 (2.07) 0.89 .03 Note: *p < .10; n = 66. Tenure in the Job Position Table 8 presents the effect sizes (r2) for 'tenure in the job position,‘ for each of the Cl subscales across countries. 36 Table 8 Analysis of Variance by Tenure in the Job Position Tenure in the Job Position Independent less than two Three to ten more than ten Variables yea_r_s years ye_a§ E r2 CISubscale Accidental 11.4 (3.93) 12.6 (4.22) 14.0 (7.07) 0.76 .02 Deceit 4.94 (1.03) 4.94 (2.31) 7.00 (4.24) 1.94 .06 Illicit 3.68 (2.07) 5.13 (2.08) 3.00 (1 .41) 4.96 .14“ Misleading 11.6 (3.08) 12.0 (3.66) 7.00(2.83) 2.04 .06 Paid Cl 7.68 (2.59) 8.13 (2.62) 5.50 (3.54) 1.05 .03 Public domain 28.0 (3.24) 27.9 (2.98) 24.5 (4.95) 1.16 .04 Soliciting 11.8 (4.09) 12.1 (3.30) 11.0 (1.41) 0.12 .00 Surveying 6.10 (2.10) 5.55 (2.31) 6.00 (2.83) 0.62 .02 Underhanded 4.35 (1.56) 4.80 (2.21) 6.00 (4.24) 0.91 .03 Country Acceptance Cl Training 11.7 (4.64) 10.9 (3.72) 13.5 (0.71) 0.34 .01 Cooperation 6.39 (2.64) 6.19(2.17) 10.0 (0.00) 2.37 .07 Note: ”*p< .01; n=66. Except for Illicit conversations (E = 4.96, p < .01), there were no significant effects. Illicit conversations explained 14% (r2 = .14) of the variance in ‘tenure in the job position.’ That is, 86% of the variance in length of time an individual is employed in the job position remains to be explained. According to the means (Table 8) and the significant r2, one interpretation could be that the longer 37 employees are in the job position, the more likely they consider Illicit conversations to be methical. Tenure in the Organization Table 9 presents the means, standard deviations, and effect sizes (r2) for each subscale for tenure in the organization across countries. Table 9 Analysis of Variance bv Tengre in the Organization Tenure in Organization Independent Less than two Three to ten more than ten Variables years years yeag E R2 CISubscale Accidental 12.1 (4.00) 12.2 (4.59) 11.8 (3.67) 0.05 .00 Deceit 4.44 (0.81) 4.69 (1.51) 5.21 (2.82) 0.79 .03 Illicit 3.63 (1.96) 4.83 (2.39) 4.26 (1 .88) 1.12 .04 Misleading 11.1 (3.18) 12.1 (3.15) 11.4 (3.59) 0.44 .01 Paid CI 8.00 (2.58) 7.55 (2.54) 8.11 (2.87) 0.29 .01 Publicdomain 27.1 (4.05) 28.0 (3.26) 28.2 (2.02) 0.57 .02 Soliciting 12.8 (4.39) 12.0 (3.40) 11.0 (3.28) 1.04 .03 Surveying 7.19 (1.72) 5.45 (2.25) 5.26 (2.10) 2.45 .07' Underhanded 4.94 (2.24) 4.79 (2.10) 4.11 (1.49) 0.97 .03 Country Acceptance CI Training 10.8 (4.19) 10.8 (4.20) 12.6 (3.96) 2.38 .07‘ Cooperation 6.00 (1.75) 6.38 (2.87) 6.79 (2.30) 0.45 .01 Note: *p<.10;n=66. 38 As Table 9 shows, Surveying of ones’ competitors (F = 2.45, p < .10, r2 = .07) and CI training (F = 2.38, p < .10, r2 = .07) both explain variance in the categorical variable ‘tenure in the organization.’ These findings are consistent with those above: Surveying is considered unethical by those having longer tenure in the organization, and those who endorse or have had Cl training have longer tenure in the organization. Organization Size Table 10 presents the means, standard deviations, and effect sizes (r2) for each subscale across all countries. As Table 10 shows, Soliciting information of competitors (F = 3.20, 2 < .05, r2 = .16) explains variance in the categorical variable 'organization size.‘ These findings are consistent with those above: Soliciting information is considered unethical by the organizations having the organization size of '26 to 100 employees.’ 39 .8 n c ”mo. v Q t ”202 8. one 633 eve Amos m: sea owe acme 8s and So 55588 3. one :3: r: 5.9 n: 60.3 QB ES m? 60.3 2: mean: .0 .Eooo< 9:300 8. k: as. : owe away So A8. 5 25 ES owe an. 5 one ooocoEooc: 2. see and 2e 8». 3 one So. 3 So age 9; awe So 9328 :2. owe cos «.2 66.8 m: :98 m9 5.: 3: Sci one 8:28 B. 8.? 698 RR 3.: Row 35 com 5.8 mom 653 new EoEoo 2.5.“. 8. mod and on.» so. 5 8.» $4.8 one so. : owe SSS m; .0 one me. mad ace v.5 sea or: 8mg 6.: an: 3.9 see SN? 858.22 me. one sod 8.4 am. : own 80% 8.4 no. 5 85 come 8o :2... 8. was E9 9} am. 5 Be 80.8 «to $6.3 owe awe e3 :88 8. Nos 5.: Que Amos 9: :NS one 5.3 3: No.8 02 _ecoo_oo< wm_momn:w_0 N. m SIS. 3 g g a oo_oo_5> g E32339: c~l_w. _cozm. NEmmLO cum co=m~Em .0 n 8cm_._m> ho mm .92. .0? cam... 40 Region Table 11 presented the means, standard deviations, and E statistics for each of the competitive intelligence (dependent variable) subscales and also for the two Country Acceptance (dependent variable) subscales. As Table 11 shows, variance for both Illicit (F = 1.93, p < .10) and Cl Training (F = 1.78, p < .10) are explained by differences in region. The significant F value indicates mean differences across regions. In Table 11, the mean for South American is 2.00 (SD = 0.0) whereas the mean for Western Asia is 5.67 (S_D = 4.0)-a difference of nearly four points on a scale that ranges from 3.00 to 15.00. Although all the means for Illicit are low, indicating all countries consider Illicit ways unethical, there are nonetheless relative differences that are significant. The findings are similar for CI Training (E = 1.78, p < .10). In Table 11, the lowest mean is for South America (M = 6.00; _S_D = 0.0) and the highest mean is for North America (M = 14.30; _SD = 4.97). The range is from 6.00 to 30.00; therefore, this suggests that Cl Training is limited in all regions but less so in North America. These findings are discussed further in a following section. 41 .o... v a . .883 E3555 0. m>> .385 Eefimm 0 mm ”5.86:5. u .52 ”not? 0 .5. 8.54 E2633 0 $5 8.2 82.58 u (m 8.2 Eofimm u (M ”85:2 538 n (m ”8.5:? .8200 0 <0 ”8.53:2 5.62 0 <2 ”302 .85 .8. 5 .8. 5 .E5 .25 481% .85 .85 .85 .85 e: 86 2.6 86 m: 8.4 8.4 v: 8... 8s 85 8.8.880 .85 .35 .8. 5 .85 .85 .85 .85 .85 :55 .65 a: v.2 6.2 6.9 o; 8.6 8.6 h: 86 8.6 6.: 8.55» _o 82254.8( .5550 .85 .35 .2. 5 .85 .95 .85 .35 .85 E5 .465 a... 8+ 8.4 8...” 8.6 8.4 8.6 8.6 8.6 8.» B... 885...... 655 63 63 .85 655 .85 .35 .85 .95 .85 2.6 86 26 So 63 6... 2.6 «No 8.6 8.6 So 8.623 .85 .85 .2. 5 .85 .85 .85 665 .85 655 .85 8.6 6.2 no. 6.: 8.6 6.: 5a. 6.9 6.2 86 4.: 852.8 .3. 5 .85 63 .25 .2... .25 .85 .85 .455 .455 8.6 58 6.8 New new 5.8 6.8 6.8 6.8 6.8 EN 5658 2.8.". .85 655 .85 .25 . 6.5 .85 .85 .85 .85 .85 5... 8.“ 8s 8... 8.6 he. 86 :6 8... 8e 86 .0 can. .85 .85 63 .85 .85 .85 .85 .85 .85 .85 3.. ma. 5: NS 6.: 6.2 4.: 6.9 6.2 6.2 6.2 8.86.5.2 .85 .35 .25 .85 .85 .85 .85 .85 .35 .25 8.. one 8.» 8... 8.4 8.6 :e «No SN 8.» 8.6 5.... 6m. 5 63 63 .85 63 .35 .85 .85 .85 .85 £6 8.9 8.4 8... 8... 8.4 «we to 8.6 8... we... :88 655 .85 .8. 5 655 .85 64.5 .85 .85 .95 .25 «no 3: ma. 6.: our 6.: Z: 59 8... 8.6 6.: .2828... 3.33.6 .0 m ME mm fl 8 g .5883... c0601 56. mm .683 8cm_.m> 6 6.6 :92. .2. 33m... 42 Results of Simple Regression Analysis Table 12 presents the result of simple regression analysis for eleven dependent variables from five independent variables—tenure in CI, tenure in the job position, tenure in the organization, organizational size, and region. Table 12. Results of Simple Regression Analysis for Predicting Perceptions of Ethical Boundaries and Country Acceptance from Four Independent Variables Independent Variables Dependent Variables Tenure in the Cl Tenure in the Job Tenure in the . M Position Organization Organizafion Sg‘ e Cl Subscales Accidental .02 .16 .03 .11 Deceit .04 .21 . 16 .09 Illicit .11 .26“ .08 .24' Misleading .02 .07 .02 .1 1 Paid Cl .05 .00 .02 .19 Public domain .12 .09 . 16 .07 Soliciting .08 .02 .18 .35‘“ Surveying . 10 .1 1 .23' .27” Underhanded .06 . 16 . 16 .09 Country Acceptance Cl Training .22‘ .07 .24‘ .00 Cooperation . 1 3 .09 . 12 .06 Note: ”*p<.01;**p<.05;*p<.10;N=66. 43 As shown Table 12, the r = .22 for tenure in CI field predicts a country’s acceptance of Cl training. The longer an employee works in the field of Cl, the greater the view that formal training is an acceptable practice. For job position, the r = .26 shows the prediction of Illicit Cl activities from tenure in the job position. This prediction coefficient can be interpreted as indicating that the longer an employee works in a Cl position, the more the employee perceives certain Cl activities as illicit, such as taping the conversations of competitors or tapping a competitor’s telephone, or falsely misrepresenting ones’ self as an interested candidate for an executive position in the competitor’s company. The findings for the prediction of surveying competitors from organizational tenure were surprising but consistent with the ANOVA results. For the organization tenure, r = .23. The three items for surveying inquired about the unethical (scored “1”) to ethical (scored “5”) practice of (1) survey a competitor’s employees on job satisfaction, (2) on job commitment, and (3) agreeing with a competitor on trading certain information but then withholding a small but potentially important detail. The responses to these items were summed to create the dependent variable called “surveying.” In a regression model, when responses are regressed against another variable or variables, the “validity” coefficient is positive. Thus, according to these results, the above three activities—altogether operationalized as ‘surveying,’ can be predicted from the length of time an employee is with the organization. These findings along with the others are discussed further in the discussion section. 44 Illicit, soliciting, and surveying can all be predicted from organizational size (r = .24, Illicit; r = .35, Soliciting; r = .27, surveying). That is, the extent to which of each of these three Cl practices is considered either ethical or unethical depends on the size of the organization. An interpretation could be made that larger organizations are more likely to participate in a range of CI activities that smaller companies do not. Results of Multiple Regression Analysis Table 13 presents the results of regression the nine competitive information dependent variables against the five independent variables—tenure in CI, tenure in the job position, tenure in the organization, and organization size. The multiple regression analysis revealed three significant effects. When the independent variables (tenure in the Cl field; tenure in the job position; tenure in the organization, and organization size) were all entered together in the model, the validity coefficient was .16. Also, the independent variables altogether predicted the perceptions of soliciting competitor’s information (r = .36). The predictive validity coefficient of approximately .20 or greater on scales with corresponding reliability coefficients of .70 or greater provide useful predictive estimates (Muchinsky, 1997) of the dependent variables—in this case, perceptions of ethical Cl activities. 45 Table 13. Multiple Regression Analysis Results for Subscales our Independent Variables Dependent Variables 3 Cl Subscale Accidental .24 Deceit .26 Illicit .36“ Misleading .17 Paid CI .21 Public Domain .27 Soliciting .36“ Surveying .31 Underhanded .29 Country Acceptance CI Training .34 Cooperation .19 Note: ***p < .01; **p < .05; *p < .10. 46 DISCUSSION In this thesis, I developed a risk assessment instrument that can be used to predict the extent to which competitive intelligence activities are or are not ethical. Factor analysis revealed eleven subscales that met the statistical criteria for reliability and for convergent and divergent validity. The items for these eleven scales refer to and capture a wide range of acts that people engage in to collect information on a business competitor. For example, the scales measure the perceived ethics of obtaining information available in the public domain; obtaining information from a competitor's supplier; and using information on a competitor that was accidentally obtained but which is known to be proprietary, among others. Taken altogether, these scales comprise a risk assessment instrument for estimating the risk of certain organizational behaviors; namely, competitive information gathering. In addition to providing a quantitative way to determine how to operate in other countries, a risk assessment instrument can be used to determine how organizations in other countries may tend to operate against ones' own corporate organization. The validity and reliability were established for each of the scales in a risk assessment instrument, pointing to its’ potential for organizational research and practice based on further analysis with a large sample. In addition, this thesis sought to identify how five organization variables might help to explain the statistical varianceuor the rationales for—engaging in the various (eleven) intelligence gather activities. In other words, the question 47 was asked, "to what extent do perceptions of these various activities depend on organizational variables." The hypotheses here were that the longer an individual was employed in the Cl field, the job position, and the organization, the more likely that person would be socialized to the way CI practices were condoned by his or her employer organization. Two other variables were organizational size and region; the rationale was that socialization might also depend on these factors. The results showed that four organizational variables explained variance in four ethical subscales. Tenure in the CI field was related to the possibility that the organizations that offer CI training are also those that have been actively involved in CI activities for relatively longer periods of time relative to other organizations. Tenure in the job position also explained variance in conducting illicit activities to obtain information on competitors. In this analysis, tenure in the job was a categorical variable divided into being in the job position for less than two years or from three-to-ten years, or for more than ten years. For each of these categories, the employees considered illicit activities as unethical; however, the longer tenured employees (greater than ten years) considered the illicit activities as more unethical, relative to the other two groups. In sum, the longer a person was employed in the Cl job, the more unethical the person considered illicit activities because CI professionals having longer tenure in the organization may be explosed to organizational norms and values. 48 Another finding was that tenure in organization explained variance in surveying competitors’ suppliers and customers to get business information. Tenure in the organization was measured along the same categorical continuum as tenure in the job: less than two years; from three-to-ten years; for more than ten years. The results indicated that the longer an employee had been with his or her organization, the more unethical that employee viewed the practice of surveying a competitor‘s employees to get business information on that competitor. The mean scores were in descending order from 7.19 (two year employees) to 5.45 (three-to-ten year employees) to 5.26 (more than ten years). The lower values refer to greater perceptions toward unethical behaviors. It is possible that employees who are relatively newer to the Cl field simply do not recognize some ethical boundaries or they choose to override them, but that over time they become socialized to what the organization will and will not accept as appropriate behavior. The last organizational variable, organizational size, explained variability in soliciting information from common suppliers or salespeople. In this analysis, organizational size was divided into five categories according to the number of employees in the organization, ranging from 26 employees to more than 1,000 employees. For each category, the employees considered soliciting information from a competitor’s suppliers and customers’ activities as unethical; however, as organization size increased, soliciting activities were considered more unethical. One explanation for this finding could be that larger organizations may have definite functions for competitive intelligence activities. The CI field is 49 relatively new, and smaller organizations may not formally allocate people and other resources to competitive intelligence activities. The above analyses were not conducted by country, and it was hypothesized that some countries may differ from others on acceptable competitive intelligence activities. However, the data did not include sufficiently large numbers of employees in different countries to provide the statistical analysis by country. Therefore, I divided the countries according to their geographical regions conducted the analysis by region. There were only two significant findings for region: collecting information illicitly, such as taping conversations and using pseudo executive searches to obtain information. However, the results of these analyses are uninterpretable because of the small sample sizes for some regions. For example, the findings showed that for all regions illicit activities were considered unethical, but for some regions the illicit activities were more unethical than for others. For employees in South American and Central America, the illicit Cl activities were considerably more unethical than responses from employees in Western Asia and Western Europe. However, only one person was employed in South American and two people were employed in Central America. Thus, there was a restriction of range in computing these analyses. The results were significant for the extent to which organizations with employees in different regions offered Cl training, and the data were sufficiently large to interpret these findings. Cl training was greater for organizations in 50 North America than for any other region. Next, in order of mean scores, Cl training was important for Australia, Western Europe, and Eastern Europe. These findings could be explained by the fact that perhaps these regions are more progressive insofar as the Cl field is concerned. However, these data do not reveal which countries in each region may be more or less accepting or promoting of Cl training. Further analysis by specific country is needed. Overall, however, the above results indicate that Cl differs depending on the characteristics of an organization and also on the region in which CI is practiced. In addition to the above analysis to m variability in CI activities, I computed regression models to mg perceptions of unethical CI activities from the five organizational variables: tenure in the Cl field; tenure in job position; tenure in organization, and organization size. Simple regression analyses revealed that perceptions of the extent to which Illicit activities are ethical or unethical can be predicted from tenure in the job and from organizational size. Also, the ethics associated from soliciting information can be predicted from the size of the organization—consistent with the above mean and ANOVA statistics, i.e., larger organizations may be more apt to engage in CI activities. In addition, surveying of competitors’ employees can be predicted from both length of time employed by an organization and also the size of the organization. Last, the acceptance of Cl training can be predicted from, again, the length of time an employee is involved in the Cl field and the length of time employed with an organization that practices Cl. 51 In sum, the simple regressions were consistent with the ANOVA findings of significant differences across the variables, and these results partially-- support the study hypotheses; ‘partially,’ because all of the dependent variables were hypothesized to be predicted from all of the independent variables, and not only those mentioned above. Of the single predictors, organization size is the most common predictor (for Illicit, Soliciting, and Surveying); tenure in an organization is a second common predictor (for predicting Surveying and Cl Training); and tenure in the job (for predicting Illicit activities). The validity coefficients for the prediction of ethical CI activities when all four organizational variables were entered together in a model ranged from .05 (Misleading competitors) to .23 (Cl Training). For these multiple regressions, only three CI activities were significantly predicted: Illicit activities, Surveying employees, and CI Training, and the largest predictive coefficient (i .e., the strongest predictive validity) was for CI Training. In terms of risk assessment, the instrument developed in part one of this thesis, when replicated using a large sample, can help organizations in the following ways: first, knowledge of a competitor's organizational size and tenure and other information is available in the public domain information from published annual reports, Internet web sites, and in other publicized documents and sources. Using these sources of information, organizations can legitimately estimate the extent to which competitors consider the Cl activities as being ethical or unethical—which may be an indicator of the extent to which another organization may engage in those activities. An organization can therefore 52 provide a greater security system to protect its 'information’ assets. Furthermore, using a measure of ethical perceptions acceptable in other countries can provide a guideline for expatriate employees. Although expatriates comply with their own U.S. standards, standards in other countries are not always the same, and there are as yet no standards across countries. This study did not have sufficient data from each country to make those specific estimates (of by country differences), however, the results by region suggested the importance of extending the present study, using the present variables. The basis for the hypotheses was organizational socialization, and this theory was supported in the findings. Implications are that the longer an employee is in an organization, the greater the opportunity for socialization to the policies and practices of the organization relative to Cl activities. Organizations, therefore, may introduce socialization "training" in the context of Cl activities so that the risks associated with Cl activities do not depend on either longevity in the job or in the company. Limitations There are several limitations that future research should address. First, this thesis should be viewed as a pilot study that needs to be replicated with a larger sample. The overall sample size was relatively small, and the negligible sample size for some countries prohibited by-country analysis. Although it was possible to analyze the data by region, a cross-country analysis would be more informative because, within regions, country cultures may differ. 53 Second, although the average sample size in the organizational literature is only 50 (Hunter & Schmidt, 1990), a sample size larger than the present 66 would allow other types of analysis. For example, for scale development, structural equations’ modeling is more efficient than principal factor analysis. However, structural equations’ modeling requires a substantially larger sample size. In the multiple regression models, I included only five independent variables; an acceptable sample-size rule-of-thumb is 10 subjects for each independent variable in a regression model (Nunnally, 1978). However, I computed both multiple and simple regressions, and l computed many simple regressions. The results of the simple regression analyses, therefore, should be consider preliminary. A third limitation is that the data were available for only 30 countries, and there are over 150 countries worldwide. Multinational corporations are increasingly expanding to countries around the world: it would be information to have information on ethical boundaries for competitive intelligence for all countries. Last, I only tested the predictive validity for four organizational variables. However, the prediction of perceptions of ethics or the acceptance of CI practices in different countries could more conceptually meaningful with other variables in the model. For example, measures of individual differences such as integrity may be an important variable. Furthermore, a validity greater in magnitude may be obtain with both person and organization variables in the prediction model. CONCLUSION Knowledge of acceptable ethical and unethical boundaries of collecting competitive information can reduce the likelihood that an organization's employees will breach the confidences of another company or country, and this type of knowledge can be obtained through a risk assessment based on the one developed in this thesis. In addition, knowledge of predictors of perceptions of ethical and unethical boundaries, and a country’s acceptance of those boundaries, can be used to forecast the extent to which a competitor may violate one's own corporate boundaries. Further, socialization practices can be formalized through organizational training programs to include the management of competitive intelligence collecting activities. Through formal methods of training rather than informal word—of-mouth Ieaming, employees can adapt to what an organization promotes and expects. These are areas that should be addressed in future research—extending the present results with large sample studies and research on organizational socialization. 55 REFERENCES Baranauskas, T. (1998). Insights into Brazilian competitive intelligence practices. Compgtitive Intelligence Magazine, 1(1), 41 -43. Berger, A. (1998). Getting the goods, without selling your soul. Competitive Intelligence Magazine, 1(2), 38-40. Broder, J. (1984). Risk Analysis and the §ecurity Survey. Newton, MA: Buttenrvorth-Heineman. Burgin, A. L. (1975). Organizational socialization in the induction of new police officers. Qisserta_tion Abstracts lntemational 35(11-B). 5691 -5692. Burrows, J. (1997). Criminology and Business Crime: Building the Bridge. In M. Felson 8. R. V. Clark (Eds), Business and Crime Preventig! (pp. 13-34). Monesy, NY: Willow Tree Press, Inc. ' Calof, JV. and Skinner, B. (1999). Govemment’s Role in Competitive Intelligence. Competitive Intelligence Magazine (April), 20-23. CERT Coordination Center. (1997). Security of the lntemet. E9 Frgehlick/Ket Engrclopedia of Telecommunications 15. New York: Marcel Dekker. 231- 255. Online Available: http:/Iwww.sei.cmu.edu/publications/articleslsecurity_intemet.html. Collins, J. M. & Chung, S. S. (1999). Competitive Intelligence: Ethical and Legal Boundaries for the Strategic Management of Corporate Security and Success. Proceeding oflhe SCIP Conference. Montreal. Canada 1999. Collins, J. M. (1999). A Theoretical Framework for Security Management. Manuscript in progress. Crites, J. O. (1976). Work and Careers. In R. Dubin (Ed), Handbook of work organizations and society. Chicago: Rand McNally. Crone, G. (1999, February 11). Welcome to the war room. Financial Post V1 (99), C15. Evan, W. M. (1963). Peer-group Interaction and Organizational Socialization: A Study of Employee Turnover. American Sociolggical Review, 28(3). 436- 440. Feedtuffs. (1993, March 9). Trade secret protection affected by new federal law. Copy Editor, 70 (10), 1. Feldman, D. C. (1981). The multiple socialization of organization members. Academy of Management Review, 6(2). 309-318. Felson, M., and Clarke, R. V. (1997). Business and Crime. In M. Felson & R. V. Clark (Eds.), Business and Crime Pregention (pp. 13—34). New York: VWIow Tree Press, Inc. Gidsag, R., Santorelli, B, Walsh, E., and Wells, G. (1999). Intellectual Property Crimes. The American Crimina_l Review. 103 (2_6_), 835-879. Hunter, J. E. & Schmidt, F. L. (1990). Methods of Meta-Analysis: Cgrrecting Error Ms in Research Fingpgs. Newbury Park, CA: Sage. lnbau, F. E., Farber, B. J., and Arnold, D. W. (1996). Protective Security Law. Newton, MA: ButtenNorth-Heinemann. Jones, D. (1999, Feburary 10). FBI: Spies cost US. Firms $2B a month. USA Today, p. 2B. 57 Jones, G. R. (1983). Organizational Socialization as Information Processing Activity: A Life History Analysis. Human Organization, 42(4). 314-320. Kahaner, L. (1996). Competitive Intelligencg New York: Simon & Schuster. Klein, C. (1998). Cultural differences in competitive intelligence techniques. Competitive Intelligence Magazine, 1(2), 21 -23. Lewandowsk, R. (1999, March 1). Corporate Confidential. The Financial Post Magazin, 18-25. Litton, A. (1997). Crime Prevention and the Insurance Industry. In M. Felson & R. V. Clark (Eds), Business and Crime Prevention (pp. 151 -1 96). Monesy, NY: Willow Tree Press, Inc. Louis, M. R. (1990). Acculturation in the workplace: Newcomers as lay ethnographers. In B. Schneider (Ed.), Organizational climate and culture. San Francisco, CA: Jossey-Bass Inc., Publishers. Malhota, Y. (1996). Competitive Intelligence Programs: An Overview. @BRINTInstitute [Online]. Available httpzllwww.briant.com/paperslciover.htm. Muchinsky, P. M. (1996). Psycholpgy Applied to Work. Pacifica Grove, CA: Brooks/Cole Publishing Co. Mumford, MD. and Owens, WA (1987). Methodology Review: Background Data Measures. Applied Psychological Measurement 11(1). 1-31. Nunnally, J. C. (1972). Educational Measurement and Evaluation. NY: McGraw- Hill. 58 Potter, E. H. (1998). Economic Intelligence and National Security. Ottawa, ON: Carleton University Press. Sarbin, T. R., Carney, R. M., and Eoyang, C. (Eds). (1994). Citizen Espionage: Studies in Trust and Betrayal. Westport, CT: Praeger Publishers. SAS Institute, Inc. (1995). The Little §AS Book: A Primer. Cary, NC: SAS Institute, Inc. Sawka, K. (1999). The Analyst’s Comer: Training Intelligence Analysts. Competitive Intelligence Magazine 2(2). 35-36. Schein, E. H. (1968). Organizational socialization and the profession of management. The Industrial Management Review. 9(2). 1-15. Shearing, C. D., and Stenning, P. C. (Ed.). (1987). Private Policing. NY: Sage. Simon, N. J. (1999). Individual “Subscription” to the Cl process. Competitive Intelligence Magazine, 2(3), 43-45. Simpkins, R. A. (1998). The global mind. Competitive Intelligence Magazine, 1(2), 34-36. Stuart, A. (1996). Click and Dagger. Issue of WebMaster Magazine [Online], Available: ww.cio.com/archivelwebbusiness/0796_CIick_1_content.html, July/Aug. 1996. Toffier, A. (1990). Powershift: Aw_areness. Wealth, and Violence at the Edge of the 218t Century. New York: Bantam Books. Unknow Arthor. (July 1998). Industrial Espionage Cost Companies Billions. m Training Report [On-Line]. Available: http:l/www.trainingreport.ca/spies.htm. 59 Vella, C. M., and McGonagle, J. J. (1987). Competitive Intelligence in the Computer Age. New York: Quorum Books. Van Maanen, J. (1976). Breaking in: Socialization to work. In R. Dubin (Ed), Hfldpook pf work. organizations and society. Chicago: Rand McNally. Van Maanen, J. (1975). Police Organization: A Iongitudianl examination of job attitudes in an urban police department. Administrative Science Quarterly, 29(2). 207-228. Wrnfree, L. T, Jr., Kielich, L., & Clark, R. E. (1984). On Becoming a Prosecutor: Observations on the Organizational Socialization of Law Interns. Work and Occupations, 11(2). 207-226. 60 APPENDIX A Items for Part I: Competitive Intelligence Boundaries Item Number Content of Item Company A... 1. Surveys 3 competitor’s employees on job satisfaction. 2. Fixes prices for a similarly manufactured product. 3. Do airily surveillance of competitor. 4. Spreads misinformation about the product of competitor. 5. Uses business consultants. 6. Uses employees of competitor to obtain confidential information. 7. Secretly tapes meetings of competitor. 8. Pays employees to hire into competitors company. 9. Uses financial report for the purpose of misleading competitor. 10. Solicits competitors' employees to obtain inside information. 11. Surveys competitors’ employees on theirjob commitment. 12. Pays a current employee of competitor for confidential information. 13. Unauthorizedly uses a competitor's trade secret(s). 14. Interviews competitor's employees with no intent to hire. 15. Agrees with competitor on a price range. 16. Solicits product information feedback competitor's customers. 17. Fixes bidding. 18. Spreads false information about a competitor. 19. Has an employee falsely apply for a job with a competitor, and providing that competitor with false information. 20. Hires a third party to obtain information about competitors. 21. Observes goings on of a competitor from property off-premises. 22. Cheats on an agreement with competitor to trade information. 23. Elicits information about a competitor from a salesman. 24. Uses accidentally obtained information of competitor, to its own advantage. 25. Runs "help wanted" ads to mislead competitor, when no hiring is actually planned. 26. Records telephone conversations of competitors. 61 Table 2. (Continued) Item Number Content of Item Company A... 27. Interviews executive applicants of competitors with no intent to hire, but with intent to gain information on competitor. 28. Uses phony credentials to obtain information on competitor. 29. Pays ex-employees of competitors for confidential information. 30. Operates in the same country as competitor. 31. Spreads misinformation about competitor‘s product. 32. Uses an assumed name to become a shareholder of competitors company. 33. Hires a third party to sort through competitor's discarded information, which is located on public property. 34. Pays a recycling company to obtain competitor's waste paper. 35. Creates personality profiles of managers of competitors using handwriting obtained in conference meetings with competitor. 36. Creates personality profiles of managers of competitors by analyzing publicly obtained documents. 37. Obtains end-prices of competitor's products from a common supplier. 38. Forms a partnership with ones' government agency, in order to outperform a competitor. 39. Collects information on medical trials from a medical conference and other sources. 40. Uses information from published material and public documents. 41. Uses camouflaged questioning at conferences and other meetings. 42. Uses financial reports and broker’s research studies. 43. Analyzes competitors' products. 44. Legitimately interviews ex-employees of competitors. 45. Directly observes a competitor under secret conditions. 46. Uses disclosures without subterfuge. 47. Hires a professional investigator to obtain information on competitor. 48. Engage in false negotiations with a competitor, to obtain information. 49. Hires an employee away from a competitor, to get information. 50. Uses reports of ones' own sales and other personnel. 51. Trespasses a competitor's property. 62 Table 2. (Cont’d) Item Number Content of Item Company A... 52. Bribing competitor’s employees or suppliers, to obtain information. 53. Planting an agent on the competitor‘s payroll. 54. Eavesdropping on a competitor, using electronic devices. 55. Using a commercial satellite to spy on competitors. Items for Part II: Country Acceptance of CI Item Number Content of Item Most companies... 56. Have CI programs. 57. Share CI each other. 58. Are targets of Cl. 59. Collaborate with one another for CI purposes. 60. Distrust their governments. 61. Work together to achieve common goals. 62. Consider CI a national imperative. 63. Have advanced Cl systems. 64. Training managers in CI procedures and processes. 65. Have Cl employees in other countries. 66. Offer college courses in CI. 67. Consider CI equivalent to industrial espionage. 68. Govemments assist with Cl gathering. 69. Consider CI to be unethical. 70. Have an interest in CI, but CI is still in its' infancy. 63 APPENDIX B List of Countries and Regions Region Country North America (NA) Canada United States of America (US) Central America (CA) Mexico South America (SA) Brazil Eastern Asia (EA) China Japan Taiwan Southern Asia (SA) Indonesia Malaysia India Singapore The Philippines Western Asia (WA) Israel Saudi Arabia Africa (AFR) South Africa Australia (AUS) Australia New Zealand Eastern Europe (EE) Croatia Poland Russia Slovenia Turkey Ukraine Western Europe (WE) Germany Italy Spain Sweden Switzerland The Netheriands The United Kingdom (U.K.) 64 APPENDIX C §cree Plot of Eigenvalues for Competitive Intelligence Items 10 1 8 E i g 6 2 e n v a l u e 4 s 3 45 6 2 7 8 901 2 345 67 89012 3 4567 89 012 34567 890 0 12 34567 89012 345 0510152025303540455055 Number of Competitive Intelligence Gathering Activitie Items 65 APPENDIX C, Continued Scrg_e Plot of Eigenvalues for Cogitrv Acceptance Items (a) mac-m<:mc-'m N O 2 4 6 8 10 12 14 16 Number of Country Acceptance of CI 66 Illllllllllllllllll