Study of Design and Construction of Transit Facilities in Rural Areas in USA

The United States rural transit facilities have a considerable impact on annual transportation expenditures as there are many of them and they are geographically dispersed. It is challenging to estimate the design and construction costs of the facilities, as the historical and documented cost database is insufficient; therefore, the ultimate aim of this study was to establish a baseline estimate for design and construction costs. Additionally, the key information associated with the different aspects of rural transit facility projects was also provided in this study. Interviews were conducted with professional managers of different departments of transportation (DOTs) who were involved in rural transit projects. A structured survey was then developed and distributed to various DOT representatives, and 26 of them were completed and returned. Two regression models were generated by utilizing the survey data to predict the design and construction costs of rural transit facilities, based on the size of the projects. Furthermore, the results revealed that issues arising from soil conditions and unexpected underground conditions are frequently risk factors for construction of transit facilities. It was also concluded that the popular approach to estimating the cost of the design and construction phases associated with transit facility projects is to use the data from similar projects. These findings support the need for additional literature to provide a baseline estimate for design and construction costs and key information of different important aspects of rural transit facilities.


Background
Many of the rural areas in the United States struggle to retain a sufficient range of facilities ensuring necessary services to the local population. These difficulties are exacerbated in areas experiencing population decline [1,2]. Existing literature provides multiple definitions for rural areas [3][4][5]. The United States Census defines a rural area based on its population density; accordingly, a rural area is defined as an area with a population of less than 1000 persons per square mile [5]. In 2001, the Federal Highway Administration (FHWA) defined a rural area as an area with a population of less than 5000 and/or the outside area of a metropolitan area with a population of less than 50,000 people [4]. In the present study, the research team adopted the FHWA 2001 definition.
Rural transportation plays a critical role globally in the socio and economic conditions of rural communities [6], as it provides them with access to education, work, and health services [7]. Rural mobility has been recently challenged by an increased demand that is due to an aging population and inadequate resources [8], as transit projects in rural areas commonly receive less attention and fewer resources than projects in urban areas [9] and have to compete with them for funding [10]. To develop rural mobility, governments need to invest in constructing various type of facilities, such as remote roads, transit stops, and park-and-ride lots, and minimize the amount and number of cost overruns, which are related to unreliable and inaccurate cost estimations [11,12].
Tran et al. [3] explained that cost overruns in rural transit facilities commonly occur because the scope of the projects is limited, the funding is constrained, and the locations are spread geographically. Accurate and reliable cost estimation is never easy to achieve because of the many uncertainties that are inherent in the construction process. Multiple researchers and practitioners [13][14][15][16] espoused that decision-makers and estimators must make correct and reasonable assumptions to minimize the consequences of these uncertainties, based on: (1) thorough information pertaining to the project's nature, (2) a traditional cost database, and (3) an acceptable estimation method or tool.
Some researchers and practitioners have indicated that there is a dearth of comprehensive information and documented data related to rural transit facility projects [7,10,17,18]. Hallowell et al. [17] indicated that many Department of Transportation (DOT) representatives, city/county managers, and transit officials are silent on construction management practices of rural transportation facility projects, and without their input and the benefit of sufficient data and information in existing literature on the characteristics of rural transit facilities and on viable current cost estimating practices, it is difficult to provide reliable and accurate cost estimates during the early stages of the projects [19].
The growth of rural transportation facilities plays a crucial role in the economic, social, and political health of modern society. The development of rural infrastructures, such as remote roads, bridges, and transit stops, improves the important transportation linkages between rural and urban areas and is necessary for improving the quality of life and maintaining economic stability in rural areas [3,20,21]. Multiple sectors, such as the agricultural sector, depend on transportation, particularly on roads [22,23]. The advantages and benefits of expanding regional transit facilities are recorded in the Regional Transit Coordination Guidebook [24], which espouses that the benefits are important for transit riders, transit travelers, transit providers, transportation systems, and the workforce.
The US DOT established the Rural Transportation Program in 1999 to improve the capacity of rural transit facility projects in the US. This program had several targets: (1) enhance the quality of life in rural areas, (2) increase the safety of the transportation system, (3) decrease material costs of transportation systems, (4) protect the rural natural environment, (5) enhance potential trade, (6) enhance the economic condition of rural areas, and (7) provide national security and border integrity.
In 2012, the Texas Department of Transportation (TXDOT) [25] classified transit facilities into three main categories and eight subcategories, as shown in Figure 1. The three main categories of transit facilities are operations and maintenance, large passenger, and small passenger. These classifications reflect the nature and characteristics of rural transit facilities and are beneficial for further analyses and efficient management [26]. researchers and practitioners [13][14][15][16] espoused that decision-makers and es make correct and reasonable assumptions to minimize the consequences o tainties, based on: (1) thorough information pertaining to the project's natu tional cost database, and (3) an acceptable estimation method or tool.
Some researchers and practitioners have indicated that there is a dea hensive information and documented data related to rural transit fa [7,10,17,18]. Hallowell et al. [17] indicated that many Department of Transpo representatives, city/county managers, and transit officials are silent on con agement practices of rural transportation facility projects, and without their benefit of sufficient data and information in existing literature on the ch rural transit facilities and on viable current cost estimating practices, it is d vide reliable and accurate cost estimates during the early stages of the proj The growth of rural transportation facilities plays a crucial role in the cial, and political health of modern society. The development of rural infrast as remote roads, bridges, and transit stops, improves the important trans ages between rural and urban areas and is necessary for improving the qua maintaining economic stability in rural areas [3,20,21]. Multiple sectors, su cultural sector, depend on transportation, particularly on roads [22,23]. T and benefits of expanding regional transit facilities are recorded in the Re Coordination Guidebook [24], which espouses that the benefits are impor riders, transit travelers, transit providers, transportation systems, and the w The US DOT established the Rural Transportation Program in 1999 t capacity of rural transit facility projects in the US. This program had seve enhance the quality of life in rural areas, (2) increase the safety of the trans tem, (3) decrease material costs of transportation systems, (4) protect the ru vironment, (5) enhance potential trade, (6) enhance the economic condition and (7) provide national security and border integrity.
In 2012, the Texas Department of Transportation (TXDOT) [25] classi cilities into three main categories and eight subcategories, as shown in Figu main categories of transit facilities are operations and maintenance, large p small passenger. These classifications reflect the nature and characteristics facilities and are beneficial for further analyses and efficient management [ The American Public Transportation Association recommended usefu implementing a new bus transit facility project and maintenance services in activities associated with the facility services and maintenance were categor classes and levels (Level I, Level II, and Level III), as presented in F The American Public Transportation Association recommended useful practices for implementing a new bus transit facility project and maintenance services in 2010 [27]. The activities associated with the facility services and maintenance were categorized into three classes and levels (Level I, Level II, and Level III), as presented in Figure 2. Their classifica-Sustainability 2022, 14, 1338 3 of 16 tion leads to higher performance, efficient funding allocation, and effective management of rural transit facilities [28][29][30].
leads to considerably fewer transportation cost overruns. [11] stated that transit agencies require reliable cost estimates for the following reasons: (1) reliable estimated costs of transit facilities in the form of simple averages or cost functions during the planning phase [31][32][33], assist in establishing the funding commitments, and prevent the likelihood and magnitude of cost overruns [12]; (2) cost performance is the key factor in assessing the overall cost-effectiveness of transit facility projects [34]; and (3) accurate cost estimation and prediction help transit agencies appropriately allocate and prioritize resources in situations where the demand for transit outstrips the available funding. For the stated reasons and for purposes of effective management of transit systems in general, agencies seek to continually develop and update planning-level cost models [11]. Multiple researchers and practitioners believe that the challenges associated with estimating costs for rural transit facilities include a lack of documented cost data, remote locations, and less competition [3,35,36]. To address those challenges, [3] recommended various resources and methods, such as providing pricing directories and precise cycle time charts for machinery, supplies, and manpower.
Anderson et al. [37] introduced some issues regarding rural and small transit facility projects: the inability to complete rural and small transit facility projects due to lack of funding, and the prevalent use of consultants for estimating the cost of the projects. Virtually all of the DOTs hire consultants for cost estimations of these projects, and then hire additional consultants to review the accuracy of the estimates.
Zheng [38] identified the reasons that it is difficult to accurately estimate the costs of rural transit facility projects: (1) lack of an organized procedure for cost estimating through the planning, programming, and project development phases; (2) possibility of renovation of the rural transit facilities; (3) a wide range of functions in these projects; (4) the wide size of transit facility projects; and (5) unique challenges associated with transit facility projects. Rural transit facility projects are commonly small in scope, but numerous and geographically dispersed, so these projects account for a significant portion of funds expended on transportation projects in the U.S. [3]. Accurate cost estimation of these projects leads to considerably fewer transportation cost overruns. [11] stated that transit agencies require reliable cost estimates for the following reasons: (1) reliable estimated costs of transit facilities in the form of simple averages or cost functions during the planning phase [31][32][33], assist in establishing the funding commitments, and prevent the likelihood and magnitude of cost overruns [12]; (2) cost performance is the key factor in assessing the overall costeffectiveness of transit facility projects [34]; and (3) accurate cost estimation and prediction help transit agencies appropriately allocate and prioritize resources in situations where the demand for transit outstrips the available funding. For the stated reasons and for purposes of effective management of transit systems in general, agencies seek to continually develop and update planning-level cost models [11].
Multiple researchers and practitioners believe that the challenges associated with estimating costs for rural transit facilities include a lack of documented cost data, remote locations, and less competition [3,35,36]. To address those challenges, [3] recommended various resources and methods, such as providing pricing directories and precise cycle time charts for machinery, supplies, and manpower.
Anderson et al. [37] introduced some issues regarding rural and small transit facility projects: the inability to complete rural and small transit facility projects due to lack of funding, and the prevalent use of consultants for estimating the cost of the projects. Virtually all of the DOTs hire consultants for cost estimations of these projects, and then hire additional consultants to review the accuracy of the estimates.
Zheng [38] identified the reasons that it is difficult to accurately estimate the costs of rural transit facility projects: (1) lack of an organized procedure for cost estimating through the planning, programming, and project development phases; (2) possibility of renovation of the rural transit facilities; (3) a wide range of functions in these projects; (4) the wide size of transit facility projects; and (5) unique challenges associated with transit facility projects.

Research Gap
Accurate cost estimating is one of the main challenges associated with rural transportation facility projects [39]. Furthermore, the estimation of design and construction costs for transportation facilities in rural areas have rarely been studied. Therefore, the ultimate aim of this study was to provide a baseline estimate for the design and construction costs, based on the size of the rural transit facilities during the early stages of projects. In addition, multiple objectives were formulated to support the literature: (1) analyze the types of design and construction contracts, (2) investigate the cost estimating methods used in the design and construction phases, (3) assess the cost of each construction system, (4) review construction risk factors, and (5) generate statistical models to facilitate the estimation of design and construction costs. This study will provide an important resource for establishing baseline estimates for the design and construction costs of rural projects and will assist DOTs in minimizing cost overruns.

Research Framework
In this step, a conceptual framework was provided to illustrate the objective of each step of the research process and to show how the objectives relate to one another. The conceptual research framework, presented in Figure 3, began with a thorough review of transit agencies' related documents and transportation agencies' websites, such as the Metropolitan Transportation Commission's online library. These resources include important details on common types and sizes of rural transit facilities, site factors, and statistical data on the design and building costs. Following a review of the literature, characteristics of rural transportation facilities were described, as well as the scope of the state of practice was identified.

Research Gap
Accurate cost estimating is one of the main challenges associated with rural transportation facility projects [39]. Furthermore, the estimation of design and construction costs for transportation facilities in rural areas have rarely been studied. Therefore, the ultimate aim of this study was to provide a baseline estimate for the design and construction costs, based on the size of the rural transit facilities during the early stages of projects. In addition, multiple objectives were formulated to support the literature: (1) analyze the types of design and construction contracts, (2) investigate the cost estimating methods used in the design and construction phases, (3) assess the cost of each construction system, (4) review construction risk factors, and (5) generate statistical models to facilitate the estimation of design and construction costs. This study will provide an important resource for establishing baseline estimates for the design and construction costs of rural projects and will assist DOTs in minimizing cost overruns.

Research Framework
In this step, a conceptual framework was provided to illustrate the objective of each step of the research process and to show how the objectives relate to one another. The conceptual research framework, presented in Figure 3, began with a thorough review of transit agencies' related documents and transportation agencies' websites, such as the Metropolitan Transportation Commission's online library. These resources include important details on common types and sizes of rural transit facilities, site factors, and statistical data on the design and building costs. Following a review of the literature, characteristics of rural transportation facilities were described, as well as the scope of the state of practice was identified. As aspects of rural transit facilities were required that were not addressed in existing literature, multiple interviews were conducted to help the research team develop a structured survey. Telephone interviews are commonly considered an additional source and useful method for procuring information for reviewing and assessing rural transit facilities (NCHRP 20-65 Task 53) [36], and were utilized in this research. An interview protocol was developed, based on the information that the research team obtained through review of the related documents from the transit agencies. It consisted of 13 in-depth questions belonging to the categories of general information, project size, project cost, project duration, historical cost data, risk factors, contingency estimation, and critical items.
A structured survey was then developed to collect comprehensive information and data regarding the various aspects of rural and small transit facilities. As shown in Figure  3, the survey protocol consisted of questions related to the project memorandum, survey instructions, survey declarations, respondent information, characteristics of the project, As aspects of rural transit facilities were required that were not addressed in existing literature, multiple interviews were conducted to help the research team develop a structured survey. Telephone interviews are commonly considered an additional source and useful method for procuring information for reviewing and assessing rural transit facilities (NCHRP 20-65 Task 53) [36], and were utilized in this research. An interview protocol was developed, based on the information that the research team obtained through review of the related documents from the transit agencies. It consisted of 13 in-depth questions belonging to the categories of general information, project size, project cost, project duration, historical cost data, risk factors, contingency estimation, and critical items.
A structured survey was then developed to collect comprehensive information and data regarding the various aspects of rural and small transit facilities. As shown in Figure 3, the survey protocol consisted of questions related to the project memorandum, survey instructions, survey declarations, respondent information, characteristics of the project, cost estimating, project schedules, project risks, change orders, and others. The research team distributed the survey to more than 1600 personnel working in 52 DOT agencies that were involved in rural small transit projects, and sent them two follow-up emails. They then called the DOTs to ask them to complete the survey, but were told that the DOT agencies did not have any historical or documented data or information pertaining to rural and small transit facility projects.
Next, the research team expanded their efforts by locating private firms nationwide who conducted construction projects in rural communities and sending the survey to them. After sending two follow-up emails, they received a few responses to the effect that they could not share the information and data due to confidentiality mandates.
Finally, the research team contacted the National Rural Transit Assistance Program (National RTAP) regarding the low response rate, and the executive director published an announcement in their newsletter. After all of the mentioned efforts, only 26 completed surveys were collected.
Then, descriptive data analyses were performed for the design and construction phases to describe the basic features of the data in this study. To provide useful information and data for different project parties, such as owners and contractors, different aspects of rural transit facility projects were assessed and analyzed. The contract types were reviewed and the implemented cost estimating methods were assessed and analyzed. Additionally, cost breakdowns corresponding to each construction system of rural transportation facilities were studied. Finally, the regression analysis technique was used to develop a predictive modeling technique for cost estimates. Two regression models were generated to examine the relationship between the design costs/construction costs and project size.

Interview
The research team made concerted efforts to develop an interview protocol that considered critical questions, for which there were few, if any, literature resources. The protocol consisted of 13 questions, classified into eight categories: general information, project size, project cost, project duration, historical cost data, risk factors, contingency estimation, and critical items. These could be covered in a one-hour interview.
The interview protocol and project memorandum were sent to 13 professional consultants and DOT employees (five DOT employees, six transit managers, and two consultants) who were actively involved in transit management teams, had more than ten years of experience in rural and small urban transit facility projects, and were located in different regions of the United States. The project memorandum presented information pertaining to the background of this study, such as objectives, procedures, and time and date of the interview. The purpose of the interview was explained in the research background, and the characteristics of the rural transit facility projects were explained in the expectations and instructions. The interviewees were expected to review the interview protocol and project memorandum so that they could be prepared for the interview questions. Of the 13 invited to participate, only six were willing to be interviewed: three DOT personnel, two consultants, and one transit manager. Multiple articles and books have suggested that anywhere from 5 to 50 participants as adequate for conducting in-depth interviews [40].

Survey
After conducting the phone interviews, the research team utilized the results of interviews and developed a survey that was designed to gather historic project-specific cost estimates from the consulting firms, state DOTs, and transit agencies associated with transit facilities: (1) types and sizes of facilities, (2) features of facilities, (3) locations of facilities, (4) actual design costs, (5) actual construction costs, (6) design schedules, (7) construction schedules, (8) unexpected conditions, and (9) cost of construction components. Brief descriptions of the 11 survey sections are presented in Table 1.

Project Memorandum
The purpose of this section is to discuss the objectives of the research, email address of the research team, and the timeline for completing the online survey.

Survey Instructions
This section discussed specific types of transit facilities that have been constructed within the past five years.

Survey Declaration
The survey declaration was created to ensure that participants had a working knowledge of cost estimate procedures for rural and small urban transit facility projects and consented voluntarily to participate in the survey.

Respondent Information
The purpose of this section is to collect general information regarding the respondents such as full name, agency name, agency type, etc.

General Project Information
The purpose of this section is to collect general information regarding the project such as project location, start of the design and construction, funding source, etc.

Characteristics of the project
The purpose of this section is to collect information regarding the project characteristics such as size, main facility systems and elements, etc.

Cost Estimating
The purpose of this section is to collect actual data regarding the cost estimating methods, database, and documentation, etc.

Project Schedule
The purpose of this section is to collect information regarding the project duration, and delays.

Project Risk
The purpose of this section is to collect information regarding the project risk factors.

Change Orders
The purpose of this section is to collect information regarding the time, reason, and cost of design and/or construction changes.

Others
The purpose of this section is to collect any lesson learned about cost estimating of the project.
Before distributing the survey, a pilot survey was conducted by sending the questionnaire to four practitioners who were managers of rural transit facility projects and were introduced to the research team by a member of the RTAP. The research team asked the practitioners to complete the survey in order to ensure: (1) the efficacy of the questionnaire; (2) the clarity of the survey instructions; (3) the efficiency of the survey's format; and (4) the prevention of any errors in the survey. The feedback was used to revise the survey questions.
Next, the research team sent the survey to the personnel of 52 state DOTs and consulting firms who were program managers, transit managers, or consultants of rural transit facilities. The survey was conducted online, and the survey invitations were sent through email. Follow-up survey requests were sent to the same group of people. Despite two follow-up emails, only 26 surveys, belonging to 26 different state DOTs, were returned because of the respondents' difficulty in accessing actual project data.

Preliminary Data Analysis
The information and data retrieved from the 26 collected surveys are shown in the five columns of Table 2: project number, transit facility type, project size based on square feet (sf), baseline budget of design phase, and baseline budget of construction phase. As indicated in Table 2, the project sizes ranged between 30 sf and 75,000 sf, with 16 of the 26 projects (61.5%) exceeding 10,000 sf. Eleven (11) of the 26 projects (42%) were facility types for administration, operations, maintenance, and vehicle storage. The baseline design budgets ranged between $1850 and $2,000,000, with 18 of the 26 (69.5%) projects having design costs that equaled or were greater than $100,000. The baseline construction costs of the projects ranged from approximately $6500 to $30,000,000, with 17 of the 26 (65.5%) having construction costs equal to or greater than $1,000,000.

Descriptive Data Analysis
The results corresponding to the descriptive data analyses for the cost and schedule of design and construction phases are shown in Table 3. The mean value of the actual design cost was $387,544, ranging between $2365 and $2,000,000. The value of standard deviation for the actual design cost was $580,439, indicating a large variation of projects' actual design cost, because the value of the standard deviation indicates how closely or widely each actual cost value spreads around the cost mean value.  Table 3 indicates that the mean value of actual construction cost was $5,278,808, and ranged between $7531 and $40,000,000. The value of the standard deviation for actual construction cost was $10,011,426, which indicates the large variation in projects' actual construction cost.
The actual schedule of the design phase ranged from 2 months to 82 months; the mean value and standard deviation of the actual schedule for the design phase were 20 months and 23 months, respectively. These values illustrate that the projects' actual design phase schedule covers a wide range. Actual construction schedules ranged between 4 months and 53 months, with the mean value and standard deviation being 11 months, indicating a large data variation.

Cost Difference (Overruns/Underruns)
The results of the descriptive data analyses, conducted to determine the absolute difference between design and construction baseline budgets and actual costs of the transit facility projects, are presented in Table 4. This table indicates that the design cost difference ranged between $0 and $103,678, and the mean value and standard deviation of design cost difference were $11,506 and $24,861, respectively. The value of the standard deviation showed large variations in the design cost overruns and/or underruns. As presented in Table 4, the construction cost difference ranged from $0 to $10,000,000. The mean value and standard deviation of the construction cost difference were $598,576 and $1,946,754, respectively, which showed the wide variations in the projects' construction cost difference.
The large variations in the cost overruns and/or underruns of both the design and construction phases indicate that the DOT agencies should focus on accurate cost estimation for rural transit facility projects in order to minimize cost overruns/underruns and improve the projects' success.

Contract Type
This step investigated the contract type associated with the 26 collected surveys. As presented in Figure 4, the contract types of the design phase for rural transit facilities were lump sum, unit price, and cost plus fee. This figure shows that 66% of the contract types were lump sum for the design phase. The contract types for the construction phase were lump sum, unit price, and cost plus fee; most of them (73%) were lump sum. Obviously, lump sum is the dominant contract type in both the design and construction phases of rural transit facility projects. In a lump sum contract, the contractor agrees to complete the project for a fixed price. This contract type has less risk for the contractor when the scope of the project is well-defined [41,42].

Cost Estimating Methods
The utilized design and construction cost estimating techniques were studied and the results are presented in Table 5. This table indicates that roughly 48% of the design and 55% of the construction cost estimating methods were utilization of similar projects. In this cost estimating method, the actual costs of previous similar projects are utilized for estimating the cost of the present project. Accordingly, one or multiple similar projects might be used as a reference. In other words, historical data of similar projects is the foundation and basis of estimating the cost of a project [44]. This method of estimation is not commonly considered accurate, because there is no provision for uncertainties and unknowns [45].
For the design phase, 30% of the projects used hours to design for cost estimating. Other methods employed for design cost estimating were historical percentage of construction cost (9%), architect's estimate (5%), and contractor's estimate (4%). For the construction phase, the cost estimating techniques used were historical bid data (20%), contractor's estimate (10%), combination of similar projects and historical data (5%), architect's estimates (5%), and consultant's estimates (5%).  Nineteen percent (19%) of the contracts were cost plus fee for the design phase, and 12% were cost plus fee for the construction phase. Cost plus fee is a type of contract in which a client pays the actual cost of the delivered service, plus a fixed fee or a percentage of actual cost. In this contract type, the owner is at risk if cost overruns occur [43].

Cost Estimating Methods
The utilized design and construction cost estimating techniques were studied and the results are presented in Table 5. This table indicates that roughly 48% of the design and 55% of the construction cost estimating methods were utilization of similar projects. In this cost estimating method, the actual costs of previous similar projects are utilized for estimating the cost of the present project. Accordingly, one or multiple similar projects might be used as a reference. In other words, historical data of similar projects is the foundation and basis of estimating the cost of a project [44]. This method of estimation is not commonly considered accurate, because there is no provision for uncertainties and unknowns [45].
The transit facility projects, comprised of four Categories (administration, operation, maintenance, and vehicle storage), are referred to as Combination I. The percentage associated with the cost breakdown of each construction system is indicated in Figure 5. For the design phase, 30% of the projects used hours to design for cost estimating. Other methods employed for design cost estimating were historical percentage of construction cost (9%), architect's estimate (5%), and contractor's estimate (4%). For the construction phase, the cost estimating techniques used were historical bid data (20%), contractor's estimate (10%), combination of similar projects and historical data (5%), architect's estimates (5%), and consultant's estimates (5%). The transit facility projects that include two or three types of facilities are referred to as Combination II. The cost breakdown for each construction system of Combination II is illustrated in Figure 6. The percentages of substructure and special construction and demolition of Combination II facilities were similar to those of Combination I facilities. The primary reasons for the mentioned differences are as follows: • A higher percentage of cost breakdown was recorded for building site work, because of the necessity for more mechanical and electrical utilities at the site.

•
Combination I facilities had a higher percentage of cost of interior, equipment, and furnishings than those of Combination II, because a higher number of walls, floors, and ceiling finishes, interior doors, partitions, and furnishings were required.
• Combination II facilities had a higher shell and service cost percentage than Combination I facilities, owing to the fact that their function, maintenance, and vehicle storage typically require additional heating, airflow, conditioning systems, pipework, and electrical constructions to provide effective and reliable assistance. The transit facility projects that include two or three types of facilities are referred to as Combination II. The cost breakdown for each construction system of Combination II is illustrated in Figure 6. The percentages of substructure and special construction and demolition of Combination II facilities were similar to those of Combination I facilities. The primary reasons for the mentioned differences are as follows: • A higher percentage of cost breakdown was recorded for building site work, because of the necessity for more mechanical and electrical utilities at the site. • Combination I facilities had a higher percentage of cost of interior, equipment, and furnishings than those of Combination II, because a higher number of walls, floors, and ceiling finishes, interior doors, partitions, and furnishings were required. • Combination II facilities had a higher shell and service cost percentage than Combination I facilities, owing to the fact that their function, maintenance, and vehicle storage typically require additional heating, airflow, conditioning systems, pipework, and electrical constructions to provide effective and reliable assistance.
furnishings than those of Combination II, because a higher number of walls, floors, and ceiling finishes, interior doors, partitions, and furnishings were required.
• Combination II facilities had a higher shell and service cost percentage than Combination I facilities, owing to the fact that their function, maintenance, and vehicle storage typically require additional heating, airflow, conditioning systems, pipework, and electrical constructions to provide effective and reliable assistance.

Construction Risk Analysis
In this step, the frequency of observed construction risk factors was investigated. As shown in Table 6, soil conditions (23%) and unexpected underground conditions (17%) were the highest construction risk factors. Buried debris, contaminated soil, and

Construction Risk Analysis
In this step, the frequency of observed construction risk factors was investigated. As shown in Table 6, soil conditions (23%) and unexpected underground conditions (17%) were the highest construction risk factors. Buried debris, contaminated soil, and unexpected utilities can increase costs and cause schedule delays of transit facility projects. The cost of risks must be added to the baseline budget of the mentioned projects.

Contingency Estimating
The results demonstrated that 75% of studied rural transit facility projects used a percentage of construction cost for estimating the contingency value. The percentage of contingency ranged from 4% to 15%. The median contingency was roughly 10%.

Estimation of Design and Construction Cost
A regression statistical technique was generated to predict the design and construction costs based on the project size (sf) of the rural transit facilities. Although this study was exploratory, the authors attempted to develop a reliable and valid model for both design and construction costs. The adopted criteria were as follows:

•
The sample size was limited to data and information of 26 rural transit projects. The research team was aware that the small number might negatively affect the models' predictability; however, there were not enough solid criteria associated with the required number of data to develop a regression model. Some researchers and authors, such as Roscoe [46], followed a rule of thumb, and assumed that 10 samples were required for each dependent variable. The rule of thumb is also commonly used for exploratory studies [47].

•
The coefficient of determination was considered for the developed models to evaluate how accurately the regression models could predict the design and construction costs for rural transit facilities. If there was a strong relation between the independent variable and the dependent variable(s), the coefficient of determination became 1.
On the contrary, if there was no relation between the independent variable and the dependent variable(s), the coefficient of determination became 0. The authors made an effort to develop regression models for design and construction costs with a coefficient of determination closer to 1.
A straight-line regression model was generated for the design cost with 90% confidence level, as shown in Table 7. In this table, design cost and project size (sf) were presented by Z and X, respectively. The result of the straight-line regression model was obtained as Z = 31.64 X (X > 0). This equation clearly shows that the design cost of a rural transit facility is approximately 32 times its size, based on square feet. The result of p-value was obtained as 0.0001 < α = 0.1. Additionally, the R-square was recorded as R-square = SS mdeol SS total = 1.6421×10ˆ13 1.8027×10ˆ13 = 0.91. The result of R-square shows that the straight-line regression model is a good fit for the normalized design cost data. The results of a generated straight-line regression model for prediction of construction cost are presented in Table 8. Similar to the design cost, the construction cost and project size (sf) are Y and X, respectively. The straight-line regression model was obtained as Y = 172.70 X (X > 0) with 90% confidence level. The equation clearly shows that the construction cost of rural transit facilities is roughly 173 times their size, based on square feet. Additionally, the p-value is recorded as <0.0001 < α = 0.1. The R-square was attained as R-square = SS mdeol SS total = 1.3637×10ˆ14 1.4494×10ˆ14 = 0.94. The R-square suggests that the straight-line regression model fits for the standardized construction cost of rural transit infrastructure reasonably well.

Discussion
The key components of rural transit facilities, design and construction cost estimating methods, construction risk factors, and design and construction contract types were identified in this study. The nature of rural transportation projects is very different from that of large transportation projects, which makes it more difficult to establish a clear and detailed overall scope and accurately estimate the cost of the project during its early stages [11,12]. As these are different types of facilities than the ones that we have in urban areas, the models that are developed for the urban areas might not be accurately applicable for these types of projects in rural areas.
Therefore, two regression models were generated to predict the design and construction costs. The regression model was adopted to develop design and construction costs of rural transit facilities inspired by [47][48][49]. A straight-line regression model was generated for design cost and construction cost and is presented in Tables 7 and 8. The developed equations help to estimate the actual cost and overrun in construction and design of transit facilities in rural areas. This enhances the proper planning, which helps in allocating sufficient contingency so the projects will not face any funding issues or will not be suspended.

Conclusions
As there is insufficient information and data associated with rural transit facilities, the ultimate aim of this study was to provide a baseline estimates for both design and construction costs of rural transit projects. In addition, this study identified the key components of rural transit facilities, identified design and construction cost estimating methods, determined construction risk factors, assessed the cost of each construction system, and identified design and construction contract types associated with rural transit facilities.
Two regression models were developed to predict the design and construction costs. It was concluded that size is a significant predictor in determining the conceptual design and construction costs of rural transit facilities. Moreover, the results demonstrated that the "similar projects' cost technique" was the most frequently used method in which the actual cost of previous similar projects are utilized for estimating the cost of the present project. The lump sum contract was employed most often for both the design and construction phases. In this contract type, the contractor agrees to complete the project for a fixed price. The data collected from interviews were used to identify risk factors, and their frequencies were determined using an online survey. The most frequently encountered construction risk factors were soil conditions and unexpected underground conditions. To manage project risks, contingency is projected as a percentage of construction cost. The contingency percentage ranges provided by interview respondents and survey results served as a guide for establishing the default contingency range for the cost estimating prototype tool. Lastly, a straight-line regression model was generated to estimate both the design and construction costs based on the project size of the rural transit facilities. The findings of this study provide a beneficial baseline estimate for the design and construction costs in the early stages of the rural transit facility projects, and will assist DOTs in accurately analyzing bidders' proposed cost estimates, enabling them to select the one that most effectively minimizes the potential for cost overruns.
Although this research was exploratory in nature and somewhat limited by the relatively small number of interviews and completed surveys, the authors made every effort to provide valid and reliable outcomes. It is, however, recommended that future studies query a larger number of DOT representatives to obtain more comprehensive results. The models were not generated as a replacement for estimates of design and construction costs associated with rural transit facilities, which require detailed construction cost techniques. In addition, this study relied on information pertaining to U.S. rural transit facility projects and lacks cross-professional, cross-organizational, and cross-national studies from rural transit facility projects worldwide.