Fuzzy analytic hierarchy process applied to technology credit scorecard considering entrepreneurs’ psychological and behavioral attributes

Abstract

As an innovative financing scheme for small and medium enterprises, credit guarantee is given for the loan of firms based on their technologies. To screen the applicant-firms, typical technology credit scoring models consider technology-related factors such as level of technology, profitability, and marketability of technology along with management characteristics, firm characteristics, and economic factors. However, psychological and behavioral attributes of entrepreneurs of the applicant-firms have received comparatively less attention, though an entrepreneur’ personal traits are likely to significantly affect the management in a small and medium enterprise. In this paper, we propose a technology credit scorecard that additionally accommodates applicant’s intelligence, personality, integrity, verbal communication, and non-verbal behaviors. Fuzzy analytic hierarchy process allows to determine the relative importance of these attributes. The proposed scorecard is expected to decrease the risk involved in technology credit funding.

Keywords

Technology evaluation model entrepreneurs’ traits fuzzy analytic hierarchy process

1 Introduction

Economy of most developed countries has similar features, according to Bass and Schrooten [1]. In such countries, 90% of firms are small and medium enterprises (SMEs). Ayyagari et al. [4] also showed that SMEs have a significant effect on GDP of each country. But, many of them face considerable financial difficulties during past decades. Subsequently, many types of credit programs have been made available for SMEs. While consumer credit scoring considers the financial state of an applicant as relatively important factor (see [7]), the technology can be a more important determinant for credit scoring of SMEs (see [46, 47]). Technology is a valuable intangible asset that has potential for various profit opportunities. It can also strengthen the competitive positions of firms (see [6]). With this background, technology-based credit scoring has been used for screening SMEs in Korea (see [37 , 66–69]).

Typical technology credit scoring model evaluates the attributes of a firm’s technology, the firm’s individual characteristics, and economic indicators (see [47]). One of the important aspects used for evaluating firm’s technology is management: how knowledgeable the top managers are about technology, how much experience they have in particular technology area, along with their organizational skill and finding ability. However, they are essentially technology related and ignore the personal characteristics of the CEO or top managers, even though it is essential to establish a recipient’s credibility prior to issuing a credit warranty for a firm. Although the credit is for a firm, a CEO’s personal traits are more likely to directly affect management in an SME than in a large firm (see [77]). In particular, entrepreneurs’ personality traits are closely related to business success (see [54 , 84]), and their intelligence has important implications for the success of their companies (see [8, 32]). Also, according to Klinger et al. [38], higher integrity in an entrepreneur has a negative association with loan default. Thus, these types of psychological traits have a significant impact on a company’s success and even loan-default. An entrepreneur’s behavioral traits during an interview are also an essential part of the evaluation for the issuance of credit (see [24, 55]). However, there has been no detailed study to incorporate these significant factors, such as an entrepreneur’s psychological and behavioral traits, into technology credit scoring model for SMEs.

Here, we argue that the technology credit scoring model for SMEs can be enhanced by considering the traits of entrepreneurs, such as their psychological and behavioral attributes; there has been little effort to apply these traits to credit scoring when also considering a firm’s technology. The basic premise of this research is that the psychological attributes of entrepreneurs can be used to predict a company’s success, which translates to loan repayment. In addition, the behavioral attributes of loan applicants during interviews can be used to predict the future of their companies, and their continued solvency.

The main purpose of this study is to construct a scorecard that takes into account all of these factors. To find the weights of the newly proposed attributes (psychological and behavioral attributes) relative to those of existing attributes (technology, SMEs characteristics, and economic indicators), we utilize both past findings (see [47]) and the results of a fuzzy analytic hierarchy process (AHP). Leung [42] argued that the preferences used for the AHP are essentially judgments of human beings based on perception and that decision makers tend to present their opinions not as “deterministic preferences” but as “perception-based judgment intervals.” Therefore, fuzzy AHP can allow a practical description of the decision making process (see [42 , 82]). By proposing a Fuzzy AHP approach, this study makes a major contribution to research on technology credit scoring. It enables to construct new scorecards that combine psychological and behavioral traits of entrepreneurs into the existing technology credit scoring model.

The following section reviews previous studies related to psychological and behavioral attributes. Section 3 introduces the data obtained through a survey and explains the process of constructing a scorecard through the fuzzy AHP methodology. Section 4 explains the application of our scorecard. Finally, Section 5 discusses conclusions and future work.

2 Psychological and behavioral attributes of entrepreneurs

While, typical credit scoring model mainly dealt with technology, SMEs characteristics, and economic indicators, this study considers entrepreneurs’ psychological and behavioral attributes that were not handed in typical model. The following literatures in this part explain how the entrepreneurs’ psychological and behavioral attributes can affect the future solvency of a company and are used to establish a basis and framework for this study.

2.1 Entrepreneurs’ psychological attributes

Following Klinger et al. [38] and Rauch and Frese [54], we subdivide psychological attributes into three categories: intelligence, personality, and integrity.

2.1.1 Entrepreneurs’ intelligence

Cognition refers to the composite activity of mental processes, including attention, memory, learning, solving, and decision making. Cognitive ability is the most important factor associated with the hiring of employees. A general mental ability (GMA) test (i.e., an intelligence or general cognitive ability test) is typically used to evaluate an individual’s intelligence. The score on such a test is one of the best predictors of future performance and learning (see [33, 56]). Additionally, the GMA test is stronger than any other method in predicting job performance (see [34, 56]). Sternberg [71] argued that intelligence contributes to successful entrepreneurship. Entrepreneurs’ intelligence is the strongest predictor of the success of their companies and is referred to as “successful intelligence.” Successful intelligence consists of three elements: practical intelligence, analytical intelligence, and creative intelligence. These elements interact with each other and lead to entrepreneurial self-efficacy, which motivates successful entrepreneurial behavior (see [8]). Furthermore, venture growth depends on the characteristics of the entrepreneur (see [65]).

However, few studies have examined the relationship between intelligence and loan default. Klinger et al. [38] measured the relationship between specific types of intelligence and loan default based on loan default data and intelligence test data obtained from 1,500 entrepreneurs in several countries. They used two popular intelligence tests, the digit span recall test and Raven’s progressive matrices, which were also used by Djankov et al. [23] and De Mel Mckenzie and Woodruff [20]. The relationship between scores on the digit span recall test and loan default was found to be insignificant, but fluid intelligence as measured by Raven’s progressive matrices was positively related to default. In our study, we consider intelligence in the more widely understood sense by considering an entrepreneur’s score on a universal GMA test.

2.1.2 Entrepreneurs’ personality

We define an entrepreneur as a person who not only establishes an organization but also recognizes and exploits opportunities to create goods and services that do not yet exist (see [64]). According to Smith [70], there are two types of entrepreneurs: the craftsman-entrepreneur and the opportunistic-entrepreneur. The latter is closer to the stereotype of entrepreneurs, tending to have advanced education and an orientation toward the future, whereas the former has low confidence but high flexibility. It has been established that successful founders have entrepreneurial personality traits characteristic of Type A behavior, defined by unique features such as a higher need for achievement, a risk-taking propensity, and tolerance for ambiguity. Entrepreneurs receive higher Type A scores than non-founders such as managers. Furthermore, their entrepreneurial attitudes and psychological traits lead to high profitability for their firms (see [9]).

The most widely used psychometric model is the Big Five model, in which an individual’s personality is assessed as a composite of the following five factors: neuroticism, extraversion, openness to experience, agreeableness, and conscientiousness (see [50]). In order to apply the Big Five model to the proposed technology scoring model, we must understand the impact of the Big Five factors on entrepreneurs’ or firms’ success. Zhao and Seibert [83] showed that entrepreneurs tend to have higher scores for openness to experience, conscientiousness, and emotional stability and lower scores for agreeableness than non-entrepreneurs. Zhao et al. [84] performed a meta-analysis of entrepreneurs’ personality traits based on the Big Five model and added a new factor, risk propensity. They found that conscientiousness, openness to experience, emotional stability, extraversion, and risk propensity were each positively related to the intention to become an entrepreneur and that risk propensity was the most influential factor. Conscientiousness and openness were each positively related to firm growth. Ciavarella et al. [19] and Zhao et al. [84] concluded that the more conscientious an entrepreneur is, the more successful his or her company will be, while greater openness to experience is not likely to be related to business success.

However, Klinger et al. [38] reported somewhat different results from those of Zhao et al. [84]. Klinger et al. [38] analyzed how the Big Five personality traits are related to business profits and loan default. The authors argued that some Big Five attributes considered to have a positive relationship with success in Zhao’s study [84] are in fact not related to the possibility of default. Furthermore, whereas Zhao et al. [84] reported that conscientiousness is a common predictor of success, Klinger et al. [38] concluded that conscientiousness is not related to business performance or default risk.

2.1.3 Entrepreneurs’ integrity

Integrity testing is used to predict overall counterproductive behaviors such as theft, absence, disciplinary action, accidents, drug abuse, and violence (see [73]). Sackett et al. [59] distinguished between “overt” and “personality-oriented” tests. The former pertains to the frequency and extent of theft, while the latter predicts work behavior by measuring certain aspects of personality, such as dependability and conscientiousness.

In this study, we focus on the latter type. An integrity test can predict job performance and behaviors (see [52]). According to previous studies (see [10 , 57]), integrity testing is significant during personnel selection, as many companies try to prevent their employees from committing unethical behaviors that may lead to enormous losses. However, studies of the relationship between integrity and entrepreneurship are insufficient to apply the results to typical credit scoring models, and the debate over the relationship between integrity and entrepreneurial success is ongoing, according to Klinger et al. [38]. The authors argued that Integrity has a negative relationship with the probability of loan default. So, one can expect that intelligence, personality and integrity of entrepreneurs can be essential for credit scoring. To sum up, we construct the following psychological attributes of loan applicants in this study: 1) intelligence, 2) personality, 3) integrity.

2.2 Entrepreneurs’ behavioral attributes

We explore entrepreneurial applicants’ behavioral attributes that may affect the credit scoring process by reviewing related studies (see [38]). There are several studies on what types of behavioral attributes will influence creditworthiness and loan repayment (see [24, 55]). Because there were no studies about direct relationship between applicants’ behavioral traits during the interview and future solvency, we explore previous studies of general behavioral traits observable during interviews for personnel selection. Entrepreneurial or general applicants’ behavioral attributes that represents creditworthiness during interviews can be classified into verbal communication and nonverbal behaviors, which include eye contact, facial expressions, attitudes, and appearances. First, verbal communication is directly expressed during interviews (see [5]) and is a fundamental factor for determining an applicant’s reliability. If the interviewee speaks fluently for a long time, this can be considered an expression of self-confidence. In contrast, inconsistent utterances point to a negative self-perception (see [76]). Secondly, eye contact is a typical characteristic of intimacy and proximity during an interview, according to Argyle and Dean [3]. Amalfitano and Kalt [2] examined the effects of eye contact on job interviewers’ evaluations and showed that interviewees are more likely to be hired if they look straight ahead rather than down. Third, maintaining a smile has a significant effect during interviews, according to Imada and Hakel [35]. Forbes and Jackson [26] concluded that maintaining a bright facial expression has a significant effect on the chances of being chosen by interviewers. Fourth, according to Forbes and Jackson [26], an upright attitude leads to positive assessments during interviews. Finally, a decent appearance may have an impact on the final decision during an interview (see [12]). Ravina [55] concluded that more beautiful applicants tended to have higher probability of getting loans but their loan-default rate was similar to that of the rest Duarte et al. [24] identified the positive effect of decent appearance of borrowers on better credit scores and higher chances for their loanfunded.

In sum, we propose to construct the following behavioral attributes for the classification of loan applicants in this study: 1) verbal communication, 2) eye contact, 3) facial expressions, 4) attitudes, and 5) appearances. These attributes directly and indirectly reflect an applicant’s personality, credibility, and even the potential for solvency in the near future.

2.3 Newly proposed attributes

We select the attributes discussed above— intelligence, personality, integrity, verbal communication, eye contact, facial expressions, attitudes, and appearances— as eight attributes that can be implemented during the technology-based credit scoring of a firm.

We assume that the attributes are in the same hierarchy. However, we need to investigate if these attributes interact with each other. Hall et al. [29] performed a meta-analysis of the relationship between verbal and nonverbal behaviors and personality and concluded that an interviewee’s verbal and nonverbal behaviors reflect his or her personality. The relationship between personality traits and observable attributes was first explored by Snyder et al. [72] and Scherer [61], who concluded that personality traits may be expressed by verbal and nonverbal behaviors. It is reasonable to assume that an applicant’s speech-related behavior is related to his or her disposition (see [62]). Conversely, an individual’s personality predicts his or her behavior (see [27]). Borkenau et al. [13] argued that extraversion and conscientiousness are significantly related to one’s face, voice, and overall appearance. Furthermore, the personality traits of entrepreneurs may have important implications for the long-term success of ventures because their personalities direct their behaviors (see [32]).

Although personality can be assumed to interact with behavioral attributes, specifically verbal expressions and nonverbal behaviors, few studies have dealt with personality and behavioral attributes at the same hierarchical level, as our fuzzy AHP does (see also Fig. 2). In this study, we add the following entrepreneurial applicants’ psychological and behavioral attributes to a typical technology credit model: 1) intelligence, 2) personality, 3) integrity, 4) verbal communication, 5) eye contact, 6) facial expressions, 7) attitudes, and 8) appearances.

As such, we use all available psychological and behavioral attributes to confirm the credibility of entrepreneurial applicants. Overall entrepreneurial traits dealt with in this study practically imply the willingness to repay in terms of credit warranty assessment. For credit scoring, two important considerations are capacity to repay and willingness to repay (see [25, 45]). Especially, willingness to repay depends on psychological variables (see [14]) and assessing willingness to repay is essentially subjective and qualitative through face-to-face meetings with borrowers (see [28]). So, the attempt to examine psychological and behavioral traits of entrepreneurs through some tests and interviews is absolutely essential for evaluators to grasp the willingness to repay of borrowers. Considering the essence of credit scoring, this study proposes the direction of improvement of technology credit scoring model.

3 Integration with fuzzy AHP

3.1 Fuzzy AHP

In order to add new psychological and behavioral attributes to the existing model, their relative weights are obtained using a fuzzy AHP. The AHP is a popular multi-criteria decision-making technique that has been utilized in many previous studies (see [30 , 53]). Specifically, various fuzzy set approaches have been applied to Saaty’s AHP (see [11 , 79]). According to Demirel et al. [22], there is no one perfect approach when using a fuzzy AHP.

In this paper, we applied the extent analysis method proposed by Chang [16], which is a widely used method designed to minimize the weaknesses of the AHP and strengthen the decision-making process (see [16, 18]). Saaty’s [58] lambda max method cannot be used to calculate a fuzzy number. The basic concept of the extent analysis is its use of (1) a triangular fuzzy number, (2) a synthetic extent value, and (3) the principle of fuzzy number comparison (see also Appendix C).

According to Chang [16], extent analysis method on fuzzy AHP, the value of fuzzy synthetic extent with respect to the ith criterion or alternative is defined as

$W_{i} = G_{i} \otimes {(\sum_{i = 1}^{k} G_{i})}^{- 1}, i = 1, 2, \dots, k_{h}$ ; the value of fuzzy synthetic extent

where $\begin{matrix} G_{i} & = & {(\prod_{j = 1}^{h} M_{ij})}^{\frac{1}{h}} ≅ (α_{i}, β_{i}, γ_{i}) \\ = & ({(\prod_{j = 1}^{k} α_{ij})}^{\frac{1}{h}}, {(\prod_{j = 1}^{k} β_{ij})}^{\frac{1}{h}}, \\ {(\prod_{j = 1}^{k} γ_{ij})}^{\frac{1}{h}}) {(\sum_{j = 1}^{k} G_{j})}^{- 1} \\ ≅ & ({(\sum_{j = 1}^{k} γ_{i})}^{- 1}, {(\sum_{j = 1}^{k} β_{i})}^{- 1}, \\ {(\sum_{j = 1}^{k} α_{i})}^{- 1}) \end{matrix}$

However, Wang et al. [75] criticize that the step of obtaining weights of criteria in the Chang’s method may allow the weight of criteria to be zero. So, this study difuzzifies the synthetic extent value by utilizing the method known as “Converting Fuzzy Data into Crisp Scores” (CFCS), presented by Opricovic and Tzeng [51] for defuzzification (see also Appendix D). This method was originally applied to a multi-criteria decision model and proven as a better defuzzification method than others (see [18 , 78]). This approach has been used to draw a moderate crisp number from multiple fuzzy numbers (see [43]).

3.2 Integration process

This section describes the methods used in this research. We should determine the relative weights of psychological and behavioral attributes introduced in previous section. Fuzzy AHP is suitable to determine the relative weights of new attributes when no past data exist. On the other hand, when the payback history data are available along with the attribute score, logistic regression is often used to update the relative weight (see [47]). We already have updated weights of existing attributes (technology attributes, economic indicators, and SME-specific characteristics), but need to find those of new attributes (psychological and behavioral attributes). Therefore, we utilize both fuzzy AHP and the results of logistic regression (Table A.1; see [47]) for a new scorecard in the following steps.

Three steps are employed to construct a new scorecard. First, we find the relative importance levels of the existing attributes (technology attributes, economic indicators, and SME-specific characteristics) and the newly proposed attributes (psychological and behavioral attributes) with the fuzzy AHP. Second, we also determine the weights of the sub-attributes of the psychological and behavioral attributes (verbal communication, eye contact, facial expressions, attitudes, appearances, intelligence, personality, and integrity) using the fuzzy AHP. Finally, we complete the new scorecard by combining the past findings of a typical credit scoring model and result of the each step (see also Fig. 1).

3.2.1 Constructing attributes and fuzzy AHP survey

The concept of linguistic variable was presented by Zadeh [80]. The value of a linguistic variable is expressed by words or sentences in natural language and can be subjectively represented by the approximate reasoning for fuzzy set theory (see [80]). The verbal judgment for pair-wise comparison is converted into a fuzzy set with a triangular membership, as described in Table 1 (see [40]). Whenever we conduct a survey for the fuzzy AHP, we use this information in Table 1. We conducted two surveys. The first is used to determine the relative weights of the following attributes: technology attributes, economic indicators, SME-specific characteristics, and psychological and behavioral attributes. The responses to the second survey are used to determine the relative weights of the eight sub- attributes of the psychological and behavioral attributes. The hierarchical structure of this study is presented in Fig. 2.

3.2.2 Estimating the overall weight of all attributes

In order to estimate the relative weights of four attributes (technology attributes, a specific firm’s characteristics, economic indicators, and psychological and behavioral attributes), we employed both fuzzy AHP and the logistic regression analysis results (see [47]). Moon and Sohn [47] introduced a technology credit scoring model by applying a logistic regression analysis to technology attributes. To reflect the results of the fuzzy AHP and the logistic regression analysis equally, we calculated the sum of the equally weighted weights of the fuzzy AHP and the logistic regression results (see also Table 3).

The fuzzy AHP was initially employed to estimate the relative weights among the four aforementioned attributes. The survey we administered to experts included six questions about relative preferences. After the overall weight of the four attributes was determined by the fuzzy AHP, we sought to consider the logistic regression results in the typical technology credit scoring model. First, we calculated the log odds ratio of each criterion’s coefficient based on Table 3 and aggregated the absolute value of the log odds ratio minus one for each criterion. Second, the relative weights of the typical model’s three attributes were measured through normalization. Because the psychological and behavioral attributes were new factors and therefore not accounted for in the logistic regression results, their overall weight was assigned by the fuzzy AHP and omitted in this step. Finally, we calculated the sum of the equally weighted weights from the fuzzy AHP and the logistic regression for the four attributes (technology attributes, a specific firm’s characteristics, economic indicators, and psychological and behavioral attributes). This provided us with an excellent estimate of the overall relative weight of each criterion.

As shown in Table 3, the calculation of the overall weights is quite different between the fuzzy AHP and the logistic regression. In the logistic regression analysis using real data, the proportion of the SME-specific weight is far greater than that of any other attribute, accounting for 70% of the total weight. Technology attributes and economic indicators comprised 17% and 13% of the total weight, respectively. On the other hand, in the results of the fuzzy AHP, the proportion of the SME-specific weight is relatively low, whereas the weight of technology attributes is greater than any other attribute.

As a result, the high-ranking attributes were the SME-specific characteristics and the technology attributes, which represented 37% and 35% of the total weight, respectively. Low-ranking attributes were psychological and behavioral attributes and economic attributes, at 15% and 12% of the total weight, respectively. An analysis of the overall weight suggested that the most influential attributes are SME-specific characteristics and the technology held by a company.

3.2.3 Estimating the specific weights of the sub-attributes

As discussed above, we estimated the weights of the sub-attributes based on Table 2. The sub-attributes weights of the three attributes in the typical model equaled the absolute value of the log odds ratio of each coefficient minus one. In order to estimate the sub-attributes’ weights for the psychological and behavioral attributes, we employed the fuzzy AHP. Our survey targeted expert committee and included 28 questions about the relative preferences between sub-attributes.

In the process of the extension analysis of the fuzzy AHP, the fuzzy synthetic extents of the psychological and behavioral attributes were calculated, as shown in Table 4 and Fig. B.1 in Appendix B. These fuzzy numbers should be converted into crisp weights by employing the CFCS defuzzification method. The specific weights of the newly proposed attributes are summarized in Fig. 3; the results indicate that an entrepreneur’s integrity and verbal communication account for a large portion of the importance of the newly proposed attributes.

We determined the weights of the three attributes in the typical model, as shown in Table 5. Variables that did not have any coefficient were omitted during the numerical calculation of the weights because they were non-significant variables (see also Table A.1 in Appendix A). The factor analysis of the typical model (see Appendix A) revealed that the technology attribute factors (F1, F3, F4, F6, F8, and F11) were meaningful, and they were then weighted. F1 and F11 were the highest ranking factors, representing 43% and 24% of all technology attributes, respectively. The factor analysis of the economic indicators (see Table A.3) indicated that ECO1 was the highest ranking at 58% , while ECO2 and ECO3 respectively accounted for 22% and 13% of the specific weight. Additionally, whether or not the candidate is a venture company was the most influential sub-attributes among SME-specific characteristics, representing 57% of the weight of these characteristics. The other factors’ weights were similar to each other.

3.2.4 Proposed technology credit scorecard

Each criterion (factor) and sub-criterion has a total possible score of 100 points. However, as discussed above, each criterion (factor) and sub-criterion had a different weight. Therefore, the overall and specific weight multiplied by each point out of 100 was used as the weighting point for each criterion and sub-criterion. The sum of these points yielded the final score for a company and entrepreneur.

Based on our findings, we propose a new technology score card (Table 6). An applicant can be scored using the new scorecard with a total of 100 possible points for each sub-criterion. The sub-attributes of SME-specific characteristics have 0 points (No) or 100 points (Yes). For example, when the company i not listed on the stock market, the company is assigned the lowest score (0 points). Thus, the example scores in the Table 6 are based on the specific weight results. The final score can be calculated by summing the weighted scores (Table 7).

4 Conclusion

Technology credit scoring model used to aid decision making for guaranteeing a loan based on a technology of a SME has been playing an important role in supporting technology based SMEs. Although top-manager’s psychological and behavioral traits can affect business success in a SME along with willingness of payback, this aspect has not been seriously considered in the technology credit scoring model.

In this study, we proposed a new scorecard incor-porating psychological and behavioral traits of entrepreneurs into the existing technology credit scoring model. We utilized the Fuzzy AHP process to determine the relative weights of entrepreneurs’ traits in technology credit scoring model. The result of this study shows that the importance of the newly added psychological and behavioral attributes turn out to be 15.22% compared to 37.09% of SME specific characteristics, 35.10% of technology attributes and 12.59% of economic indicators. Among all entrepreneurial traits, their integrity and verbal communication were found to be the most important factors for technology credit scoring. It indicates that integrity that has been mainly applied to the field of personnel selection and individual job performance can be utilized for the technology credit scoring.

This is the first study obtaining the relative weight of psychological and behavioral traits of entrepreneurs as screening factors in technology credit scoring model. The updated technology credit scoring model is expected to improve the performance by considering and evaluating psychological and behavioral attributes of entrepreneurs. Consequently, huge reduction of fund loss caused by entrepreneurs’ psychological and behavioral traits is also expected. Furthermore, the study provides the practical procedure that integrates the weights of psychological and behavioral attributes into those of other screening factors of a typical credit scoring model under fuzzy environment.

However, the current study has some limitations. The main issues are measurement system for psychological and behavioral attributes. For the areas in which various measurement methods are available, proper selection is an issue whereas new development is necessary for those areas that do not have well established measurement system. In addition, it is recommended that further research be undertaken for multi-criteria decision-making in addition to the fuzzy AHP approach. These areas are left for further research.

Footnotes

Appendix

Acknowledgments

This work was supported by a grant from the National Research Foundation of Korea (NRF) funded by the Korean government (MSIP) (2013R1A2A1A09004699). We appreciate the helpful discussion that we had with Yong-han Ju. Graduate students of the Industrial Statistics Lab. in the Dept. of Information & Industrial Engineering at Yonsei University.

References

Aczél

and Saaty

T.L.

, Procedures for synthesizing ratio judgments, Journal of Mathematical Psychology27(1) (1983), 93–102.

Amalfitano

J.G.

and Kalt

N.C.

, Effects of eye contact on the evaluation of job applications, Journal of Employment Counseling14 (1977), 46–48.

Argyle

and Dean

, Eye contact, distance and affiliation, Sociometry28 (1965), 289–304.

Ayyagari

, Beck

and Demirguc-Kunt

, Small and medium enterprises across the globe, Small Business Economics29(4) (2007), 415–434.

Argyle

, Bodily communication, Routledge (2013).

Ballester

, Garcia-Ayuso

and Livnat

, The economic value of the R&D intangible asset, European Accounting Review12(4) (2003), 605–633.

Baas

and Schrooten

, Relationship banking and SMEs: A theoretical analysis, Small Business Economics27(2-3) (2006), 127–137.

Baum

J.R.

and Bird

B.J.

, The successful intelligence of high-growth entrepreneurs: Links to new venture growth, Organization Science21(2) (2010), 397–412.

Begley

T.M.

and Boyd

D.P.

, Psychological characteristics associated with performance in entrepreneurial firms and smaller businesses, Journal of Business Venturing2(1) (1988), 79–93.

10.

Berry

C.M.

, Sackett

P.R.

and Wiemann

, A review of recent developments in integrity test research, Personnel Psychology60(2) (2007), 271–301.

11.

Boender

C.G.E.

, De Graan

J.G.

and Lootsma

F.A.

, Multi-criteria decision analysis with fuzzy pair-wise comparisons, Fuzzy sets and Systems29(2) (1989), 133–143.

12.

Boor

, Wartman

S.A.

and Reuben

D.B.

, Relationship of physical appearance and professional demeanor to interview evaluations and rankings of medical residency applicants, The Journal of Psychology113(1) (1983), 61–65.

13.

Borkenau

and Liebler

, Observable attributes as manifestations and cues of personality and intelligence, Journal of Personality63 (1995), 1–25.

14.

Bowers

, Consumer credit use by low income consumers who have had a consumer education course: An exploratory study, Journal of Consumer Affairs13(2) (1979), 334–345.

15.

Buckley

J.J.

, Fuzzy hierarchical analysis, Fuzzy sets and Systems17(3) (1985), 233–247.

16.

Chang

D.Y.

, Applications of the extent analysis method on fuzzy AHP, European Journal of Operational Research95(3) (1996), 649–655.

17.

Chou

T.Y.

, Chou

S.C.T.

and Tzeng

G.H.

, Evaluating IT/IS investments: A fuzzy multi-criteria decision model approach, European Journal of Operational Research173(3) (2006), 1026–1046.

18.

Chou

C.H.

, Liang

G.S.

and Chang

H.C.

, A fuzzy AHP approach based on the concept of possibility extent, Quality & Quantity47(1) (2013), 1–14.

19.

Ciavarella

M.A.

, Buchholtz

A.K.

, Riordan

C.M.

, Gatewood

R.D.

and Stokes

G.S.

, The Big Five and venture survival: Is there a linkage?Journal of Business Venturing19(4) (2004), 465–483.

20.

De Mel

, McKenzie

and Woodruff

, Who are the microenterprise owners? Evidence from Sri Lanka on Tokman versus De Soto, In International differences in entrepreneurship, University of Chicago Press, 2010, pp. 63–87.

21.

De Korvin

and Kleyle

, Fuzzy analytical hierarchial processes, Journal of Intelligent and Fuzzy Systems7(4) (1999), 387–400.

22.

Demirel

, Demirel

N.C.

and Kahraman

, Fuzzy analytic hierarchy process and its application, In Fuzzy Multi-Criteria Decision Making, Springer US, 2008, pp. 53–83.

23.

Djankov

, Miguel

, Qian

, Roland

and Zhuravskaya

, Who are Russia’s entrepreneurs?Journal of the European Economic Association3(2-3) (2005), 587–597.

24.

Duarte

, Siegel

and Young

, Trust and credit: The role of appearance in peer-to-peer lending, Review of Financial Studies25(8) (2012), 2455–2484.

25.

Emel

A.B.

, Oral

, Reisman

and Yolalan

, A credit scoring approach for the commercial banking sector, Socio-Economic Planning Sciences37(2) (2003), 103–123.

26.

Forbes

R.J.

and Jackson

P.R.

, Non-verbal behaviour and the outcome of selection interviews, Journal of Occupational Psychology53(1) (1980), 65–72.

27.

Funder

D.C.

, Explaining traits, Psychological Inquiry5(2) (1994), 125–127.

28.

Golin

and Delhaise

, The bank credit analysis handbook, a guide for analysts, bankers and investors, John Wiley & Sons, 2013.

29.

Hall

J.A.

, Coats

E.J.

and LeBeau

L.S.

, Nonverbal behavior and the vertical dimension of social relations: A meta-analysis, Psychological Bulletin131(6) (2005), 898.

30.

Heo

, Kim

and Cho

, Selecting hydrogen production methods using fuzzy analytic hierarchy process with opportunities, costs and risks, International Journal of Hydrogen Energy37(23) (2012), 17655–17662.

31.

Hung

M.L.

, Yang

W.F.

, Ma

H.W.

and Yang

Y.M.

, A novel multiobjective programming approach dealing with qualitative and quantitative objectives for environmental management, Ecological Economics56(4) (2006), 584–593.

32.

Hunt

R.E.

and Adams

D.C.

, Entrepreneurial behavioral profiles and company performance: A cross-cultural comparison, International Journal of Commerce and Management8(2) (1998), 33–49.

33.

Hunter

J.E.

and Hunter

R.F.

, Validity and utility of alternative predictors of job performance, Psychological Bulletin96(1) (1984), 72.

34.

Hunter

J.E.

and Schmidt

F.L.

, Intelligence and job performance: Economic and social implications, Psychology, Public Policy, and Law2(3-4) (1996), 447.

35.

Imada

A.S.

and Hakel

M.D.

, Influence of nonverbal communication and rater proximity on impressions and decisions in simulated employment interviews, Journal of Applied Psychology62(3) (1977), 295.

36.

Jensen

J.M.

, Opland

R.A.

and Ryan

A.M.

, Psychological contracts and counterproductive work behaviors: Employee responses to transactional and relational breach, Journal of Business and Psychology25(4) (2010), 555–568.

37.

Y.H.

and Sohn

S.Y.

, Updating a credit-scoring model based on new attributes without realization of actual data, European Journal of Operational Research234(1) (2014), 119–126.

38.

Klinger

, Khwaja

A.I.

and del Carpio

, Enterprising Psychometrics and Poverty Reduction, Springer, 2013.

39.

Kutlu

A.C.

, Ekmekçioğlu

and Kahraman

, A fuzzy multi-criteria approach to point-factor method for job evaluation, Journal of Intelligent and Fuzzy Systems25(3) (2013), 659–671.

40.

Kwong

C.K.

and Bai

, A fuzzy AHP approach to the determination of importance weights of customer requirements in quality function deployment, Journal of Intelligent Manufacturing13(5) (2002), 367–377.

41.

Lee

M.Y.

, Lee

A.S.

and Sohn

S.Y.

, Behavior scoring model for coalition loyalty programs by using summary variables of transaction data, Expert Systems with Applications40(5) (2013), 1564–1570.

42.

Leung

L.C.

and Cao

, On consistency and ranking of alternatives in fuzzy AHP, European Journal of Operational Research124(1) (2000), 102–113.

43.

Lin

C.J.

and Wu

W.W.

, A causal analytical method for group decision-making under fuzzy environment, Expert Systems with Applications34(1) (2008), 205–213.

44.

Liu

, Zhang

and Jin

, A multi-attribute group decision-making method based on interval-valued trapezoidal fuzzy numbers hybrid harmonic averaging operators, Journal of Intelligent and Fuzzy Systems23(5) (2012), 159–168.

45.

Marquez

, An introduction to credit scoring for small and medium size enterprises, World Bank, Washington, DC, 2008.

46.

Mestiri

and Hamdi

, Credit Risk Prediction: A comparative study between logistic regression and logistic regression with random effects, International Journal of Management Science and Engineering Management7(3) (2012), 200–204.

47.

Moon

T.H.

and Sohn

S.Y.

, Technology credit scoring model considering both SME characteristics and economic conditions: The Korean case, Journal of the Operational Research Society61 (2010), 666–675.

48.

Moon

T.H.

, Kim

and Sohn

S.Y.

, Technology credit rating system for funding SMEs, Journal of the Operational Research Society62(4) (2010), 608–615.

49.

Murphy

K.R.

and Lee

S.L.

, Personality variables related to integrity test scores: The role of conscientiousness, Journal of Business and Psychology8(4) (1994), 413–424.

50.

Norman

W.T.

, Toward an adequate taxonomy of personality attributes: Replicated factor structure in peer nomination personality ratings, The Journal of Abnormal and Social Psychology66(6) (1963), 574.

51.

Opricovic

and Tzeng

G.H.

, Defuzzification within a multicriteria decision model, Journal of Uncertainty, Fuzziness and Knowledge based Systems11(5) (2003), 635–652.

52.

Ones

D.S.

, Viswesvaran

and Schmidt

F.L.

, Comprehensive meta-analysis of integrity test validities: Findings and implications for personnel selection and theories of job performance, Journal of Applied Psychology78(4) (1993), 679.

53.

Onut

and Efendigil

, A theorical model design for ERP software selection process under the constraints of cost and quality: A fuzzy approach, Journal of Intelligent and Fuzzy Systems21(6) (2010), 365–378.

54.

Rauch

and Frese

, Let’s put the person back into entrepreneurship research: A meta-analysis on the relationship between business owners’ personality traits, business creation, and success, European Journal of Work and Organizational Psychology16(4) (2007), 353–385.

55.

Ravina

, Love & loans: The effect of beauty and personal characteristics in credit marketsNew York Universityworking paper, 2008.

56.

Ree

M.J.

and Earles

J.A.

, Intelligence is the best predictor of job performance, Current Directions in Psychological Science1(3) (1992), 86–89.

57.

Ryan

A.M.

, Schmit

M.J.

, Daum

D.L.

, Brutus

, McCormick

S.A.

and Brodke

M.H.

, Workplace integrity: Differences in perceptions of behaviors and situational factors, Journal of Business and Psychology12(1) (1997), 67–83.

58.

Saaty

, A scaling method for priorities in hierarchical structures, Journal of Mathematical Psychology15 (1977), 234–281.

59.

Sackett

P.R.

, Burris

L.R.

and Callahan

, Integrity testing for personnel selection: An update, Personnel Psychology42(3) (1989), 491–529.

60.

Saaty

T.L.

, Analytical planning, the organization of systems (Vol. 7). RWS Publications, 1985.

61.

Scherer

K.R.

, Personality inference from voice quality: The loud voice of extroversion, European Journal of Social Psychology8(4) (1978), 467–487.

62.

Scherer

K.R.

and Scherer

, Speech behavior and personality, Speech Evaluation in Psychiatry (1981), 115–135.

63.

Schmitt-Rodermund

, Pathways to successful entrepreneurship: Parenting, personality, early entrepreneurial competence, and interests, Journal of Vocational Behavior65(3) (2004), 498–518.

64.

Shane

and Venkataraman

, The promise of entrepreneurship as a field of research, Academy of Management Review25(1) (2000), 217–226.

65.

Shepherd

D.A.

, Venture capitalists’ assessment of new venture survival, Management Science45(5) (1999), 621–632.

66.

Sohn

S.Y.

, Doo

M.K.

and Ju

Y.H.

, Pattern recognition for evaluator errors in a credit scoring model for technology-based SMEs, Journal of the Operational Research Society63(8) (2011), 1051–1064.

67.

Sohn

S.Y.

and Kim

Y.S.

, Behavioral credit scoring model for technology-based firms that considers uncertain financial ratios obtained from relationship banking, Small Business Economics41(4) (2013), 931–943.

68.

Sohn

S.Y.

and Ju

Y.H.

, Updating a credit-scoring model based on new attributes without realization of actual data, European Journal of Operational Research234 (2014), 119–126.

69.

Sohn

S.Y.

, Moon

T.H.

and Kim

, Improved technology scoring model for credit guarantee fund, Expert Systems with Applications28(2) (2005), 327–331.

70.

Smith

, The entrepreneur and his firm: The relationship between type of man and type of company, Occasional Papers, Bureau of Business and Economic ResearchMichigan State University, 109, 1967.

71.

Sternberg

R.J.

, Successful intelligence as a basis for entrepreneurship, Journal of Business Venturing19(2) (2004), 189–201.

72.

Snyder

C.R.

, Shenkel

R.J.

and Lowery

C.R.

, Acceptance of personality interpretations: The Barnum Effect and beyond, Journal of Consulting and Clinical Psychology45(1) (1977), 104.

73.

Tharenou

, The relevance of industrial and organisational psychology to contemporary organisations: How far have we come and what needs to be done post-2000?Australian Psychologist36(3) (2001), 200–210.

74.

Thomas

L.C.

, A survey of credit and behavioral scoring: Forecasting financial risks of lending to customers, International Journal of Forecasting16(2) (2000), 149–172.

75.

Wang

Y.M.

, Luo

and Hua

, On the extent analysis method for fuzzy AHP and its applications, European Journal of Operational Research186 (2008), 735–747.

76.

Wiens

A.N.

, Harper

R.G.

and Matarazzo

J.D.

, Personality correlates of nonverbal interview behavior, Journal of Clinical Psychology36(1) (1980), 205–215.

77.

Wincent

and Westerberg

, Personal traits of CEOs, inter-firm networking and entrepreneurship in their firms: Investigating strategic SME network participants, Journal of Developmental Entrepreneurship10(3) (2005), 271–284.

78.

W.W.

and Lee

Y.T.

, Developing global managers’ competencies using the fuzzy DEMATEL method, Expert Systems with Applications32(2) (2007), 499–507.

79.

Van Laarhoven

P.J.M.

and Pedrycz

, A fuzzy extension of Saaty’s priority theory, Fuzzy sets and Systems11(1) (1983), 199–227.

80.

Zadeh

L.A.

, The concept of a linguistic variable and its application to approximate reasoning, Information Sciences8(3) (1975), 199–249.

81.

Zadeh

L.A.

, The role of fuzzy logic in the management of uncertainty in expert systems, Fuzzy Sets and Systems11(1) (1983), 197–198.

82.

Zadeh

L.A.

, Generalized theory of uncertainty (GTU)— principal concepts and ideas, Computational Statistics & Data Analysis51(1) (2006), 15–46.

83.

Zhao

and Seibert

S.E.

, The Big Five personality dimensions and entrepreneurial status: A meta-analytical review, Journal of Applied Psychology91 (2006), 259–271.

84.

Zhao

, Seibert

S.E.

and Lumpkin

G.T.

, The relationship of personality to entrepreneurial intentions and performance: A meta-analytic review, Journal of Management36(2) (2010), 381–404.