Building information modeling implementation strategies for public infrastructure projects in emerging markets: The case of Ethiopia

Abstract

The purpose of this study is to investigate the critical BIM adoption strategies for public construction projects in developing countries to enhance successful BIM implementation schemes in the AEC industry. Initially, an empirical study was conducted using a structured questionnaire survey which is first validated though a pilot test. Then, the study investigated the discrepancy of professional groups on the perception of critical adoption strategies using different statistical tools. Further, factor analysis was conducted to identify potential latent factors associated with public construction projects. The result indicates that the top ranked BIM adoption strategies are strategic & adequate IT infrastructure, availability of standards and guidelines for BIM adoption, and government policy. Similarly, six latent constructs were developed from the factor analysis: organization, application & tools, market, information management, project, and process. The study highlighted critical strategies that are deemed most important to various professionals and business owners, in an effort to improve BIM adoption in the developing countries. The novelty of this study is in relation to BIM implementation strategies towards infrastructure construction projects in developing countries, particularly in the east African region. This paper can also be particularly beneficial to construction firms involved in public construction projects to facilitate the implementation of BIM in their own organizational nature. Moreover, the cross-country comparative analysis provides a global perspective to improve the gap in the understanding and adoption of BIM in the Sub-Saharan African region.

Keywords:

• BIM Adoption
• construction projects
• developing countries
• factor analysis

1. Introduction

The efficient delivery of construction projects require sufficient budget, effective design, and innovation and collaboration among parties (Asma et al., 2020; Kassem & Succar, 2017; Yin et al., 2019). One of the factors that enhances the information flow, communication between stakeholders, and overall success of construction projects is the extent to which innovation is used in different project phases (Hochscheid & Halin, 2019; X. P. Ma et al., 2020).

Adopting BIM and innovative procurement methods such as integrated project delivery method (IPD) in construction projects has been effective in managing workflow and enhancing productivity of the built environment (Chang et al., 2017; Piroozfar et al., 2019). In this respect, studies highlighted that developed countries/regions such as United Kingdom, Germany, United States, Australia, and Hong Kong have advanced their BIM adoption strategies in the AEC industry (Chan et al., 2019; Liao & Teo, 2017; Won & Cheng, 2017). These countries/regions have been developing their own BIM adoption policies and strategies to encourage and facilitate BIM adoption process across their respective markets in the past decade (Derrick & Mohamed, 2011; Le et al., 2020).

In recent years, however, BIM adoption studies in developing countries have expanded rapidly (Olanrewaju et al., 2020). Most studies focuses on exploring the awareness of BIM by professionals (Nasila & Cloete, 2018) and investigating the application, challenges, and barriers of BIM implementation in the construction industry (Ullah et al., 2019). A study by Ngowtanasawan (2017) reported that factors including relative advantage and compatibility of BIM software’s influence the successful adoption of BIM in the AEC industry. Whereas, P. Ma et al. (2020) reveal cultural barriers, in-country construction policies, and attention given to public and private projects influence the successful adoption of BIM in the construction sector.

Similarly, prior studies reported various challenges associated with BIM adoption in construction projects,; including unreliable technological infrastructures and unavailability of professional skills (Ozorhon & Karahan, 2017). However, there is still a dearth of studies and methodologies in developing countries that focuses on BIM adoption in relation to organization and project characteristics.

In this regard, the purpose of the present study is to identify and explore the critical BIM adoption strategies for public construction projects. The study also aims to investigate the perception of each participant groups (client, consultant, contractor, and academia) on the potential BIM adoption strategies. In addition, the paper discusses cross-country comparative analysis between developing countries. Moreover, the study also addresses the limitation of existing BIM adoption strategies in developing countries by highlighting potential practical implications and recommendations on how the findings can help speed up the BIM adoption process in public construction projects.

The findings of this study are believed to provide significant research data for construction firms to give attention for successful BIM adoption in organizations as well as project levels. This study will also contribute to policy makers and government regulators in developing countries to encourage BIM adoption in the construction sector.

2. Overview of BIM in Ethiopia

The Ethiopian construction industry is the second largest industry exceeded by agriculture in term of capital spending, the resources it consumes, and the number of jobs it creates for skilled and unskilled workers (Ayalew et al., 2016). However, even though the industry is one of the biggest, studies show that the construction industry is still tied with the traditional contractual and delivery methods, weak communication and collaboration between parties involved in projects, claims and disputes, and low productivity (Ayalew et al., 2016; Gebremariam & Dinku, 2018).

To address the above challenges and improve the performance of construction projects, the Ethiopian government took an initiative to encourage BIM adoption in big public construction projects around the country. However, the limitation of studies regarding BIM adoption and BIM understanding by professionals and stakeholders become a hindrance to implement different strategies and policies and capacitate construction firms. Nuramo and Haupt (2016) reported that low performance, fragmentation of the construction sector, and lack of BIM researches and trainings in universities have been a challenge for a wide range BIM of adoption in infrastructure projects. Hence, the current study aims to address the aforementioned BIM implementation focused studies in developing countries, particularly in Ethiopia by suggesting fact-based strategies for various stakeholders.

3. Literature review

Systematic literature review was conducted to identify the potential BIM adoption strategies using two of the biggest databases: Scopus, and Google Scholar (Ahmed & Kassem, 2018). Keywords such as, “BIM Adoption Strategies,” “BIM Adoption in Developing Countries,” and “Critical Success Factors of BIM adoption” were used in the search. Then, the authors filtered 44 papers published in year 2016 and later. From this, 13 publications were further taken for the questionnaire development based on their methodological approaches and geographical location and the remaining 31 for analysis throughout the paper.

3.1. BIM adoption strategies in construction projects

Successful adoption of BIM in construction projects requires an unprecedented effort and approach by professionals, business partners, and government support across the industry (Ngowtanasawan, 2017; Seyis, 2019). Recent developments pinpoint the BIM adoption strategies within different contexts: from industry, organization, and project levels (Chan et al., 2019; P. Ma et al., 2020).

Hochscheid and Halin (2019) studied the factors that influence the BIM adoption and diffusion process in the construction sector. Similarly, Zhou et al. (2019) explored the potential applicability of the advanced lessons and BIM experiences cultivated from different countries to the development of concise BIM adoption strategies in China. Whereas, in the context of organizations, Liao and Ai Lin Teo (2018) explored organizational working culture and experts’ role as BIM adoption drivers through people management in organizational structures and attributes. Similarly, the process of successful BIM adoption strategies in terms of capacitating construction firms was explored by Ahn et al. (2016). More so, Asma et al. (2020) investigated the governing criteria of client firm in BIM based construction projects.

In recent years, studies are emphasizing on exploring strategies that focus on BIM adoption in construction projects (Abdulmumin et al., 2020; Le et al., 2020). For instance, a study by Ozorhon and Karahan (2017) examined the critical factors of adopting BIM in construction projects located in Turkey, using a questionnaire survey. P. Ma et al. (2020) studied the BIM adoption strategies and diffusion techniques in the Chinese construction projects. Moreover, Aljobaly and Banawi (2020) evaluated the capacity of Saudi Arabia construction firms in adopting BIM to construction projects, using three major BIM attributes. The authors argued that cultural contextualities significantly affect BIM adoption in construction projects. Thus, having a country-wise BIM adoption policy with an emphasis of global adoption strategy is an important element.

Considering the majority of studies mainly focus on BIM strategies from market level and the fact that more recent advancements emphasize project conditions, the present study aims to analyze and enhance potential BIM strategies from project perspectives in developing countries. Moreover, prior studies in developing countries donot differentiate between public and private sector. In this respect, this study fills the gap in the literature by exploring critical BIM adoption strategies in the context of public construction projects. The critical strategies identified from the literature are outlined in Table 1.

Table 1. Summary of BIM adoption strategies from literature review

Item	BIM Adoption Strategical Factors	References
BS1	Strategic & adequate IT infrastructure	(Olawumi & Chan, 2019), (X. P. Ma et al., 2020) (Liao & Teo, 2017; Ozorhon & Karahan, 2017; Yin et al., 2019; Zhou et al., 2019),
BS2	Availability of financial resources for BIM adoption	(X. P. Ma et al., 2020) (Aljobaly & Banawi, 2020; Chan et al., 2019; Ozorhon & Karahan, 2017),
BS3	Effective senior leadership	(Chen et al., 2016; Liao & Ai Lin Teo, 2018; Olanrewaju et al., 2020; Olawumi & Chan, 2019; Yin et al., 2019)
BS4	Availability of technical support for employees	(X. P. Ma et al., 2020) (Ismail et al., 2017; Liao & Teo, 2017; Olanrewaju et al., 2020; Zhou et al., 2019),
BS5	Collaborative working environment in firms	(Chan et al., 2019; Liao & Ai Lin Teo, 2018; Liao & Teo, 2017; Olawumi & Chan, 2019; Ozorhon & Karahan, 2017)
BS6	Government policy	(Olawumi & Chan, 2019), (X. P. Ma et al., 2020) (Aljobaly & Banawi, 2020; Ismail et al., 2017; Liao & Ai Lin Teo, 2018),
BS7	Availability of quality control and Specifications	(Chan et al., 2019; Chen et al., 2016; Olanrewaju et al., 2020; Ozorhon & Karahan, 2017; Yin et al., 2019)
BS8	Appropriate legal parameters	(Ahn et al., 2016; Aljobaly & Banawi, 2020; Ismail et al., 2017; Ozorhon & Karahan, 2017; Yin et al., 2019)
BS9	Sufficient standards and guidelines for BIM adoption	(Olawumi & Chan, 2019) (X. P. Ma et al., 2020), (Chan et al., 2019; Liao & Teo, 2017; Zhou et al., 2019),
BS10	Extent of BIM awareness	(Ahn et al., 2016), (X. P. Ma et al., 2020) (Aljobaly & Banawi, 2020; Liao & Ai Lin Teo, 2018; Ozorhon & Karahan, 2017),
BS11	Coordination among stakeholders	(Ismail et al., 2017; Liao & Ai Lin Teo, 2018; Liao & Teo, 2017; Olawumi & Chan, 2019; Yin et al., 2019)
BS12	Highly competent consultancy service	(X. P. Ma et al., 2020) (Liao & Teo, 2017; Olanrewaju et al., 2020, 2020; Ozorhon & Karahan, 2017),
BS13	Sufficient interoperability data & information	(Ahn et al., 2016; Chen et al., 2016; Ismail et al., 2017; Olawumi & Chan, 2019; Ozorhon & Karahan, 2017)
BS14	Regular communication between parties	(Olawumi & Chan, 2019), (X. P. Ma et al., 2020) (Ismail et al., 2017; Ozorhon & Karahan, 2017; Yin et al., 2019),
BS15	Appropriate choice of delivery & contract types	(Chen et al., 2016), (X. P. Ma et al., 2020) (Aljobaly & Banawi, 2020; Liao & Ai Lin Teo, 2018; Olanrewaju et al., 2020),
BS16	Managing risks associated with BIM	(Chen et al., 2016), (X. P. Ma et al., 2020) (Liao & Ai Lin Teo, 2018; Ozorhon & Karahan, 2017),
BS17	Clear roles & responsibility of BIM usage	(Olawumi & Chan, 2019), (X. P. Ma et al., 2020) (Liao & Ai Lin Teo, 2018; Ozorhon & Karahan, 2017; Yin et al., 2019),
BS18	Managing workflow throughout project life cycle.	(Aljobaly & Banawi, 2020; Ismail et al., 2017; Liao & Teo, 2017; Olawumi & Chan, 2019; Yin et al., 2019)

4. Methodology

The purpose of this study is to investigate the critical BIM adoption strategies that focus on public construction projects in developing countries. The empirical data were collected from professionals working in the Ethiopian construction sector and analyzed through factor analysis (Figure 1).

Figure 1. Systematic Literature Review Flow Chart.

4.1. Reasons for using Factor Analysis (FA)

In recent years, factor studies have been emerged as an important analytical tool in construction and innovation studies (Chang et al., 2017; Husain et al., 2018). Prior studies adopted factor analysis to investigate BIM adoption and diffusion in the construction sector (Chen et al., 2019; Ngowtanasawan, 2017). A recent study by Derrick J.Z. & Mohamed F.E., (2018) used factor analysis to explore the barriers related to adopting BIM in prefabricated construction sector in China. Whereas, Chang et al. (2017) applied factor study to investigate the behavior of individual in relation to BIM usage and application.

Literature have highlighted the importance of applying factor analysis in BIM adoption strategy studies. The reason being (1) factor analysis is helpful to cluster observed variables in to a significant smaller set factor variable, which in turn helps to ease the analysis process (Liao & Ai Lin Teo, 2018), and (2) it allows for the examination of latent constructs that are measured by several observed variables in the BIM adoption strategies (Ahmed & Kassem, 2018). Thus, based on the above reasons, the present study aims to examine the critical BIM adoption strategies in public construction projects based on a comprehensive empirical factor analysis.

4.2. Identification of BIM adoption strategies from review of literature

The current study conducted a systematic review of relevant literature to gain a broad view of the strategies from different countries and avoid biased and inaccurate empirical analysis. In this regard, a total of 18 critical BIM adoption strategies were summarized taking into account recent studies conducted from specific project perspectives (Table 1).

4.3. Data collection

A 5-point Likert scale ranging from 1 = not critical, 2 = not quite critical, 3 = fairly critical, 4 = very critical, and 5 = extremely critical, was adopted in this study to get the perception of professionals that consisted of clients, consultants, contractors and academia and are currently working in public construction projects across Ethiopia. Prior studies suggest that a 5-point Likert scale has been proven to be useful for a reliable data analysis in similar investigations around the world (Ozorhon & Karahan, 2017; X. P. Ma et al., 2020).

The draft questionnaire then sent to four experienced professionals (2 in academia and 2 in the industry) for content validation. The content validation helps to validate certain BIM adoption strategies taken from the literature against the cultural and other related factors and the overall language usage in the questionnaire. Figure 2 outlines the methodological flowchart of the study.

Figure 2. Methodological Flow Chart.

4.4. Pilot study

The current study adopted a two-stage data collection technique. The first stage of the data collection was based on a pilot study consisting of 12 experts working in various construction projects in and around the city of Bahir Dar. The pilot study was conducted as a pretest of the draft questionnaire for the purpose of checking overall language usage, assessing the average time needed to fill out the questions, and checking the compatibility of statistical tools against the questionnaire format for the main data collection (second stage) and analysis.

4.5. Data analysis

The analysis of data in the current study comprised of different statistical methods, including Kolmogorov-Smirnov, Cronbach’s alpha, chi square test, spearman’s rank correlation test, Mann-Whitney, and Kaiser-Meyer-Olkin test. These statistical analysis techniques provide an all-inclusive understanding of the key data collected from various professionals across the Ethiopian construction sector. Using multiple data analysis techniques is beneficial to understand and conceptualize the data from different respondent groups perspectives as well as to triangulate the data as part of the validation process (Ding et al., 2014; Enegbuma et al., 2014). A statistical software package IBM® SPSS® Statistics 23 was used to analyze the data.

4.6. Demography of respondents used in this study

After the completion of the pilot test, the revised questionnaire was then sent out to 181 experts randomly that are working in the field of architecture, civil engineering, and construction using face to face meetings and emails. A total of 110 questionnaires were filled and returned back, which implies a 61% overall response rate. From these, after discarding incomplete questionnaires, 96 valid responses were taken for further analysis. In order to enhance the validity and reliability of responses, experts who has a 5 year or more professional experience in public construction projects in the Ethiopian construction sector were considered for the study. Table 2 describes the demographic profile of participants.

Table 2. Demographic profile of participants

Profile	Frequency	Percentage
Gender
Male	77	80%
Female	19	20%
Experience (Years)
6–10	52	54%
11–15	34	35.5%
>15	10	10.5%
Education
BSc	32	33.3%
MSc	64	66.7%
Organization (Firm)
Client	13	13.5%
Consultant	41	43%
Contractor	27	28%
Academia	15	15.5%

5. Findings

This section summarizes the findings of data collected from professionals working in public construction projects.

5.1. Normality and reliability tests

Initially, Kolmogorov-Smirnov normality test was conducted to check whether the data are normally distributed or not. The result indicated that the variables violated the assumption of normal distributions with a confidence interval of 95%. Thus, the null hypothesis is rejected, and it is concluded that non-parametric statistical tests will be employed for further analysis.

Similarly, Cronbach’s alpha (α) test is used to test the reliability (internal consistency) of data collected through a questionnaire survey. The coefficient of α ranges between 0 and 1. The rule is, if the α-value is greater than or equal to 0.7, the measurement is said to be acceptable (Hayter, 2012). In this study, an overall Cronbach’s alpha value of 0.715 was recorded. Hence, the α value is in acceptable range, and the result is reliable.

5.2. Mean score rank

Mean score method is adopted to rank the responses of professionals working the Ethiopian public construction sector. Mean score ranking (M) is computed based on the mean (average) values associated with each factor/attribute (Chan et al., 2019). In this study, the mean value ranges between M = 4.48 and M = 3.23. Based on the result, “Strategic & adequate IT infrastructure” (M = 4.48), “Sufficient standards and guidelines for BIM adoption” (M = 4.47), and “Government policy” (M = 4.42), perceived as the top critical strategies of BIM implementation in Public construction projects. In contrast, strategies such as “Managing risks associated with BIM” (M = 3.29), “Sufficient interoperability model data and information” (M = 3.28), and “Managing workflow throughout project life cycle” (M = 3.23), has the lowest mean values and perceived as the least critical BIM adoption strategies.

Furthermore, as it is shown in Table 3, the result illustrates that there is a similarity and consensus in the perception of the top 5 rankings by all participant groups (client, consultant, contractor, and academia). This is also in line with prior studies (Liao & Teo, 2017; Olawumi & Chan, 2019). Table 3 illustrates the mean score results of all participant groups.

Table 3. Mean score ranking for BIM adoption strategies

Item	BIM Adoption Strategies	All Participants		Client		Consultant		Contractor		Academia
		Mean	Rank	Mean	Rank	Mean	Rank	Mean	Rank	Mean	Rank
BS1	Strategic & adequate IT infrastructure	4.48	1	4.54	3	4.46	3	4.41	2	4.60	2
BS9	Sufficient standards and guidelines for BIM Adoption	4.47	2	4.61	2	4.59	1	4.30	4	4.47	3
BS6	Government policy	4.42	3	4.62	1	4.34	6	4.48	1	4.32	8
BS12	Highly competent consultancy service	4.40	4	4.38	8	4.58	2	4.09	8	4.40	5
BS10	Extent of BIM awareness	4.35	5	4.52	5	4.39	5	4.00	9	4.73	1
BS11	Coordination among stakeholders	4.33	6	4.46	6	4.24	7	4.33	3	4.46	4
BS5	Collaborative working environment in firms	4.31	7	4.38	7	4.41	4	4.11	6	4.33	7
BS15	Appropriate choice of delivery & contract types	4.25	8	4.36	9	4.15	8	4.26	5	4.39	6
BS8	Appropriate legal parameters	4.09	9	4.53	4	3.98	10	4.10	7	4.00	10
BS17	Clear roles & responsibility in BIM usage	3.77	10	3.46	14	4.02	9	3.33	14	4.13	9
BS3	Effective senior leadership	3.56	11	3.62	12	3.66	11	3.41	12	3.53	11
BS7	Availability of quality control and specifications	3.46	12	3.69	10	3.54	14	3.37	13	3.20	15
BS4	Availability of technical support for employees	3.36	13	3.23	15	3.56	13	3.52	10	2.67	18
BS14	Regular communication between parties	3.32	14	3.68	11	3.41	15	3.22	15	2.93	17
BS2	Availability of financial resources for BIM adoption	3.31	15	2.92	18	3.61	12	3.07	16	3.27	14
BS16	Managing risks associated with BIM	3.29	16	3.15	17	3.22	17	3.44	11	3.33	13
BS13	Sufficient interoperability model data & information	3.28	17	3.15	16	3.37	16	3.06	17	3.52	12
BS18	Managing workflow throughout project life cycle	3.23	18	3.62	13	3.21	18	3.05	18	3.19	16

5.3. Ranking of agreement within participant groups

Kendall’s coefficient of concordance was used to measure the level of agreement within each participant group on the ranking of BIM adoption strategies. The value of Kendall’s coefficient of concordance (W) ranges from 0 to 1, where 0 means there is no agreement (consensus) and 1 stands for there is perfect agreement level (Hayter, 2012). However, if the number of items to be ranked is greater than or equal to 7, chi square test (x²) shall be deployed instead (Chan et al., 2019). In this case the hypothesis to be tested is:

The null hypothesis (Ho) is as follows: There is no relationship between the sets of rankings within each expert groups.

Based on the analysis, the Kendall’s coefficient of concordance (W) for all experts, client, consultant, contractor, and academia are 0.296, 0.483, 0.273, 0.301, and 0.419, respectively, which reveals that all the values are closer to “0” and there was a disagreement between the responses. Similarly, the significant values for all experts is resulted as 0.000, which is less than the allowable significance level = 0.05. Thus, the null hypothesis will be rejected.

Correspondingly, the chi-square test was also conducted since the number of items to be ranked were greater than seven. So, the computed values are for all respondents (483.41), client (106.83), consultant (190.31), contractor (138.12), and academia (106.93). The allowable critical value (from chi square table) with degree of freedom (df) = 17 and a “ρ value” of 0.05 is read as 27.59. All the above values are greater than the critical value of 27.59, which means that the null hypothesis will be rejected as well. Therefore, the result can be interpreted as there was a relationship within the sets of rankings of each expert groups.

5.4. Ranking agreement between participant groups

Spearman’s rank correlation coefficient (r_s) was adopted to test the extent of agreement between expert groups (Olawumi & Chan, 2018). The value of r_s ranges between − 1 and + 1. A higher r_s negative/positive value shows stronger linear relationship between rankings. In contrast, when the value of r_s is “0”, the result reveals no relationship between sets of rankings. If the computed value of r_s is lower than a predetermined significance level (i.e 5%), the null hypothesis (Ho) will be rejected, which means there is statistically significant correlation between the rankings of expert groups.

The null hypothesis (Ho) in this is as follows: There is no correlation on the sets of rankings between expert groups.

After computation, the values of r_s at the significant level of 0.05; a) between client and consultant, b) client and contractor, c) client and academia, d) consultant and contractor, e) consultant and academia, and f) contractor and academia is 0.761 (ρ = 0.000), 0.822 (ρ = 0.000), 0.713 (ρ = 0.001), 0.779 (ρ = 0.000), 0.870 (ρ = 0.000), and 0.721 (ρ = 0.001), respectively. All the above “ρ” values are lower than the significance level of 0.05, and the null hypothesis (H0) will be rejected. Thus, there is a correlation between the rankings of all pairs of expert groups.

5.5. Statistical differences between groups

The Mann-Whitney U test is employed to determine any statistically significant differences or divergences in the mean values of similar BIM adoption strategies ranked by a pair of expert groups. If the computed ρ value is less than the allowable significance level (0.05), then the null hypothesis (Ho) will be rejected (Chan et al., 2019).

5.5.1. The null hypothesis (Ho) is: there is no statistical difference between the mean values of items filled by expert groups

The “U” analysis shows that for the pair of expert groups, client vs consultant, client vs contractor, client vs academia, contractor vs academia and consultant vs academia, all ρ values of BIM adoption are greater than 0.05, and the null hypothesis will be rejected (meaning there is a statistical difference between the mean values of items filled by expert groups).

In contrast, for the case of consultant vs contractor groups, all items have a ρ value of less than the allowable 0.05 except No. 2, 12, and 17, where the null hypothesis will not be rejected (Table 4).

Table 4. Mann - whitney test output for consultants vs contractors

No.	Mean Rank		Mann-Whitney U	Z Value	ρ value	Decision (Ho)
	Consultant	Contractor
BS1	34.54	34.44	552.000	−0.02	0.98	Accept
BS2	38.95	27.74	371.000	−2.41	0.02	Reject
BS3	36.24	31.85	482.000	−0.94	0.35	Accept
BS4	35.12	33.56	528.000	−0.34	0.73	Accept
BS5	37.30	30.24	438.500	−1.58	0.11	Accept
BS6	33.88	35.44	528.000	−0.35	0.72	Accept
BS7	35.21	33.43	524.500	−0.39	0.70	Accept
BS8	32.29	37.85	463.000	−1.21	0.23	Accept
BS9	36.85	30.93	457.000	−1.37	0.17	Accept
BS10	37.48	29.98	431.500	−1.67	0.10	Accept
BS11	32.84	37.02	485.500	−0.93	0.35	Accept
BS12	38.15	28.96	404.000	−2.09	0.04	Reject
BS13	36.98	30.74	452.000	−1.33	0.18	Accept
BS14	36.11	32.06	487.500	−0.86	0.39	Accept
BS15	33.30	36.31	504.500	−0.66	0.51	Accept
BS16	32.22	37.96	460.000	−1.25	0.21	Accept
BS17	40.50	25.39	307.500	−3.29	0.00	Reject
BS18	35.78	32.56	501.000	−0.72	0.47	Accept

5.6. Factor analysis

Factor analysis (FA) also known as principal component analysis is used to test the principal component of BIM adoption strategies and relationship of variable factor loadings. FA is a statistical procedure which is used to obtain a simplified structure for a large data set (Chang et al., 2017). It is employed when various data measurements are analyzed on a certain item/factor (Liao & Teo, 2017). The lower limit for a factor loading value is 0.5. P. Ma et al. (2020) emphasized the importance of FA and determination of correlation coefficients for studies related to BIM implementation and the area of construction management in general.

Two rounds of factor analysis using varimax rotation method was conducted for BIM adoption strategies in Ethiopia. In the first round, both BS10 and BS11 had a higher component factor loading value of 0.468 and 0.361, respectively. Thus, since both strategies were less than 0.5 (the lower threshold value), they were discarded for further analysis. In contrast, after eliminating both strategies, the second round of FA analysis result shows an acceptable value for all factor loadings.

As shown in Table 5, the Kaiser-Meyer-Olkin test result also reveals an acceptable value of 0.637; any value larger than 0.6 is considered as good (Chang et al., 2017). Similarly, the ρ value of Bartlett’s test result is 0.000. The rule is that any ρ value less than 0.001 is considered to be acceptable (Chan et al., 2019). Thus, the test result is acceptable.

Table 5. KMO & Bartlett test result for BIM adoption strategy

Test Type	Result
Kaiser-Meyer-Olkin measure of sampling adequacy	0.637
Bartlett’s test of sphericity
Approx. chi-square	339.44
df	120
Sig.	0.000

The screen plot indicates the eigenvalue (column) result output of BIM adoption strategies against the component number (row). As shown in Figure 3, the plot line is somewhat flat except the first four items, meaning each successive component is accounting for smaller and smaller amounts of the total variance. The rule is that components with an eigenvalue lower than 1, considered less variance when compared to the initial variable.

Figure 3. Screen plot SPSS output.

Moreover, this study further analyzed the total variance explained, as summarized in Table 6. The initial eigenvalue result reveals a 6.332% of more variance explained and is 63.499% of the total variance explained is reasonable for six components.

Table 6. Eigen value output for components (total variance explained)

Component	Initial Eigenvalues			Rotation Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	2.943	18.394	18.394	2.547	15.919	15.919
2	2.357	14.734	33.129	1.891	11.820	27.739
3	1.429	8.932	42.061	1.714	10.711	38.450
4	1.289	8.055	50.116	1.447	9.045	47.494
5	1.128	7.051	57.166	1.328	8.299	55.794
6	1.013	6.332	63.499	1.233	7.705	63.499

Table 7 points out the rotated component matrix using Varimax. The rule is that item component values greater than 0.5 are considered to be acceptable in BIM adoption related studies (X. P. Ma et al., 2020). The values labeled in bold are all greater than 0.5, which in turn are acceptable. Based on the analysis, six latent factor components were identified, namely, Organizational, Application & Tools, Market Related, Information Management, Project, and Process (Table 7).

Table 7. Rotated component matrix using varimax

BIM Adoption Strategies	Components
	Organizational	Application & Tools	Market Related	Information Management	Project	Process
Availability of financial resources for BIM adoption	0.912	0.135	0.074	0.102	−0.009	−0.051
Effective senior leadership	0.870	0.140	0.079	0.055	−0.115	0.005
Availability of technical support for employees	0.784	0.010	−0.062	−0.054	0.222	−0.032
Sufficient interoperability model data & information	−0.105	0.690	0.028	0.382	0.346	−0.046
Highly competent competence consultancy	0.055	0.685	0.099	−0.108	−0.133	−0.006
Regular communication b/n parties	0.075	0.601	0.124	0.085	0.269	0.250
Strategic & adequate IT infrastructure	0.224	0.589	−0.133	0.119	−0.112	−0.034
Appropriate legal parameters	−0.022	−0.002	0.824	−0.106	0.083	−0.079
Managing risks associated with BIM	0.012	0.276	0.668	0.180	−0.057	0.164
Government policy	0.066	−0.102	0.546	0.216	0.087	0.010
Managing workflow throughout project life cycle	−0.090	0.266	0.186	0.722	−0.109	0.053
Clear roles & responsibility in BIM usage	0.142	−0.024	0.072	0.692	0.185	0.033
Availability of quality control and Specifications	−0.073	0.113	0.327	0.121	0.768	−0.087
Collaborative working environment in firms	0.434	−0.142	−0.268	−0.018	0.622	0.173
Sufficient standards and guidelines for BIM adoption	−0.175	0.098	0.090	−0.161	0.031	0.796
Appropriate choice of delivery & contract types	0.125	−0.027	−0.062	0.346	−0.028	0.674

Organizational related factor is the first component identified in the analysis. This component is related to organization’s structure, resource standing, and competency (Ozorhon & Karahan, 2017). In this regard, the organizational related factors comprised of three strategies; Availability of financial resources for BIM adoption (0.912), Effective senior leadership (0.870), and Availability of technical support for employees (0.784).

The second component on the Varimax rotated matrix is Application & Tools. It is related to the extent of software and hardware usage during each project stage and the corresponding communication media used among experts and as well as parties. The Application & Tools component includes Sufficient interoperability model data & information (0.690), Competence consultancy (0.685), Regular communication b/n parties (0.601), and Strategic & adequate IT infrastructure (0.589). In addition, the third one is Market Related component. It includes BIM adoption strategies such as Appropriate legal parameters (0.824), Managing risks associated with BIM (0.668), and Government policy (0.546). Market strategies focus on the role of government in devising BIM policies and encouraging BIM adoption across the construction industry.

Similarly, Information Management component involves the flow of information and work accountability in relation to BIM usage in construction projects (Ismail et al., 2017). Information Management component involves strategies: Managing workflow throughout project life cycle (0.722), and Clear roles & responsibility in BIM usage (0.692). Moreover, Project component is the fifth component in the Varimax rotated component matrix table. It is comprised of two strategies: Availability of quality control and Specifications (0.768) and Collaborative working environment in firms (0.622). Project includes institutional working structure and activities with respect to coordination and quality assurance.

Lastly, the six component is outlined as Process. This component details the process within the choice of contract types and guidelines and the compatibility of project delivery systems in BIM projects (Olawumi & Chan, 2019). Both Standards and guidelines for BIM (0.796), and Appropriate choice of delivery systems and contract types (0.674) were identified as a Process component.

6. Discussion and practical implications

The first stage of the results section was to conduct a mean score analysis for the purpose of ranking the critical BIM adoption strategies derived from the systematic literature review. Based on the findings, strategic & adequate IT infrastructure is the first BIM implementation driver. The extent of development of IT and the availability of related infrastructures directly affects BIM adoption in developing countries (Hamma-Adama et al., 2018). In Nigeria for instance, studies highlighted that insufficient infrastructure is the major hinderance of BIM adoption in projects (Olanrewaju et al., 2020). To overcome the problem, these developing nations should consider the involvement of private sector in the development of IT infrastructure, especially in the Sub-Sharan African region. Similarly, the experts outlined that the government’s involvement and initiative in devising policies and sufficient standards is vital to encourage BIM adoption. To support the above argument, the Ethiopian Project Management Institute (ECPMI) designed BIM adoption roadmap in 2016, with the aim of enhancing professionals BIM awareness, and encourage universities to include BIM courses in their curriculums (Assefa, 2019).

The second stage of the results section focuses on evaluating the perception of different expert groups on BIM adoption strategies. In this regard, the Mann-Whitney test result indicated that except for consultant Vs contractor and academia Vs contractor groups, the rest of pair of expert groups agreed on the rankings of critical adoption strategies. This is also evident in the current Ethiopian construction setting where design-bid - build project delivery method creates a hole for a fragmented industry and also allows major parties to differ on project goals (Gebremariam & Dinku, 2018).

In the third stage, the authors examined the FA of the factors associated with BIM strategies in public construction projects. FA is particularly important to pinpoint and prioritize critical BIM adoption strategies by providing clustering mechanisms (Ozorhon & Karahan, 2017). In this study, the FA results highlighted six latent factor components: Organizational, Application & Tools, Market Related, Information Management, Project, and Process.

Organization is the first component which emphasizes the organizational aspect of BIM adoption in public construction projects (Chen et al., 2016). Organizations are the key elements a construction project. Capacitating firms and enhancing the competency of each professionals in organizations allows for a smooth BIM adoption in construction projects (Bosch-Sijtsema et al., 2019). The organizational aspect of BIM adoption in public construction projects mainly focuses on organizational culture, organizational structure, and organizational change management.

Organizational culture plays a crucial role in BIM adoption, as it affects the willingness and readiness of stakeholders to embrace BIM and collaborate with each other. A positive organizational culture that supports innovation, learning, trust, communication, and teamwork is essential for successful BIM adoption (Succar & Kassem, 2015). However, many public construction projects suffer from a negative organizational culture that hinders BIM adoption, such as resistance to change, lack of trust and commitment, and poor communication and coordination (Khosrowshahi & Arayici, 2012).

On the other hand, organizational structure refers to the formal and informal arrangements of roles, responsibilities, authority, and relationships among individuals and groups within an organization (Ahn et al., 2016). Organizational structure influences the efficiency and effectiveness of BIM adoption, as it determines how information and resources are shared and coordinated among stakeholders. A flexible and adaptive organizational structure that facilitates integration, collaboration, and coordination is desirable for BIM adoption (Ahuja et al., 2017). In addition, a systematic and proactive organizational change management approach that involves stakeholder engagement, vision and strategy formulation, change leadership, training and education, incentive and reward systems, and performance measurement and feedback is recommended for BIM adoption (Ahuja et al., 2017).

The second component of the FA analysis focus on Application & tools. Availability of hardware, software packages and the overall IT infrastructure is an integral part of BIM adoption process in projects (Chan et al., 2019). Similarly, market related factors are mainly related to government’s role with respect to devising BIM policies and its engagement in the successful implementation of BIM across the market. Literature outlines that legal issues are still one of the challenges of BIM adoption in the construction sector (Ozorhon & Karahan, 2017). In this respect, the regulatory body should play a vital role towards devising appropriate legal and contractual obligations when it comes to BIM implementation in public construction projects (Olanrewaju et al., 2020).

Having clear roles and responsibilities is one of the fundamental principles of BIM adoption in projects and requires careful attention with respect to project information management (Morlhon et al., 2014; Olawumi & Chan, 2019). Studies suggest that a smooth BIM data communication can enhance the accuracy and compatibility of flow of information within organizational structures, as well as between project stakeholders (Ahn et al., 2016). In addition, managing project workflow information also helps professionals and top management team to make timely projects decisions in every aspect of the project (X. P. Ma et al., 2020).

Project aspect is also one of the critical constructs of BIM adoption process. As it is indicated in Table 4, contractors perceived that creating a collaborative working environment among major parties in a certain project in essential for successful BIM adoption. The reason could be the successful adoption of BIM in construction projects directly affects the contractor (Ozorhon & Karahan, 2017). Moreover, It’s been known that public infrastructure projects take up a huge budget in developing countries (Ismail et al., 2017). Thus, availability of sufficient BIM adoption guidelines and careful choice of contract types during the procurement stage greatly affects the overall Process of BIM implementation in public construction projects.

6.1. Cross – country comparative analysis

One of the objectives of this study was to compare the findings with other countries. P. Ma et al. (2020) reported the benefits of construction engineering contextual comparative studies to compare findings across markets and obtain results from multiple methodological cross validations. In this regard, the top 5 ranked BIM adoption strategies were compared against three recent studies published in countries such as Ghana, Nigeria, and South Africa.

Level of BIM awareness and stakeholder’s involvement towards BIM adoption are common BIM adoption strategies in all four countries. Extent of availability of IT infrastructure and sufficient standards and guidelines for BIM adoption are the common critical strategy for Ethiopia and Nigeria, whereas government’s policy and encouragement for adoption of BIM and availability of competent consultancy are similar strategies in Ethiopia, Saudi Arabia, and South Africa. Moreover, availability of BIM education and training are shared by Saudi Arabia, Nigeria, and South Africa.

From the summary in Table 8, it can be clearly shown that the distinct BIM adoption critical strategies for BIM adoption are extent of BIM need and aims (ranked fifth) in Nigeria and encourage BIM adoption in the construction supply chain and encourage stakeholder collaboration (ranked first and second) in South Africa.

Table 8. Cross-country comparative analysis

Rank	This Study	(Aljobaly & Banawi, 2020) (Saudi Arabia)	(Amuda-Yusuf, 2018) (Nigeria)	(Odubiyi et al., 2019) (South Africa)
1	Strategic & adequate IT infrastructure	Stakeholders BIM awareness	Extent of IT infrastructure and business process	Encourage BIM adoption in the construction supply chain
2	Sufficient standards & guidelines for BIM Adoption	Role of consulting service in BIM adoption	Standards of BIM adoption	Encourage stakeholder collaboration
3	Government policy	Market competition by firms	Stakeholders understanding and involvement	Stakeholders BIM understanding
4	Highly competent consultancy service	Stakeholders understanding & commitment	Education and training	Extent of employee BIM training
5	Extent of BIM awareness	BIM training for workers	Extent of BIM needs & aims	Implementation by the government

In Addition, BIM adoption strategies in Latin and south Asian developing countries are diverse and dynamic. Some countries such as Brazil, Chile, Mexico, and Peru have issued mandates or guidelines that require or encourage the use of BIM in public projects (Machado et al., 2021). These mandates aim to improve the quality, efficiency, and transparency of public works and services. In contrast, BIM trainings and capacity development programs, innovation, and technological infrastructures are the major drivers of BIM Adoption in South Asian countries (Ahuja et al., 2020).

7. Conclusion

The current study outlined the critical BIM adoption strategies for public construction projects in the Ethiopian construction industry. The major findings include the following:

• The most significant BIM adoption strategies in Ethiopia are strategic & adequate IT infrastructure, availability of sufficient standards and guideline, and government policy.

• In contrast, managing risks associated with BIM, interoperability model data, and workflow management are perceived as the least critical strategies of BIM adoption in Ethiopia.

• Similarly, the study explored the level of agreement between the rankings of expert groups participated in the survey. Evidently, the findings reveal that the majority of expert groups agreed on their response except contractor vs consultant pair of groups.

• More so, the FA also contributes critical theoretical evidences and empirical justifications of BIM adoption strategies of construction projects by identifying key components such as organization, application & tools, market, information management, project, and process. These constructs alone are not sufficient enablers of BIM adoption, although they are considered to be baseline mechanisms for incorporation of BIM in public construction projects.

• Prior studies have been focusing on exploring the perception of client, consultants and contractor while neglecting academia in BIM adoption studies. However, this study incorporated the perception of experts in academia in examining critical BIM adoption strategies to get a better scope of results.

• The findings of the study provide theoretical implications and empirical validations to encourage BIM adoption initiatives in public construction projects. Similarly, the results can also be used as a benchmarking BIM adoption strategy to guide key business decisions for top tier management team and CEOs in the Ethiopian public construction sector.

• In addition, considering there are a dearth of BIM adoption studies in the developing nations, the paper can contribute practical implications to the existing BIM adoption strategical approaches in the literature.

• Further, the cross-country comparative analysis helps to contextualize the findings in diverse construction markets, and helps in the development of a comprehensive BIM adoption strategy across the Sub-Saharan African region.

• The current study doesn’t consider project types (residential, commercial, industrial) in the analysis, though it is common to explore BIM strategies in terms of generalized project concept.

• Future studies can be extended on organizational BIM adoption case studies using different project types, project life cycle and exploitation of the relationship between BIM adoption and sustainability in public construction projects for a better understanding of BIM usage.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data underlying the results presented in the study are available within the manuscript.

Additional information

Funding

The authors didn’t receive any specific funding for this work.