Insights on Individual's Risk Perception for Risk Assessment in Web ...

13 downloads 133 Views 222KB Size Report
CENTERIS 2013 - Conference on ENTERprise Information Systems / PRojMAN 2013 - ... Keywords: web-based tools, risk perception; risk management; SME's; .... of a hypothetical lottery ticket and asking individuals to rate themselves on a ...
Available online at www.sciencedirect.com

ScienceDirect Procedia Technology 9 (2013) 886 – 892

CENTERIS 2013 - Conference on ENTERprise Information Systems / PRojMAN 2013 International Conference on Project MANagement / HCIST 2013 - International Conference on Health and Social Care Information Systems and Technologies

Insights on individual’s risk perception for risk assessment in web-based risk management tools Luís Pereiraa,*, Alexandra Tenerab, João Wemansc a,b,

Faculdade de Ciências e Tecnologia (FCT), Universidade NOVA de Lisboa (UNL), Lisbon, Portugal b UNIDEMI, Department of Mechanical and Industrial Engineering c WS Energia, Founder and Senior Developer

Abstract In order to identify, assess and operate risks in early stages, risk management practices are essential in innovative projects. Small and Medium Enterprises (SME’s) are still lacking empirical models, metrics and tools to manage project risk, which must be overcame. For this purpose, this paper briefly exhibits an application that can enable SME’s to engage with risk management practices through an existing online platform - Spotrisk®. However, in this work is showed that it’s important to control potential bias that can prevail in this or similar tool, whenever surveys are employed. So, throughout the gathering of a literature review, this paper aims to lift a framework intended to correct potential biases, caused by different risk attitudes among users, and draw conditions to render a concise model to assess project risks within SME’s. © 2013 The Publishedby by Elsevier Elsevier Ltd. access under CCpeer-review BY-NC-ND license. © Authors 2013 Published Ltd.Open Selection and/or under responsibility of Selection and/or peer-review under responsibility of SCIKA – Association for Promotion and Dissemination of CENTERIS/ProjMAN/HCIST Scientific Knowledge Keywords: web-based tools, risk perception; risk management; SME’s; project management; innovation management.

* Corresponding author. Tel.: +351914888747 E-mail address: [email protected]

2212-0173 © 2013 The Authors Published by Elsevier Ltd. Open access under CC BY-NC-ND license. Selection and/or peer-review under responsibility of SCIKA – Association for Promotion and Dissemination of Scientific Knowledge doi:10.1016/j.protcy.2013.12.098

Luís Pereira et al. / Procedia Technology 9 (2013) 886 – 892

1. Introduction EU’s economies are gradually more reliant on Small and Medium Enterprises (SME’s), which to subsist in the increasingly globalized and demanding marketplace are sustained on the benefits inherent to the integrated capability to innovate and to draw and implement innovative projects [1]. Under the act of innovate lies, as its essence, the concept of creating something new. Thus, risk is bindingly inherent to the innovation process. As a consequence, early risk identification and management is vital in innovative projects [2]. However, several studies shown severe discrepancies in the attitude towards project risk management between SME’s and large firms, in the sense that SME’s tend to show some disregard concerning the importance of risk management practices, unlike large companies [3] [4]. Also, the restraining lack of resources and the reduced availability usually associated with SME’s precludes them to perform an accurate project risk management [5]. Nonetheless, little effort has been made to develop empirical models, metrics and tools to assist SME’s in an accurate and adequate management of their project’s risks [6]. Thus, this paper will first present and discuss an integrated risk perception, management and response tool, designed to SME’s and startups, which can provide an early risk assessment performed through a web-based platform and cloud database. Secondly it will be displayed an oversight across the main bias that can occur while the users perform the information inputs on the platform, as well as a literature insights to correct the pending issues. 2. A Risk perception and management support tool: the Spotrisk® This tool consists in a web-based risk management application designed to SME’s and startups, conceived to an early identification, assessment and control of risks inherent to innovative projects. This tool was developed under the support of a project co-financed by QREN, FEDER and PorLisboa, in WS Energia Lda., a New Product Development (NPD) Portuguese SME that conducts product and service development for the solar photovoltaic industry. This tool is now on the launch phase to the market. The application Spotrisk® is mainly grounded on the model of Risk Diagnosing Methodology (RDM) developed by Keizer, Halman and Song [7], and assesses a project’s risks by delegating the respective project team members to answer an adapted checklist questionnaire. The questionnaire is composed by 35 goal oriented questions, each of them approaching potential wide-ranging risks typically adjacent to innovative projects through its conceptualization, feasibility, capability and launching stages, embracing issues from technology, market, finance and operational areas [8]. Each proposed question is answered over three different parameters: its level of implementation within the assessed project; the capacity of the project team to influence the achievement of the specified goal within the time and resource limits; and finally the severity of the negative consequences on the project’s performance in the case the specified goal is not attended. All questions are answered on a Likert five-point-scale, as for “Very low” representing the lowest respect and “Very high” being the highest respect regarding the defined goal. Each response represents a numeric quantity to be used in the risk profile calculation, as shown in the table 1. The questionnaire results are then conducted into a database, where each goal is categorized into a risk class, returning from the data base the respective categorization: “Safety”, “Low Risk”, “Medium Risk”, “High Risk” or “Failure”. Also, for each goal is proposed an adequate risk response. In the end an overall rate of the project’s risk profile is calculated through a weighted average of each answered question. Now, for comprehensive purposes it will be taken an example sighted in the Figure 1, addressing over the goal no.2 from the idea stage - “2. The idea has springboard potential” - and analyzing each of the three parameters according to a hypothetical project. First, regarding the “Level of Implementation”, let’s say that the specific idea has springboard potential, meaning that it has good documented prospects to become a product or service. So the rating may be “High”. Now regarding the “Capacity to Influence”, let’s say the project team is hypothetically capable to influence the foundation of the idea’s potential, being able to add valuable features to

887

888

Luís Pereira et al. / Procedia Technology 9 (2013) 886 – 892

the project. In that case the influence could be also “High”. Finally, regarding the “Severity of the Consequences”, let’s suppose that the lack the idea’s springboard potential jeopardizes the project and that also a strong springboard potential will benefit the project. Then the impact shall be “High”. Thus the goal results indicates a “Low” state of risk and generates a “Protect” advice, leading the project manager to safeguard the project’s conditions by taking measures such as getting insurances or reviewing contracts.

Fig. 1. Print screen from the idea stage assessment of a certain project.

The same assessment is made through each goal of the whole questionnaire and in the end it is performed a project’s risk profile analysis and the platform generates several graphics and charts reflecting the risk evaluation of each stage, as well as charts that differentiates the behavior of each parameter. Such evaluation is made according to the metrics presented in Table 1. Table 1. Spotrisk’s questionnaire qualitative risk metrics

Level of implementation Answer very low low medium high very high

Value 1 2 3 4 5

Capacity to influence Answer very low low medium high very high

Value 1 2 3 4 5

Severity of the consequences Answer very low low medium high very high

Value 5 4 3 2 1

However this independent analysis may not be enough. The risk perception can differ according to each individual, being that differences in perceived risk preferences may reflect different estimates of risk inherent to a specific decision [9]. Nevertheless the differences found in the answers from different users among the same project team led to a deeper leaning over this issue. 3. Potential cognitive biases and risk perception For decades, conjectures were made in decision-making literature asserting that individuals generally perceive risk the same way when contemplating identical decision-making scenarios, presuming that individuals were fully rational, profit-maximizing, information processors. However studies shows that complex managerial decisions can be a function of behavioral factors [10]. In order to perform a decision within uncertain events, people can rely on a limited number of heuristic principles which reduce the complex task of asserting probabilities and predicting values to simpler judgmental operations. In general these heuristics are quite useful but sometimes they lead to severe and systematic errors called biases, usually found

Luís Pereira et al. / Procedia Technology 9 (2013) 886 – 892

in the intuitive judgment of probability [11]. Because cognitive biases influence the information individuals notice, and the interpretation they reach, biases may affect risk perception, causing individuals to discount on negative outcomes and in the uncertainty associated with their decisions [10]. Therefore, cognitive biases are intimately related with individual risk attitude, which is directly linked to the actual behavior of specific assortments within a goal oriented questionnaire answering. So, in this paper we seek to bring forward available literature insights about how to identify, control and reduce potential cognitive biases, in order to improve a model’s consistency and to correct cognitive biases brought in the heuristic processes that shapes the risk perception, while answering to surveys of risk tools such as the Spotrisk’s goal oriented questionnaire. While answering this questionnaire, users are challenged to compare their real project with an ideal project with specific features and evaluate its level of similarity with it. This process accounts a given subjectivity and uncertainty and it was drawn upon by what can be designated as “Representativeness heuristic” [11], in which probabilities are evaluated by the degree to which A is representative of B, i.e. by the degree to which A resembles B. For example, when A is very representative of B, the probability that A is originated from B is judged to be high. On the other hand, if A is different of B, the probability that A is originated from B is judged to be low. So, this process will lead to different assessments and to different attitudes towards uncertainty among individuals. 4. Some Available Models and Studies Several studies have been carried out in order to help validating methods to measure an individual’s risk attitude and many of them were analyzed empirically with data collected in surveys [12], [13], [14], [15]. Hypothetic survey questions bring advantages such as offering a direct measure of individual attitudes, avoiding the need to recover behavioral parameters by making restrictive identifying assumptions, and they also bring the possibility of reaching very large samples at a relatively low cost. A potentially disadvantage of using hypothetical survey questions, is that they might not predict actual behavior. However some work has been developed in order to validate survey measures by combining large surveys with real field experiments, ending up with both statistical power and confidence in the reliability of the measures [12], [13]. For example, Dohmen, Falk, Huffman, Sunde, Schupp and Wagner [12] conducted a study in Germany regarding individual risk attitudes, primarily using a survey with a sample of 22.000 individuals, where they first asked about their “willingness to take risks” on an 11-point scale, in general and in specific contexts such as car driving, financial matters, sports and leisure, career, and health. Secondly respondents also indicated their willingness to invest in a hypothetical lottery with explicit stakes and probabilities, being possible to calculate a parameter describing the curvature of the individual’s utility function. Then it was carried out a complementary field experiment based on a representative sample of 450 individuals where they actually participated on a lottery with real prizes. The results showed that the survey measures could predict actual risktaking behavior in the field experiment. Other findings were that the distribution of willingness to take risks exhibited substantial heterogeneity across individuals, which was partially explained through four exogenous factors: the willingness to take risks was negatively related to age and to being female and positively related to and height and parental education [12]. Also Ding, Hartog and Sun [13] developed their work in a similar way but with a slightly different approach, challenging Chinese respondents in a hypothetical lottery game, where the possibility to win 1,000 yuan was 10%, and asking then how much would they be willing to pay at most to buy a lottery ticket. Then, they changed the game into asking how high a probability should be at least, for the respondents to take the lottery ticket rather than the 100 yuan in cash. In the end, similarly to [12], a field experiment took place, where similar results and similar conclusions were drawn. Also, similar studies and conclusions were headed by Fausti and Gillespie [14].

889

890

Luís Pereira et al. / Procedia Technology 9 (2013) 886 – 892

The majority of these studies converged into putting forward two methods: asking for the reservation price of a hypothetical lottery ticket and asking individuals to rate themselves on a scale of risk attitude, either in general or for specific domains of life [12]. Both methods have been used to explain actual choices, such as the relationship of risk attitude with several aspects such as smoking, drinking, self-employment, employment status, financial investments, investment in human capital, the probability of unemployment, attained level of education employment in occupations differing by earnings risk, wage growth, among other things [13]. The work developed by Weber, Blais and Betz [15] brings sharper information to help us attaining the desired objective of applying a qualitative representation of an individual’s risk attitude. First it provides information about the nature of the content-specificity of risk taking. It provides a risk-attitude scale that allows researchers and practitioners to assess both conventional risk attitudes and perceived-risk attitudes in 6 commonly encountered content domains: financial, health/safety, recreational, ethics, social, and investment/gambling. This risk attitude-scale is divided into a risk perception scale, where respondents are asked to indicate how risky they perceive certain situations, providing a rating from 1 to 5, being 1 not risky at all and 5 extremely risky; and an expected benefits scale, where respondents are asked to indicate the benefits they would obtain from certain situations, providing a rating from 1 to 5, being 1 no benefits at all and 5 great benefits. This scale is applied onto 90 questions that embrace the specified 6 commonly encountered domains. Their studies also provide additional evidence for the hypothesis that perceived-risk attitude, which factors domain differences in risk perception out of risk behavior, is significantly more consistent across domains for a particular respondent than conventional risk attitude [15]. Therefore, the literature review performed indicates that employing a survey with a scale similar as the one developed in the work of Weber, Blais and Betz [15] should translate as a possible approach into making an appropriate qualitative analysis with the database inputs like the ones used in Spotrisk. Thus, by merging the data inputs from project team assessment with the existing data regarding the typical risk profiles of standard perceived risks, the platform should be able to assemble the inputs filtered by a weighted outlook, as well as to generate graphics and charts from typical views and perspectives of perceived risk and risk attitude. The amplitude of results gathered will then provide to the database a deviation from which it will be possible to classify the responses from each project team member, according to the decision inherent to his risk perception. Thus, each standard perceived risk profile is conceived through a certain tendency in the parameter answering, which leads to an overall risk profile. Some samples of current standard risks perceptions are shown in Figure 2.

Fig. 2. Sample of standard stage perceived risks profiles

Luís Pereira et al. / Procedia Technology 9 (2013) 886 – 892

For example, a “Confident” type of decision maker has a tendency to overstate the parameter “Level of Implementation”, inputting a very optimistic kind of thinking and leading the assessment for a non-risk outcome. The standard perceived risk profiles can be showed and analyzed according to an overall risk assessment, generating antagonistic profiles such as “Confident” vs. “Insecure”; “Controller” vs. “Out of Hand”; or “Protected” vs. “Vulnerable”. Yet, they can also be viewed according to each stage and with the level of consistency referred at each phase of the project, generating profiles such as “Idea Ace” or “Afraid from Market” antagonizing both beginning and project conclusion. 5. Main Conclusions This paper sought to present a literature review consistent with the actual requirements of the project risk assessment and management tool - Spotrisk® - regarding the users risk attitudes, in order to provide a solution to correct potential cognitive bias brought in the assessment of the several issues in the goal oriented questionnaire. The studies identified on the performed literature review showed that survey measures can be behaviorally relevant and consistent on the prediction of actual risk-taking behavior in field experiments. Therefore a model under the form of a survey strikes as the most appealing solution, along with a risk-attitude scale resultant from the Weber, Blais and Betz work, providing the necessary information to incur on the creation of a solution to the problems found. Thereby it should be possible to perform a useful and pragmatic approach to correct the bias on the judgments of risk perception in the goal-oriented questionnaire assessment. Acknowledgments The authors gratefully acknowledge the funding by QREN, FEDER and PorLisboa, Ministério da Ciência, Tecnologia e Ensino Superior, Portugal, under grants WS NPT QREN - LISBOA-01-0202-FEDER-011999. References [1] N. A. Ebrahim, S. Ahmed and Z. Taha, "SMEs, Virtual research and development (R&D) teams and new product development: A literature review," International Journal of the Physical Sciences, vol. 5, no. 7, pp. 916-930, 2010. [2] J. G. Vargas-Hernández and A. García-Santillán, "Management in the Innovation Project," Journal of Knowledge Management, Economics and Information Technology, vol. 1, no. 7, pp. 1-24, 2011. [3] A. Brancia, "SMES Risk Management: An Analysis of the Existing Literature Considering The Different Risk Streams.," The 8th AGSE International Entrepreneurship Research Exchange, pp. 225-239, 2011. [4] Jayathilake, "Risk Management Practices in Small and Medium Enterprises: Evidence From Sri Lanka," International Journal of Multidisciplinary Research, vol. 2, no. 7, pp. 226-234, 2012. [5] M. S. Freel, "Patterns of innovation and skills in small firms," Technovation, vol. 25, no. 2, p. 123–134, 2005. [6] G. G. Aleixo and A. B. Tenera, "New Product Development Process on High-Tech Innovation Life Cycle," World Academy of Science, Engineering and Technology, vol. 58, pp. 794-800, 2009. [7] J. A. Keizer, J. I. Halman and M. Song, "From experience: applying the risk diagnosing methodology," The Journal of Product Innovation Management, vol. 19, p. 213–232, 2002. [8] L. Pereira, A. Tenera, J. Bispo, J. Wemans, “Risk Management in Innovative SMS’s: a Web-Based Model”, International Conference on Integrity, Reliability and Failure, vol. 4, pp. 769-777, 2013. [9] J. J. Janney and G. G. Dess, "The risk concept for entrepreneurs reconsidered: New challenges to the conventional wisdom," Journal of Business Venturing, vol. 21, pp. 385-400, 2006. [10] M. Simon, S. M. Houghton and K. Aquino, “Cognitive Biases, Risk Perception, and Venture Formation: How Individuals Decide to Start Companies,” Journal of Business Venturing, vol. 15, p. 113–134, 1999. [11] D. Kahneman and A. Tversky, "Judgment under Uncertainty: Heuristics and Biases," Science, New Series, vol. 185, pp. 1124-1131, 1974. [12] T. Dohmen, A. Falk, D. Huffman, U. Sunde, J. Schupp and G. G. Wagner, "Individual Risk Attitudes: New Evidence from a Large,

891

892

Luís Pereira et al. / Procedia Technology 9 (2013) 886 – 892 Representative, Experimentally-Validated Survey," Journal of the European Economic Association , vol. 9, no. 3, pp. 522-550, 2010. [13] X. Ding, J. Hartog and Y. Sun, "Can We Measure Individual Risk Attitudes in a Survey?", IZA Discussion Papers , Institute for the Study of Labor (IZA), vol. 4807, 2010. [14] S. Fausti and J. Gillespie, "Measuring risk attitude of agricultural producers using a mail survey: how consistent are the methods?," The Australian Journal of Agricultural and Resource Economics, vol. 50, p. 171–188, 2006. [15] E. U. Weber, A.-R. Blais and N. E. Betz, "A Domain-specific Risk-attitude Scale: Measuring Risk Perceptions and Risk Behaviors," Journal of Behavioral Decision Making, vol. 15, p. 263–290, 2002.