Comparative analysis of managers’ perception in overseas construction project risks and cost overrun in actual cases: a perspective of the Republic of Korea

ABSTRACT As the number of overseas projects increase, capabilities of managers should be comprehended by handling risks. Therefore, comparison of project managers’ risk perception and risk handling costs from actual cases is necessary to set the management direction. Thus, eleven risk factors that occur frequently in overseas construction were classified into three categories . Perception of 54 international construction managers were collected by survey and ranked by using relative importance index. Consequently, managers perceived that the impact of risks occurring due to the wrong plan was high and conflict risk was low. Then, the data from 20 cases were amplified by applying Monte-Carlo simulation with statistical verification of reliability and normality . Through multivariable regression analysis, the cost impact of the legal conflict risks with the local subcontractors and engineers was the highest,intensely differed with managers’ perception. Specifically, the risks involving various stakeholders with sufficient time to resolve were not recognized, despite the high-cost impact. Though, fully understanding perception of the host-clients or site labors was difficult, this study is vital in that the conceptual gap was narrowed by numerically comparing the difference between risk perception of managers and actual risk cost. , It would be used as a guideline by focusing on factors with high impact for educating managers at overseas sites.


Introduction
Overseas construction projects are double-edged swords for construction companies from developing countries. Evidently, expanding shares in the global construction market contributes to the economic development of their countries (Lee et al. 2011). However, failures in risk management have severely damaged the financial statuses of construction companies, given that they result in cost overruns owing to uncertainty and complexity (Ahady, Gupta, and Malik 2017). For instance, in one Arabic country, costs were overrun in 40% of the total projects (Alhomidan, 2013). In Saudi Arabia specifically, inadequate planning, resource shortages, and fluctuations in material values have caused cost overruns (Bin Seddeeq et al. 2019). In rare cases, projects with contractors from developing countries have exceeded 100% of the anticipated costs, with serious implications for the government and public (Sinesilassie, Tabish, and Jha 2018). The cost overruns for infrastructure projects are affected by factors such as political, social, legal, and technical risks (Moghayedi and Windapo 2018).
Construction companies can expect to increase their profitability, economic growth, and reputation in the global market owing to the expansion of international projects. Therefore, it is essential to focus on methods for mitigating risks to avoid financial and economic issues (Jang et al. 2019). Leading a successful overseas construction project comprising companies and related stakeholders, such as governments, owners, managers, and engineers, requires focusing on reducing or mitigating the cost overrun risk because cost performance is a critical consideration to many companies. It reflects the efficiency of a project; therefore, it is important to identify risk factors that cause cost overrun in advance for better project management (Ahmed and Anatatmula 2017). In addition, a strategic approach is often taken through in-depth discussions, aiming to stabilize participating stakeholders' positions after projects (Lee, Kim, and Kim 2017). For appropriate cost management and various risk-related decision making, it is necessary for project managers to have an in-depth understanding of the types of risks that have detrimental effects on construction projects (Wang et al. 2016). Risk identification is a stepwise procedure in risk management, and it is crucial for prioritizing responses within limited resources (Choudhry et al. 2014). Therefore, identifying and analyzing potential risks and attracting the perceptions of managers thereto in the initial stages of projects is the principal step in risk management, and it allows all participants to take measures for improving the successful completion of the projects (El-Sayegh and Mansour 2015).
The judgments from project participants in terms of risks is called risk perception, and is based on intuitions from beliefs, attitudes, and experiences (Wang et al. 2016). Numerous studies have been conducted regarding risk perception to address the uncertainty with regard to individual stakeholders, such as project managers, owner-contractors, and local communities (AlGeelani, Dabous, and Venkatachalam 2020;Iyer and Jha 2005;Wang et al. 2016;Di Maddaloni and Davis 2018). It is vital for project managers to have a clear understanding of the appropriately identified cost overrun factors because their risk perception will have direct influences on behaviors and decision making; further, they take on the majority of the responsibility for risk management. (Wang et al. 2015). Various studies have identified construction risks that managers perceive as critical success factors and key performance indices for establishing risk strategies or defining risk perceptions among managers in various countries Gupta and Verma 2020;Demirkesen and Ozorhon 2017). However, the cost impact of risk factors that managers place importance on is often separated from the actual cases, as guidelines are given nominally and risks are managed mainly by subjective judgement of their own experiences. As a result, repeated issues often occur as the site is managed according to the styles of the previous seniors . In addition, as education programs and management guidelines are provided based on experience with practical difficulties, there is a conceptual gap in reducing the private company's financial damage between practice and research. Therefore, objective data such as delay days or excess cost that can identify how much each risk actually affects the overseas project is needed, and managers who sufficiently trained to understand how to properly respond to risks (Zhong et al. 2018). Moreover, it is necessary to quantitatively compare the impacts of factors from actual cost overruns and the perceptions of managers rather than relying on managers' own models and subjective judgments (Sinesilassie, Tabish, and Jha 2018). In particular, as the Republic of Korea's overseas investment in projects is increasing, it is necessary to directly compare and validate managers' perceptions on establishing appropriate strategies for responding to risks (Lee et al. 2011;Lee, Kim, and Kim 2017). Hence, it is necessary to identify cost overrun factors and establish management methods based on the appropriate perceptions and actual cost overrun impact factors.
In this study, project managers' risk perceptions of the Republic of Korea and the risks affecting cost overruns from actual cases are compared, aiming to identify knowledge gaps by introducing various statistical methods. The literature on overseas construction projects and factors in cost overruns are reviewed to categorize and identify risks to make theoretical framework. Afterwards, the costs that were exceeded due to the risks that occurred in actual cases were classified based on the proposed framework. Based on the organized actual cost data, Monte-Carlo simulation (MCS) statistical method is introduced to amplify the data and reliability of increased data was also sought for verification of the data. With the MCS data, multivariate regression (MVR) was conducted to objectively grasp the cost impact of each risk factor. Additionally, a survey related to the risk factors impacting cost overruns is conducted with project managers from the Republic of Korea who have managed overseas construction projects. The result of survey was analyzed by applying the relative importance index (RII) to rank the significance of the risk factors how managers perceive the impact of risk factors. Finally, the subjective perception of the managers and the objective case data were compared directly through rank normalization. As the result of this study compared the impact of costs in the actual cases and managers' perceptions according to the identical proposed framework, it is novel in that it identified which areas of education and management are realistically needed to be focused. The results of the comparative analysis can be used determine the risk perceptions' of managers in an overseas construction project. This study analyzed the reasons for differences in such awareness and discusses areas that managers should focus on to improve management strategies for participating countries. By introducing various statistical analysis methods to compare the case and the perception level, it helps in understanding the ripple effect of the relationship between practitioners in the existing knowledge of management with intellectual development.

Cost overrun in overseas construction projects
Overseas construction projects are diverse, with most projects involving social infrastructure elements such as transportation structures, water facilities, and power plants for industrial development in recipient countries. Related projects often seek to promote the socioeconomic development of a nation (Ahady, Gupta, and Malik 2017). Social infrastructure projects are primarily large in scale; they include the participation of various stakeholders, such as governments, clients, contractors, and private companies, all with various levels of contribution (Sinesilassie, Tabish, and Jha 2018). Construction companies aim to earn profits and develop and expand the progress of various projects (Ahady, Gupta, and Malik 2017). In the Republic of Korea, numerous private construction companies have actively participated in overseas infrastructure projects such as waterways and expressways and in the planning for larger business developments (Kim and Kang 2017).
Participation in overseas projects is often challenged by various nontechnical issues, such as political, economical, sociocultural, and legal considerations (Utama et al. 2019;Lee et al. 2011). Overseas construction projects with greater uncertainties owing to unstable investment environments and legislation can yield cost overruns. In addition, the industry is technically sensitive, and technology must be effectively managed and operated (Bin Seddeeq et al. 2019). For example, construction companies encounter various environments and cultures in overseas projects. This scenario is particularly challenging for construction companies because it is necessary to adapt a completely different management method from domestic projects to avoid the accumulation of relative risks (Utama et al. 2019). Further, the stakeholders in overseas construction projects are primarily concern with proper management to prevent cost overruns within the contracted amount by as much as possible (Plebankiewicz 2018).
Cost overrun has been identified as the most important issue, primarily in large construction projects: approximately 90% of projects combined (Ullah, Abdullah, and Nagapan 2016). Therefore, proper cost management is important to executing a successful project. Various definitions exist for cost overruns in the field of risk management. For example, cost overruns can be considered as project risks that are excluded in the contract price (Ullah, Abdullah, and Nagapan 2016). Another definition of a cost overrun is a budget overrun or cost increase that results from unexpected excess costs, i.e., underestimations in budget measurements (Al-Hazim, Salem, and Ahmad 2017). The definition applied in this study is the total additional expenses incurred relative to the contractual amount (Derkhshanalavijeh and Teixera 2017). To respond to such cost overruns, the identification of risk factors before and during a project could ultimately determine the performance of overseas construction projects. As risks are perceived by a manager, it is necessary to establish a risk response strategy by determining the factors in cost overruns in advance (Gupta and Verma 2020;Kassem, Khoiry, and Hamzah 2020).

Project managers' risk perceptions in various countries
Risk perception is defined as a decision-maker's assessment of the risk(s) inherent in a situation (Sitkin and Pablo 1992). This means that risk perception can affect decision making because it refers to the subjective evaluation of the probability of an event occurring (Zhang and Qian 2017). Risk perception is influenced by a variety of factors, such as psychological, social, and cultural backgrounds and incorporates various subjective views derived from an individual's experience and personality (Wang et al. 2016;Lee, Kim, and Kim 2017). In particular, the risk perceptions of project managers can impact their overall decision making; accordingly, the final output is considered a reflection of a manager's capability (Liu and Chiu 2016). Therefore, management models could be developed based on the risk perceptions of managers, enabling decision support for cost contingencies (Ayub, Thaheem, and Ud Din 2016). Identifying individuals' risk perceptions and understanding risk factors by identifying stakeholders' perceptions affect project management, attitudes, judgment, and final outcomes (Wang et al. 2016;Davis 2017). Perception comes from experience; however, it is subjective and prone to errors. Accordingly, the impact of a clear cost overrun should be determined objectively, and the corresponding risk factor should be identified and responded to.
Manager perceptions of cost overruns in overseas projects differ depending on the country. China is increasingly interested in reducing cost overruns in overseas construction projects, mainly owing to a lack of knowledge and experience; in addition, perception of technical issues is increasing (Li et al. 2017). As megaprojects have been increasingly undertaken in the United Kingdom, local governance conflicts have led to a growing perception of the sentiments of different regions (Di Maddaloni and Davis 2018). Similarly, Malaysia is becoming increasingly aware of conflicts between governments and locals (Al-Sibaie et al. 2014). In developing countries, not only conflicts, but also risks such as government uncertainty, importexport difficulties, and terrorism have led to increased costs, and managers' perceptions of such costs are growing (Al- Khattab et al. 2007). The Republic of Korea considers technical risks and macroparts of politics and the economy in overseas construction; moreover, decision-making risks owing to organizational structures are also considered .
Numerous studies, including risk perception, including structured questionnaire surveys for providing and evaluating quantitative analyses, have been conducted. In a Chinese construction project were evaluated based on closed questions. The risk perceptions of managers were numerically identified based on the explanatory power of R 2 values, a standardized coefficient (β), and the p-value of each factor (Wang et al. 2016). The Relative importance index (RII) has been used to identify perceptions by ranking factors, making them easier to compare (Gündüz et al. 2013). Through unstructured interviews, information regarding managerial influences was gathered from experts in the interviewees' own words based on their varied experiences; thus, managers' perceptions were identified (Ayub, Thaheem, and Ud Din 2016;Davis 2017).

Identified overseas construction risk factors
Risk identification is the first and foremost step in risk perception, and is essential for developing a suitable risk response strategy (Lee, Kim, and Kim 2017). In overseas construction projects, the efficiency of a project should be improved by responding to risks and uncertainties in advance (Forcael et al. 2018). Moreover, because the corresponding factors and categories are often too difficult to define and abstract, it is necessary to extract factors that can be systematically and comprehensively managed (Zhong et al. 2018). To solve these difficulties, risk factors have been identified from various researchers or documents, such as "Project Management Body of Knowledge (PMBOK)" and "Projects IN Controlled Environments version 2" (PRINCE2), aiming to successfully lead projects by preventing cost overruns and delays (Pehlivan and Öztemir 2018).
Previous research studies on cost overrun factors have defined relative categories for various criteria. Risks have been categorized as behavioral, contractual, and technical (Jaffar, Tharim, and Shuib 2011), considering the local investment environment, technical risks, organization management, and economic risk (Li et al. 2017). In addition, the United Arab Emirates' publicprivate partnership has identified categories for execution, planning, financial, and operational risks (AlGeelani, Dabous, and Venkatachalam 2020). However, the risk categorizations and simplifications of the factors are subjective, and a number of theoretical frameworks exist. In the case of the Republic of Korea, risks were categorized into planning, technical, and conflict risks Lee, Kim, and Kim 2017). In general, construction managers perceive various risks occurring during the construction stage owing to poor planning (Ahady, Gupta, and Malik 2017). In the planning stage, the project cost is influenced largely by variations in material quantities, fluctuations in material prices, and changes in design (Rathi and Khandve 2013;Kim and Kang 2017). Poor project design, poor project planning, inadequate planning processes, and design deficiencies can be further categorized as planning risks, which can potentially lead to a revised project budget (Plebankiewicz 2018;Kumaraswamy and Yogeswaran 1998;Ahemd and Anantatmula 2017). In addition, planning/systemic risks related to designs are factors in the planning risk category (AlGeelani, Dabous, and Venkatachalam 2020).
Technical risks refer to risks owing to technical uncertainties, such as construction methods, procurement risks, and project completion (Li et al. 2017). Scheduling issues, specifications, and a lack of coordination among management teams are other forms of technical risks (Ahady, Gupta, and Malik 2017). Previous research has found that technical risks arising from terrain conditions, such as those concerning site conditions and land acquisition, are the main causes of cost overruns owing to environmental conditions (Sinesilassie, Tabish, and Jha 2018). Moreover, the quality of the technical specifications, information availability, and time extensions have been further defined as potential technical risks (Kumaraswamy and Yogeswaran 1998). A quality risk is a type of execution risk related to technical risk (AlGeelani, Dabous, and Venkatachalam 2020). A Nigerian public sector construction project defined the main technical risks as fluctuations in material prices and labor, plant costs, and "construction delays owing to improper technique" (Dlakwa and Culpin 1990).
Problems such as increased project costs and reductions in productivity arise from conflicts (Jaffar, Tharim, and Shuib 2011). Mismanagement from stakeholders and poor-quality communication are the causes of most conflicts. In one study, the formal conflicts between contractors, engineers, and employers were analyzed based on cases in the Vietnamese construction industry (Maemura et al. 2018). The majority of occurrences were found to result from the incompetence of project managers and claims from clients and subcontractors (Sinesilassie, Tabish, and Jha 2018). Other causes of conflict include clients' unrealistic expectations, late instructions, and over-budget requests (Kumaraswamy and Yogeswaran 1998). Conflicts from engineers owing to inspection approvals and internal issues from host countries, such as political and contractual uncertainties, are also common (Maemura et al. 2018). Subcontractor risks, such as those concerning cash flow, experience, and compliance with contract rules, have been found to yield various conflicts (Artan Ilter and Bakioglu 2018). Therefore, building trust and relationships with subcontractors is considered critical for successfully leading a project (Gupta and Thakkar 2018). Struggles and conflicts between local communities and construction companies yield cost overruns or budget deficit issues (Özen and Özen 2017). Local communities are often excluded from communication during overseas development projects despite being final users, leading to serious repercussions. Therefore, it is essential to consider local communities' interests when making essential decisions (Di Maddaloni and Davis 2018).

Research methodology
In this study, a comparison analysis was performed between international project managers' perceptions and the actual cost data for handling risk in overseas construction projects. Figure 1 shows the main processes and methodology introduced in this study. The risk factors were identified through a literature review, in which the cost impacts of risks were determined by introducing various statistical method. After classifying costs incurred due to risks in actual cases from overseas projects based on the proposed framework, these arranged data was used to conducting MCS for amplifying the data (Arnold and Yildiz 2015). As the ratio of the excess cost to the contract cost as a dependent variable, the cost impact of each risk in the actual cases was analyzed objectively through MVR. A questionnaire was then administered to identify perceptions of project managers about extracted risks based on the proposed framework. The survey participants judged the impact of cost overruns based on their own experiences of the overseas project. The results of survey were ranked through RII to identify subjective recognition of managers. Consequently, it was possible to identify which risk factors should be emphasized and managed in the overseas projects by comparing the ranks of the two analyses results.

Identification and classification of risk factors for data collection
Although there are various methods of identifying risk, in this research, the overseas risks with major impacts were classified with reference to a literature review and expert opinions. In creating the survey, the authors identified the risks arising from overseas construction projects based on various published articles (Janjić, Todorović, and Jovanović 2020). In addition, three experts and two professors with experience in overseas construction were consulted to conduct the assessment based on the identified risks. The results from the consultation provided vital insights and reliable opinions for identifying the appropriate risk factors.
Consequently, the authors proposed framework by categorized overseas construction project risks into three categories: plan risks, technical risks, and conflict risks. Plan risks occurred within a construction site owing to insufficient planning. Technical risks are associated with technological, mechanical, and onsite conditions. Conflict risks were the results of cost overruns and delays owing to claims with subcontractors, clients, and local communities. Table 1 shows various occurrences in multinational construction projects presented in the previous literature regarding the risk factors, related details, and contexts thereof defined in this study. Notably, these investigated items are subjective, and they were selected by experts as those that would be regarded as primary. The categorized risks and risk factors were coded accordingly.

Identification of cost impact of risks by simulation
The actual data analyzed in this study were 20 cases on overseas infrastructure projects from Korean engineering and construction companies such as bridges, tunnels, seaports, and power plants, as shown in Table 2 (Southeast Asia, Middle East Asia, and South America). The Korean managers managed the sites themselves by transferring the rights from the host-country. The project risks occurring in these cases were identified according to reports and data from individual sites. Based on these data, 351 risks from the sites were identified, along with the additional costs incurred while addressing them. Utilizing the opinions of five participants from the site management, the research framework of Table 1 was applied to reorganize and reclassify the risk handling costs more than planned.
To increase the accuracy of the cost impact analysis and achieve a better perspective of the long-term situation(s) based on the identified risks from Table 1, this study employed decision-making tool based on probabilistic theory from historical data. The MCS was chosen to simulate the risks for the cost impact analysis. The MCS is used in various fields to estimate the probability of an event occurring and supports the generation of probable scenarios (Pehlivan and Öztemir 2018). For example, the conceptual design of an investment in terms of capital returns and risks (Arnold and Yildiz 2015) can be simulated with the aim of optimizing the schedule variability in the cost estimation (Pehlivan and Öztemir 2018). However, results from past studies are often over emphasized, uncertain, and unavailable because they are based on soft data (Yuan et al. 2020). The definitions of the three factors mentioned can be affected by reliability (confidence intervals) rather than accuracy, variability in the information (Gupta and Thakkar 2018), and interference with the accuracy of the results, respectively.
In this study, a MCS was applied to project the risk analysis and investigate cost-related problems. In general, the MCS is one of the more efficient and effective techniques for analyzing statistical data while reducing subjectivity (Yuan et al. 2020). It is capable of initiating logic and probability processes to simulate and forecast risk impacts. Using case studies, the mean and standard deviation of each risk factor are utilized to determine its probability and potential impact. Significant random numbers based on the probability distributions are created in the MCS to analyze the effects of risk in each iteration (Bhargava et al. 2017;Pehlivan and Öztemir 2018). In this study, given the data from case studies, a normal distribution was applied to the input of MCS to extract a set of possible values after 10,000 iterations within 95% confidence. To confirm the normal distribution of data from cases, the Kolmogorov-Smirnov test was conducted. The significance was 0.127, which was greater than 0.05, therefore, it was found that the data fit a normal distribution. The cost impact of risk by performing MCS was calculated based on Equation 1, as follows: In the above, x i is the risk factor value, f(x i ) is the cost impact of the risk, σ is the standard deviation, and μ is the mean value. The cost impact is extracted from the  (2013), Dlakwa and Culpin (1990), Kim and Kang (2017), Ahmed and Anantatmula (2017) P4 Technical risks On-site condition Structure construction and procurement process delay, Additional cost for ground investigation, Unforeseen ground conditions, Delay in site handover, Bad weather conditions Rathi and Khandve (2013), Sinesilassie, Tabish, and Jha (2018), Al-Hazim, Salem, and Ahmad (2017) (2015) C4 result of the iterative process, i.e., similar to actual situations (Yuan et al. 2020). In this study, the MCS was executed using MS Excel, and the obtained values were used as input data to calculate the cost impact of risks in projects. Subsequently, the Spearman's rank correlation coefficient was used to assess the relationships between variables. In this approach, the resultant value is between −1 and 1; in particular, −1 and 1 show perfect negative or positive correlations, respectively; −1 to −0.5 or 0.5 to 1 show strong negative or positive correlations, and between −0.5 and 0.5 show weak negative or positive correlations, respectively (Gebrehiwet and Luo 2017). MVR analysis, using the risk factors from Table 1 including planning, technical, and conflict risks, was performed to investigate and determine the cost impacts. MVR analysis compares the effects of various variables, resulting in a standardized coefficient (β) . β is based on a standard deviation, a common measurement method for comparing variables, and is capable of obtaining high predictability and practical impact of independent variables affect the dependent variable (Safapour, Kermanshachi, and Kamalirad 2019;Luo, Zhang, and He 2020). To determine the level of impact from the identified factor, the dependent variable was set as the value that increases the cost as incurred by the risk per project cost, as shown in Equation 2 as follows: Here, RC is the risk cost, and PC is the total project cost. The corresponding dependent variable was used in a MVR analysis to rank the impact degrees by comparing the standardized coefficients. The Statistical Package for the Social Sciences version 24.0 was used to process the sample for further analysis.

Survey used for relative importance index (RII)
Obtaining the perceptions of managers is essential for measuring the opinions and experiences of the practitioners and representing the relationships among the opinions of the participants (Ameyaw and Chan 2015). In this study, the relevant perceptions were obtained through a survey of the risk factors affecting cost overruns. Purposive sampling was conducted to choose survey participants depending on their position and the experience levels of the project managers (from 10 to 20 years) within overseas construction projects (Janjić, Todorović, and Jovanović 2020). In addition, all participants in the survey were Korean nationals. Ultimately, 79 managers were chosen for the questionnaires, and 54 valid responses, representing 68.35% of the response rate, were received. Table 3 shows the details of the survey participants. The survey consisted of questions related to the weight of the cost impacts according to risks in overseas construction; the results ranged from low to high impact as represented on a scale from 1 to 5 points, respectively, such as in the Likert scale (Muneeswaran et al. 2020;Gebrehiwet and Luo 2017). Each survey participant was asked to evaluate the perception on the 11 risk factors suggested in the study on the 5 scale above based on their overseas construction project experiences, and participants were rewarded about $10. In this study, the reliability of the questionnaire with the five-point scale was calculated using Cronbach's coefficient alpha, which measures the internal consistency among the factors (Wang and Yuan 2011;Forcael et al. 2018). Generally, when the Cronbach's alpha is more than 0.7, the questionnaire is considered reliable (Hair et al. 2010;Janjić, Todorović, and Jovanović 2020). The questionnaire in this study yielded a Cronbach's alpha of 0.734, thereby satisfying the threshold. In general, the RII is a method for ranking the significance of factors from survey answers according to their relative importance. In this study, the RII method was used to determine the relative importance of the factors affecting cost overruns in overseas projects (Gündüz et al. 2013). Implementing the mean and standard deviation was not appropriate for evaluating the overall ranking owing to its ineffective representations of the relationships between factors (Khatib et al. 2020). The RII formula for the calculation of each factor is shown in Equation 3 as follows: RII is the relative importance index; it ranges from 0 to 1, and a greater importance is defined by a value closer to 1. W is the weight of individual factors identified by the respondent and ranges from 1 to 5; it is multiplied by the number of respondents.
A is the maximum weight, and N is the total number of respondents.

Results
Identifying the risks critical to cost overruns is a statistical process that utilizes various methods, such as MVR and RII. A comparison analysis is made between the actual risk management costs and that obtained through the questionnaire to identify the perception gaps between managers and risk handling costs from actual projects. First, it was judged whether it can be used for MVR by analyzing the MCS result of the cost data extracted from the actual case. Next, the cost impact of risk factors in actual cases was identified through the process of MVR. Then, RII was obtained to determine the rank of managers' perceptions based on the survey results. Last, the results obtained by the two analyzes were compared.

Identification of the applicability of simulation data and correlation analysis results
Cost data for each risk obtained from actual cases were classified into the proposed framework shown in Table 1. As there was a quantitative limit to directly introducing the data into statistical analysis, therefore, the MCS was implemented to amplify the number of data for overcoming the lack of data. In applying the MCS, it is important to verify the reliability and validity of the data distribution. As mentioned above, the result of the normality test was 0.127, indicating that the data were normally distributed. Moreover, to increase the reliability of the MCS results, it was compared with the origin data of the cases. Table 4 shows the obtained sample value, mean, and standard deviation of MCS for each factor, and the mean and standard deviation of cases respectively. To identify the usability of the simulation data compared to the origin case data, t-test was performed to see if there was a significant difference. As a result of each risk factor, it was found that there was no difference in the significance probability of more than 0.05. Therefore, subsequent analysis by using the data of MCS is available. Then, the correlation between the independent variables is shown in Table 5. The Spearman's correlation demonstrates the relationships among the risk factors. The results show that there is a positive relationship between factors within each category, i.e., within the planning, technical, and conflict risk categories. However, the relationships between each category are completely different. Factors affect other category factors; although a general relationship may exist, this is rarely the case.

Analyzing cost impact by factors from actual cases with simulation data
Prior to analyzing the cost impact results for each factor, the independence of the factors is analyzed using the Durbin -Watson (DW) test to validate the independence and autocorrelation in the residuals (Jang et al. 2019). DW value of less than 2.0 indicates zero correlation among the categories and factors in the regression model . In this study, the DW value obtained from the risk categories is 1.968; therefore, it can be considered that the categories are independent. The p-values from all factors are less than 0.05, indicating substantial explanatory power. Low p-values indicate a statistically significant relationship between independent and dependent variables (Moghayedi and Windapo 2018). Therefore, it can be concluded that all categories have impacts on the cost overrun in the ratios, with a confidence interval of 95%. Furthermore, the regression model produced is considered reliable for prediction, because the adjusted R 2 is 45.1%, i.e., in the range of 0.4-0.5. Table 6 shows the statistical results from the regression model for the dependent factors. The variance of inflation factor (VIF) determines the multicollinearity among independent variables. A larger VIF indicates that an independent variable is more likely to be dependent; in this study, it further shows that the variables are independent because the VIF values are all below 10 (Yum et al. 2020). In particular, the VIF of each factor (from P1 to C4) is less than 10, indicating that the variables are not dependent on each other. The standardized coefficient beta shows the independent factors with the greatest influences on the dependent variables (Yum et al. 2020). No multicollinearity exists between the independent variables used in this study; therefore, it is possible to analyze and obtain regression equations using all the independent variables. Consequently, in accordance with the guidelines of Hair et al. (2010), the conditions of linearity, multicollinearity, and normality satisfy the requirement that all independent variables should be valid variables. The p-value, which shows the significance probability, indicates that among the independent variables, construction quality errors (T3) has a value 0.074; because this is greater than 0.05, it indicates that this variable does not have a significant effect on the cost impact at a confidence interval of 95%.
As a result, the regression equation that appears based on the standardized coefficients is shown in Equation 4 as follows: From analyzing the regression equation after the exclusion of T3, the standardized coefficients indicate that the subcontractors' responsibility (C3), at 0.226, has the largest effect on cost overrun in overseas projects, with C2, T1, C4, C1, P1, P2, P3, T2, and P4 following thereafter. In addition, the significance level is smaller than 0.05 at .000, indicating that the regression model can identify the relative critical risk factors with influences on cost overruns in overseas construction projects.

Ranking of the perceptions of international construction managers by applied RII
The distribution of each response based on the survey is shown in Figure 2, and it represents the managers' risk perceptions. The left graph is organized as the frequency of the score of the survey responses in a percentage, and the right graph shows the mean and standard deviation. From the subsequent survey results, Equation (3) can be used to obtain the RII values.
The results of a detailed calculation using the RII method are shown in Table 7, where the risk importance is organized based on the perceptions of each manager. The RII can be used to calculate and rank the factors based on the degrees of reliability of the survey participants. Factors sometimes have the same mean value, in which case the factor with the lower standard deviation is considered to have a larger impact on cost overruns (Wang and Yuan, 2011). Moreover, factors with the same RII values and a higher mean factor are considered to have greater influences on cost overruns. Table 7 shows that the risk factors with higher impacts on a cost overrun are the fluctuation of material price (P4), construction quality errors (T3), and quantity errors (P3), with RII values of 0.886, 0.862, and 0.838, respectively. The risk factors with low impacts are the on-site condition (T1), local communities' complaints (C4), and client responsibility (C1), with RII values of 0.610, 0.610, and 0.643, respectively. Table 8 shows the resulting ranking from the comparison of the impacts of the risks, as analyzed based on actual cost data and the perceptions of project managers. For the averaging of the rank value, the T3 factor was excluded because the MVR revealed that it had no relation to cost overrun.  The individual factors were analyzed in detail by performing rank normalization, making it possible to compare and analyze the degrees of impact. The topranked items have a normalized weight of 1, and the bottom-ranked items have a weight of 0 (Renda and Straccia 2003). The equation for rank normalization is given by Equation 5 as follows:

Comparison of manager perception and cost overrun impacts of actual cases by rank normalization
in the above, w(i) is the value of the normalized weight, τ(i) is the ranking of the i factor, and τ is the total number of factors. According to this equation, the values in Table 8 can be arranged and diagrammed as shown in Figure 3. By analyzing the factors, a misconception of importance regarding managers' perceptions and actual risks was identified. Managers perceived that the risk category at the planning stage was crucial, but the impact of additional costs consumed by conflict risks was higher; as such, it requires increased attention from the Korean managers. In addition, the perceptions of technical risk show large differences. From the perspective of the detailed factors, for P1, P2, and T2, the Korean managers' perceptions and cost overrun impacts are similar, but there are significant differences for the "fluctuation of material price (P4)" and "construction quality errors (T3)" factors. A common risk perceived by managers is inflation and rising exchange rates, which have direct impacts on the welfare of managers and construction sites. As these issues directly affect the lives of managers at overseas construction sites, the perception of managers considered those risks relatively more impact. Moreover, since managers are in position to directly monitoring and obtaining feedback from the supervision of construction works and quality inspections, they regard it as important, and their awareness has increased. However, the results show that the cost overruns from rework or defection are insignificant  compared to the perceptions of managers. Therefore, the risks related to technical limitations are not considered significant in overseas construction projects; instead, the conflicts or cultural issues are more influential.

Discussion from findings
To control the various risks in overseas construction projects, appropriate perceptions of risks from managers are required; these perceptions must be directly related to the costs of the projects. To manage risks, the classification of risks must be clearly established. Therefore, development of framework by reviewing previous overseas construction risk studies should take precedence. The proposed framework in Table 1 classified the risks by each category, and it was verified by using correlation analysis. The correlation between factors in the category is high, but the correlation to other categories is low. Consequently, framework helps to classify risks possible, and there is a novelty in that it could be applied in other cases well. Moreover, cost overruns are a critical issue within the construction industry and require constant attention from project participants and managers (Sinesilassie, Tabish, and Jha 2018). The results from the managers' survey, representing the perceptions and actual costs, show that there are differences in terms of risk impacts. Such misperceptions can have significant impacts on practical management and can be detrimental to cost performance.
Although the importance of conflict management and relational governance has been emphasized and educated to reduce risks, this study shows the current project managers of the Republic of Korea are not sufficiently considered (Forcael et al. 2018 ;Di Maddaloni and Davis 2018). The excess costs incurred from conflicts can be reduced through the participation of the various stakeholders because it is challenging for managers to directly participate and solve such issues due to the complex structure from multinational participation. (El-Sayegh and Mansour 2015). The results from this study show that managers of the Republic of Korea perceive conflict risks as being less responsible for cost overruns compared to other risk factors. Furthermore, planning risks, such as the lack of a certain quantity, design errors, and errors in the master plan, are considered as high risks by managers because these risks are directly influenced by those highly involved in project sites. Experience from those deeply involved in management has an influence on the perception of risks when considering the impacts of the risks (Van der Linden 2014). Despite the increasing number of education of managers using PMBOK, PRINCE2, practical experience is important to understand the importance of reducing conflicts rather than acquiring management skills. Managers are more closely related to the plan risks in construction sites than the conflict risks occurring off-site in overseas projects; thus, the former are bound to be more keenly perceived (Wu et al. 2017). Moreover, the perception of the risk is greater than the management of the problem when additional opinions are requested that incorporate factors potentially solved by engaging in direct involvement.
Conflict risks often include cultural and social issues, as well as the political risks in host countries. In these cases, it is found that the focal point for managers is towards the opinions of the client and addressing risks such as negotiation with host country to procure materials, coordination with subcontractors, and resolution possible strikes with the local communities. Conflict solving is evidently constrained by their management ability. For example, in Vietnam project, there had been conflict with local engineers and even a lawsuit due to the difficulty in negotiation about the method of payment. This conflict was occurred as Korea companies did not recognize whether there were cultural differences between Korea and Vietnam and did not write it down in the contract. Although such a situation is infrequent in practice, it causes fatal impact in project progress because it takes long time and costs too much to solve the conflicts.
Furthermore, planning and technical risks in actual cases that are highly perceived by managers, management of these risks is considered appropriate by participating countries and companies. The cost overruns due to planning risks are low from actual data but the perception is high and, consequently, given more attention. For instance, in the UAE project given by the survey participants, there were frequent changes in planning stage, however, the level of the additional costs incurred was not much high because the measures to deal with risks were also continuously managed. Responding to a change in plans and applying new technologies in overseas countries aimed to directly transfer experience to recipient country. As managers directly participated in this situation, although managers' perception of related risks was high, the impact on costs overrun was not significant. Therefore, if the project is carried out in consideration of the relationship with other country to reduce conflict risks in this process based on the proposed framework, the cost overrun would be reduced. Likewise, the "fluctuation of material price (P4)" and "construction quality errors (T3)" are perceived to have high impacts on cost overruns by Korean managers, but this is not the case in practice. A study from Africa suggested a similar result, where cost estimations, based on fluctuations of material prices, were thought to be significant by managers and contractors because these risks openly affect the site situation, whereas the actual impact on cost overruns was not large (Ameh, Soyingbe, and Odusami 2010). Given that the results of this research have been similar so far, it is important for managers to focus on hard parts such as technology and quality, but it is necessary to consider and guide how to manage and establish relationships with local sub-contractors and engineers. The skills of local engineers in charge of overseas construction have increased as the more experience they have, and the government organizations responsible for taxation and fluctuations (e.g. in market prices) are becoming more advanced. Therefore, it is necessary to manage a site with respect to cultural and technical developments because national progress is not checked by the perception from a past project (Forcael et al. 2018).site condition (T1)", "subcontractors" responsibility (C3)", and "local communities' complaints (C4)" are found to have less significant influence on the perceptions of managers, despite their high proportions in risk costs. The majority of the risks are derived from cultural and social differences between the host country and international companies. To reduce the relative risks, approaches can include formal solutions, legal support, or making appropriate deals for financial compensation on procedures when providing a solution (Wang et al. 2022;. In particular, a large portion of cost overruns is a result of conflicts with lower-level teams, subcontractors, and various contractual problems. Managers of the Korea should strategically establish countermeasures against these, educate corporations and individuals to increase their management capabilities, and establish clear guidelines for managing risks occurring at a site. Therefore, it is necessary to clearly define and negotiate risk factors from the early stages of a construction project with a higher chance of success in terms of satisfactory project output (Ilyas, Li, and Ullah 2019). Neglecting conflicts at the initial stages may yield major disputes and court proceedings leading to greater cost overruns (Sinesilassie, Tabish, and Jha 2018). Therefore, managers must raise awareness on social conflicts and related risks and prepare for countermeasures.

Implications for overseas project managers
Misdirected management and ignorance are found to be significantly associated with the final project outcomes (Liu and Chiu 2016;Pehlivan and Öztemir 2018). As the structures of stakeholders become increasingly complex and the amount of information held by those at the managerial level increases, it becomes necessary for managers to foresee possible risks, e.g., to reduce conflicts directly affecting projects. Because managers better manage what they deem relevant, positive outcomes can be achieved by allowing them to manage the factors that involve high-cost overruns (Davis 2017). This study clarifies the issues that managers should consider when managing overseas construction through a practical comparison and directly evaluates managers' perceptions by reviewing the key management factors. As shown in this study, conflicts with subcontractors in the site such as bankruptcy, non-compliance with contracts, litigation settlement fee for local engineers, etc. affect cost overrun impact based on case-based objective data. However, when comparing the perception level, it was identified that project managers from the Republic of Korea were not yet clearly interested in conflicts risks. This study is differentiated in that education subject for managers could be expanded to deal with conflicts and allow them to establish a more practical direction for their overseas project management. Therefore, step-by-step education is needed to understand that managing locals is a part of roles and responsibilities of managers in the actual field, and to manage related conflict risks between host and recipient country in advance.
As the proportion of overseas projects is growing annually, the Republic of Korea mainly considers the regulation or policy of the partner country and the interactions between the governments and companies attempting to create positive relationships (Lee, Kim, and Kim 2017). However, research shows that high costs from conflicts with local people are constantly incurred at sites. Accordingly, proper consideration and education are needed, and those who manage the site need to increase their risk perceptions when such risks have potential to incur massive cost overruns. In particular, there is no clear response to all such risks arising from a site. The risks will be reduced by building positive relationships with locals and respecting cultural aspects, e.g., by deploying Korean managers to reside near the site despite the cost implications . Moreover, if it is applied to other cases based on the developed framework, managers deal with conflicts that might arise between the field community, technicians and sub-contractors in advance with more interests. The results of this research would be evident as a response guideline to reduce cost overruns that can show why overseas construction requires resolving diplomatic relations with other countries and conflicts with local residents. Doing this is expected to improve the positive impression and competitiveness of Korean companies in overseas construction projects. Therefore, careful consideration of management methods for closing the cultural and social distances between stakeholders and managers is required; additionally, the perception of these factors should be raised to reduce conflicts and prevent cost overruns.

Conclusion
In this study, the perceptions of project managers in international construction projects were examined to identify the types of risks that should be managed intensively compared to the real actual data. As cost overruns is one of the most frequently occurring risk factors within overseas construction projects, it is important to compare cost overrun simulation data from a MCS based on actual case data, as done in this study. It is also necessary to approach projects with keener attention, smarter strategies, and care to minimize damage and mitigate risks and achieve a successful project (Ahady, Gupta, and Malik 2017).
The contributed findings of this research are as follows.
(1) Previous studies have suggested the factors related to overseas construction projects that are important to management by conducting questionnaires for managers, contractors, and owners (El-Sayegh and Mansour 2015). This study has differentiation in that it compared managers' perceptions of risks and cost impact of actual risks to identify the knowledge and conceptual gap. Although it is difficult to obtain various cost and risks data from overseas projects, it is significant in that the results is identified and compared in a reliable complex statistical method. Consequently, despite the growing interest about conflicts among stakeholders academically, it was emphasized that field managers had a low level of perception in managing these issues in the overseas projects.
(2) Various descriptive analysis through statistical processes for measuring the consequential risk costs can clarify the effects of cost estimations on the overall costs of risks within an international project (Yuan et al. 2020). In this study, to increase the reliability of the data and analysis, a simulation was conducted to compensate for insufficient real data. The simulated data were extracted from reliable real-case data and applied to the analysis. As a result of performing normality verification on the data obtained in this way and performing correlation analysis, it was found that the correlation between the risk factors in each category of the proposed framework was high but less correlation among categories. Therefore, the framework could be applied in other studies. (3) Questionnaires from this study found that managers focus more on planning risk that can be directly intervened and resolved in the best possible quality. On the other hand, it was found that current Korean overseas managers do not take much consideration in relation to conflicts between local managers and subcontractors, which highly affected according to local circumstances. However, the results of descriptive analysis based on actual cost data showed that the impact of cost excess due to conflict was higher than planning and technical risks. To lead successful international construction projects, managers should increase their interest and perception of the cost overruns occurring from conflicts.
(4) It is essential that conflict risks are discussed during the initial stages of a project because the resulting cost damage could be considerable (Ilyas, Li, and Ullah 2019). In the Republic of Korea, education related to deal conflict is conducted according to a theoretical perspective, raising perception of field managers, but it is only superficial. In addition, as the management skills are learned based on the field experiences, only the technical approach is emphasized and the perception of managers is low when it comes to resolving conflicts in the site. To overcome this situation, it is necessary to reduce conflicts with subcontractors through knowledge sharing based on cultural and social approaches to reduce costs caused from these risks. Conflict risk mitigation is required to increase the level of satisfaction among stakeholders, e.g. by enhancing the social or governance structure. Furthermore, to resolve conflicts between project participants and locals and achieve a successful project, an improved cultural understanding is necessary.
This study conducts a literature review and questionnaire based on participants at high-level positions within overseas international construction projects, thereby providing important contributions for both academia and industry. Despite the vital contributions, there are limitations requiring further research. Stating the overall cost impact of a studied risk is difficult because the real data of these sectors are sensitive and, thus, validation is difficult. Therefore, it is necessary to study the various risk factors in more detail. Questionnaires were conducted to understand the perceptions of managerial positions, but the corresponding detailed opinions were not sufficient. Opinions of other participants, such as owners, clients, contractors, and designers, were not considered. Moreover, from the recipient country's perspectives, there is a limit to understanding the risks arising from the donor country's circumstances. Therefore, to understand the risk factors related to conflict, there is a need for a further study with a wider range of participants at the managerial level. This will allow for a variety of measurements and more details in terms of subjective opinions and qualitative analysis, leading to increased reliability and depth, with the aim of reducing cost overruns in international construction projects.