Factors Affecting the Accuracy of Pre-Tender Cost Estimates in the Gaza Strip

Abstract

According to them, most of the significant factors affecting project costs are qualitative, such as client priorities (e.g., completion time, procurement methods, market conditions, etc.). According to the Mann-Whitney test, there is no difference of opinion between clients and consultants in the factors affecting accuracy of pre-tender cost estimates at the significance level of 0.05.

Headnote

Abstract: Pre-tender cost estimates of construction projects require extensive knowledge and expertise. The aim of this paper is to identify, evaluate and rank essential factors affecting the accuracy of pre-tender cost estimating from the perspective of clients and consultants. A survey questionnaire was conducted to elicit professionals' views on and experiences with factors affecting the accuracy of pre-tender cost estimates; a total of 70 organisations (i.e., 46 clients and 24 consultants operating in the Gaza Strip, Palestine) responded to the survey. The results of analysing a total of 64 factors considered in the questionnaire reveal that the top five factors affecting the accuracy of pre-tender cost estimating are (1) materials (prices/availability/supply/quality/imports), (2) closure and blockade of borders, (3) project team's experience in the construction type, (4) the experience and skill level of the consultant and (5) clear and detailed drawings and specifications. Kendall's coefficient of concordance was used as a measure of agreement between the two groups of respondents (i.e., clients and consultants) who ranked various factors and it appears that they are generally in agreement. Both clients and consultant groups should focus on the main factors identified in this study to develop effective strategies for accurate cost estimating, which would ultimately lead to successful projects.

Keywords: Cost estimation, Accuracy, Clients, Consultants

(ProQuest: ... denotes formula omitted.)

INTRODUCTION

The success or failure of a project is dependent on the accuracy of several estimates done throughout the course of the project (Ahuja, Dozzi and Abou Rizk, 1994). Therefore, the preparation of a cost estimate of the project is one of the most difficult tasks in project management because it must be done before the work is accomplished (Oberlender, 1993). Pre-tender cost estimating is simply the final costing of the work carried out by a consultant (i.e., quantity surveyor or engineer) on behalf of a client (Odusami and Onukwube, 2008) before tenders are received. It sits somewhere between cost planning and post-contract cost control, provides an indication of the probable construction cost prior to contract- awarding and involves collecting, analysing and summarising all available data related to the construction of the project (Holm et al., 2005). Thus, for a contractor to secure a job, his cost estimate must be as accurate and competitive as possible (Marjuki, 2006). Inadequate estimating invariably leads to misallocation of scarce resources (Flyvbjerg, Holm and Buhl, 2002).

An estimate can be accurate, low or high. An accurate estimate generally results in the most economical project cost, while either an underestimation or an overestimation often leads to greater actual expenditures. Inaccuracy in the estimate of a project may arise from two sources: bias associated with the project itself and bias associated with the estimating techniques used and the operating environment (Aibinu and Pasco, 2008). Accurate estimation of construction costs is heavily dependent on the availability of quality historical cost data and the level of professional expertise, among other things. The limited information available at the early stages of a construction project may mean the quantity surveyor must make assumptions about the design details of a project, which may not eventuate as project design, planning and construction evolve (Liu and Zhu, 2007).

Professional estimators have access to reliable cost and productivity references for estimating labour, material, equipment and other major work components. These major cost items have a high visibility factor and consequently receive adequate attention in the preparation of the pre-tender estimate. However, there are little-known low visibility factors affecting the estimate accuracy, such as procurement forms and contract arrangements, that should be considered in the preparation of pre-tender estimates. Unfortunately, these factors are either entirely overlooked or sadly neglected by estimators in the Gaza Strip. Identification of these low visibility factors is very important for improving the overall performance of the construction industry. The purpose of this paper is to identify, evaluate and rank essential factors affecting the accuracy of pre-tender cost estimating from clients' and consultants' perspectives.

FACTORS INFLUENCING THE ACCURACY OF ESTIMATES: A LITERATURE REVIEW

Various studies have focused on identifying the factors that have some influence on the accuracy of estimating the costs of construction work. Based on previous studies, Gunner and Skitmore (1999) identified 12 factors: building function, type of contract, conditions of contract, contract sum, price intensity, contract period, number of bidders, good/bad years, procurement basis, project sector (public, private or joint), number of priced items and number of drawings. Ling and Boo (2001), using data from 42 projects in Singapore, found similar results when they compared five variables against Gunner and Skitmore's work. Skitmore and Picken (2000) studied the effects of four independent variables (building type, project size, sector and year) on estimating accuracy and tested these variables against 217 projects from a quantity surveyor based in the United State of America (USA). They found that bias existed in project size and year and consistency errors existed in project type, size and year.

By reviewing 67 process industry construction projects around the world, Trost and Oberlender (2003) identified and grouped a total of 45 factors that contribute to the accuracy of early stage estimates into 11 orthogonal factors. Of these 11 factors, the five most important are process design, team experience and cost information, time allowed to prepare estimates, site requirements and bidding and labour climate. Elhag, Boussabaine and Ballal (2005) stated that the technological and project design, the contractor's expertise and management ability and the client's desired level of construction sophistication play important roles in determining the cost of a project. According to them, most of the significant factors affecting project costs are qualitative, such as client priorities (e.g., completion time, procurement methods, market conditions, etc.).

Enshassi, Mohamed and Madi (2007) examined cost estimating practices in contracting companies operating in the Gaza Strip. Their study revealed that the most important factors affecting contractors' cost estimates are the financial status of client, the type of current contractor workload and the project's location relative to hostile "hot-spot" areas. According to Liu and Zhu (2007), two types of factors, control factors and idiosyncratic factors, influence and contribute to the cost of a project. Control factors are the factors that can be controlled by estimators to improve the performance of estimation, while idiosyncratic factors influence cost estimation but are outside the control of the estimators and include market conditions, project complexity, weather, contract size, site constraints, resource availability, type of procurement system and contract work type (Liu and Zhu, 2007). Considering the unstable political and economic conditions in the Gaza Strip, idiosyncratic factors can be regarded as more relevant and influential for cost estimation in this part of the world.

RESEARCH METHODOLOGY

Sample and Response Rate

This research was conducted in the Gaza Strip, Palestinian National Authority (PNA). The targeted research population consisted of construction engineers and managers from diverse public clients and consultancy organisations experienced in tendering and estimation. The selected organisations (either public client organisations or consulting firms) represented the top 20% of all organisations listed by the relevant government agency as ranked by their total project value procured or designed in 2010/2011. Targeted organisations are seen as key players in the provision of construction projects and consultancy services. To increase the response rate, where possible, more than one potential respondent was approached within each of these organisations to solicit the required information. However, to avoid response bias, no more than three questionnaires were sent to the same organisation. To illustrate, for a consulting organisation where a number of principal partners are listed as contact persons, no more than three questionnaires were sent out. In total, 50 questionnaires were distributed to 19 targeted clients and 28 questionnaires to 17 local consulting organisations.

The response rate from client organisations was 6% higher than from the consultancy firms, 92% (46 responses) compared to 86% (24 responses). A response rate of more than 30% is likely to produce results subject to non-response bias. Hence, the obtained response rates should produce reliable results. On average, the respondents have an average of 20 years' experience in tendering and estimating.

Questionnaire design

A questionnaire survey was undertaken to determine the opinions of clients and consultants regarding factors affecting the accuracy of pre-tender cost estimates in the Gaza Strip. The questionnaire was constructed based on a literature review, five face-to-face interviews with clients and consultants and the personal professional experience of the researchers.

A total of 85 factors that were identified and reported in 12 previous studies (Al-Khaldi, 1990; Al-Thunaian, 1996; Akintoye, 2000; Madi, 2003; Trost and Oberlender, 2003; Elhag et al., 2005; Shash and Ibrahim, 2005; Babalola and Aladegbaiye, 2006; Dysert, 2006; Enshassi et al., 2007; Liu and Zhu, 2007; Odusami and Onukwube, 2008) were considered in this research. To better reflect the nature of the local construction industry, 15 of these factors were deleted, 15 of them were modified and 12 factors were merged, while six new factors were added. In total, 64 factors were grouped into the following five groups as advised by experts who participated in a pilot study (Table 1):

1. Factors related to clients' characteristics

2. Factors related to consultants, design parameters and design information

3. Factors related to project characteristics

4. Factors related to contract requirements and procurement methods

5. External factors and market conditions

Pilot study

A pilot study, as suggested by Naoum (1998), provides a trial run for the questionnaire and involves testing the wording of questions, identifying ambiguous questions, testing the techniques used to collect data and measuring the effectiveness of standard invitations to respondents. A pilot study was conducted by distributing the modified questionnaire to six seasoned clients and four consultants. The results revealed that the questions were generally regarded as being clear, although some respondents had difficulties understanding a limited number of questions. Accordingly, these problematic questions were modified.

Instrument Validity

According to Polit and Hungler (1985), validity refers to the degree to which an instrument measures what it is supposed to be measuring. Instrument validity can be evaluated in two ways: (1) content validity and (2) statistical validity, which includes criterion-related validity and construct validity.

For this research, the content validity of the questionnaire was tested by an expert panel of ten experts in the project management field consisting of two academics, four consultants and four seasoned clients. Each expert was requested to evaluate the content validity for each item based on rating the index of content validity. Based on the comments of the experts, the final questionnaire included 64 factors grouped into five categories.

To ensure the validity of the questionnaire and following Polit and Hungler (1985), two statistical tests were applied. The first used a criterion-related validity test (Spearman test) and the correlation coefficient and p-value were calculated for each of the five factor groups. The results revealed that the p-values are less than 0.05, so the correlation coefficients of the field are statistically significant at α = 0.05 (0.01 < p-value < 0.05). Thus, it can be concluded that the factors of all groups are consistent and valid to measure what they are intended to measure.

The second test was a structure validity test (Spearman) to test the validity of the questionnaire structure by testing the validity of each group and the validity of the whole questionnaire. The test results showed that the p-value for each group is less than 0.05, so the correlation coefficients of all the groups are significant at α = 0.05 (0.01 < p-value < 0.05).

Cronbach's Coefficient Alpha test was also employed to measure the consistency of the questionnaire. This method is designed as a measure of internal consistency to determine whether all items within the instrument measure the same thing (George and Mallery, 2003). The Alpha value varies between 0.0 and +1.0, with higher values reflecting a higher degree of internal consistency. The results revealed that the calculated Alpha values for each group range from 0.676 to 0.841, demonstrating the reliability of each group in the questionnaire. For the entire questionnaire, this value was found to be 0.911, indicating very high reliability.

Respondents were asked to rate the importance of each factor using a five-point Likert scale. To measure attitudes with respect to surveyed variables, the Relative Importance Index (RII) technique was employed. This technique is widely used to analyse the factors affecting the accuracy of cost estimating (Akintoye, 2000; Madi, 2003; Elhag, Bossabaine and Ballal, 2005; Odusami and Onukwube, 2008).

In addition to the above techniques, two non-parametric methods, Kendall's Coefficient of Concordance and the Man-Whitney test, were utilised in this research. Non-parametric methods are widely used for studying populations that take on a ranked order. Kendall's Coefficient of Concordance (W) is used to evaluate the degree of agreement between the two groups, clients and consultants, regarding the ranking of key factors and it ranges from 0 (no agreement) to 1 (complete agreement) (Siegel and Castellan, 1988). The degree of agreement between the clients and consultants was tested using the following two hypotheses across the five factor groups:

Null Hypothesis, H0: There is an insignificant degree of agreement between clients and consultants.

Alternative Hypothesis, H1: There is a significant degree of agreement between clients and consultants.

The Mann-Whitney test is utilised to test the null hypothesis that two populations have identical distribution functions against the alternative hypothesis that the two distribution functions differ only with respect to location (median), if at all (Siegel and Castellan, 1988). This test is used to determine whether there is a significant difference at α ≤ 0.05 in rank means of the respondents' agreement between clients and consultants.

The relative index technique has been widely used in construction research for measuring attitudes with respect to surveyed variables. Likert scaling was used for ranking questions that have an agreement level. The respondents were required to rate the importance of each factor on a 5-point Likert scale using 1 for not important, 2 for of little importance, 3 for somewhat important, 4 for important and 5 for very important. Then, the relative importance index was computed using the following equation:

...

where W, which is the weighting given to each factor by the respondent, ranges from 1 to 5; n1 = the number of respondents for not important; n2 = the number of respondents for of little importance; n3 = the number of respondents for somewhat important; n4 = the number of respondents for important; and n5 = the number of respondents for very important. A is the highest weight (i.e., 5 in the study) and N is the total number of samples. The relative importance index ranges from 0 to 1 (Tam and Le, 2006).

RESULTS AND DISCUSSION

Group 1: Factors Related to Clients' Characteristics

Group 1 contains seven factors. The ranks and the RIIs of these factors related to client characteristics are shown in Table 2. Client experience level was ranked first with an RII of 0.889 by the clients and second with an RII of 0.758 by the consultants. The combined ranking by both clients and consultants placed this factor at the top of the list with an RII of 0.843 and 12th in the overall ranking of all 64 factors. These results are in agreement with the findings of Trost and Oberlender (2003) and Babalola and Aladegbaiye (2006) who have shown that the client experience level plays a significant role in cost estimate accuracy.

Clear scope definition for the client was ranked second by the clients and first by the consultants, with RII scores of 0.783 and 0.844, respectively. This factor was ranked second by both groups with an RII of 0.826 and was ranked 16th in the overall ranking.

The results are in agreement with previous studies, asserting that the accuracy of a cost estimate is highly dependent on the level of detail and adequate project scope because it is a definition of the client's requirements for space, function and quality of the proposed project (Akintoye, 2000; Trost and Oberlender, 2003; Babalola and Aladegbaiye, 2006; Dysert, 2006; Liu and Zhu, 2007; Odusami and Onukwube, 2008).

Client's method of payment was ranked last by both clients and consultants for this group with an RII of 0.666 and was ranked 59th in the overall ranking. Both client and consultant groups agreed that this factor does not play a significant role in pre-tender cost estimate accuracy This result somehow contradicts the findings of Al-Khaldi (1990), in Saudi Arabia, who concluded that client's method of payment has a very strong effect on the accuracy of the estimated cost.

For the Group 1 factors, W was 0.760 with a p-value (Sig.) less than the level of significance, α= 0.05. Thus, the null hypothesis (H0) was rejected and there is sufficient evidence to support the alternative hypothesis, H1. There is also a significant degree of agreement among the clients and consultants in this group. For Group 1, the Mann-Whitney test revealed that the obtained p-value (Sig.) was 0.35, which is much greater than the level of significance, α= 0.05, for this group. Therefore, there is an insignificant difference between the clients and consultants in terms of the factors related to client characteristics.

Group 2: Factors Related to Consultants, Design Parameters and Information

Relative Importance Indexs (RII) and ranks of the 23 factors included in this group are shown in Table 3. Project team's experience in the construction type was ranked first by both clients and consultants with an RII of 0.933 and 0.950, respectively. This factor was also ranked third by the combined total sample. This emphasises that this factor is considered a key factor affecting estimate accuracy. The results are consistent with the findings of two previous studies by Trost and Oberlender (2003) and Odusami and Onukwube (2008) in which this factor was ranked second and third, respectively.

The experience and skill level of the consultant was ranked second by both clients and consultants with an RII of 0.917 and it ranked fourth in the overall ranking. The clients ranked this factor third with an RII of 0.916, while the consultants ranked it second with an RII of 0.933. This result indicates that to produce an accurate estimate, those involved in the estimating process must have the relevant professional knowledge and skills. This result agrees with Odusami and Onukwube's (2008) findings in which the respondents ranked this factor first. Moreover, the obtained results are in line with Aibinu and Pasco (2008), Trost and Oberlender (2003) and Madi (2003), who stated that the accuracy of a cost estimate is highly dependent on the level of estimator experience. Having an experienced estimator is critical for producing high-quality and reliable cost estimates. Dysert (2003) emphasised that if an estimator were more professional, budget and other related problems could be greatly reduced.

Volume of consultant's workload during estimation and number of estimating team members were considered the least influential factors by both clients and consultants, with RIIs of 0.677 and 0.600 and overall rankings of 55th and 63rd, respectively. Clients and consultants are satisfied that these factors do not significantly affect the accuracy of cost estimates. These results are justified because the volume of construction projects in the last five years in Gaza has been reduced dramatically after experiencing a boom period, so time is available for consultants to adequately prepare the required cost estimate. Size of estimating team is not that influential if the estimating team members were well- qualified and experienced. These results match findings by Odusami and Onukwube (2008) and Akintoye (2000).

For the Group 2 factors, W was 0.894 with a p-value (Sig.) less than the level of significance, α= 0.05, so the null hypothesis was rejected. Therefore, it can be concluded that there is sufficient evidence to support the alternative hypothesis: there is a significant degree of agreement among the clients and consultants in this group.

According to the Mann-Whitney test, the obtained p-value (Sig.) was 0.394, which is greater than the level of significance, α= 0.05, for this group. As such, there is an insignificant difference between the responses of clients and consultants.

Group 3: Factors Related to Project Characteristics

A total of 11 factors were analysed in this group, as shown in Table 4.

Project complexity was ranked first by both clients and consultants with an RII of 0.806 and it ranked 22nd overall. Additionally, clients and consultants separately ranked this factor first, with RIIs of 0.814 and 0.783, respectively. In this survey, complexity entails the technical complexity of the task, the amount of overlap and interdependencies in the construction stage, the project organisation, the site layout and the unpredictability of work and site construction. This result is supported by Madi (2003), who illustrated that the complexity of projects in the Gaza Strip is likely to increase construction costs. In another study performed by Akintoye (2000) in the United Kingdom (UK) on factors influencing contractors' cost estimating practices, it was established that this factor was the most important. Project complexity has also been addressed by Elhag, Boussabaine and Ballal (2005), who considered it to be a key factor.

The site conditions factor was ranked second by both clients and consultants with an RII of 0.766 and was ranked 29th overall. Both clients' and consultants' separate rankings placed this factor third with RIIs of 0.769 and 0.750, respectively. This finding emphasises that the estimator must have prior knowledge of site conditions to produce an accurate estimate. The result is consistent with the case study by Enshassi, Mohamed and Madi (2007) in which the winning contractor decreased his estimate because the project location was classified as a quiet area. Moreover, this result closely matches that obtained by Elhag, Boussabaine and Ballal (2005).

Location of project and type of structures were ranked 10th and 11th by both clients and consultants with RIIs of 0.646 and 0.637 and overall rankings of 61st and 62nd, respectively. Clients and consultants believed that these factors have the least influence on the accuracy of pre-tender cost estimates. This result may be because cities and towns in the Gaza Strip are reasonably accessible by means of suitable road infrastructure and the size of projects in the Gaza Strip cannot be classified as mega projects, so these factors have little effect on tender estimation. These results are still consistent with the findings of previous studies undertaken by Akintoye (2000), Elhag, Boussabaine and Ballal (2005) and Odusami and Onukwube (2008).

For the Group 3 factors, W was 0.768 with a p-value (Sig.) less than the level of significance, α = 0.05; the null hypothesis was rejected. The alternative hypothesis, that there is a significant degree of agreement among the clients and consultants in this group, was justified.

The obtained p-value (Sig.) from the Mann-Whitney test was 0.781, which is greater than the level of significance, α= 0.05, for this group and indicates that there is an insignificant difference between the client and consultant respondents toward the factors related to project characteristics.

Group 4: Factors Related to Contract Requirements and Procurement Methods

A total of nine factors in Group 4 were analysed, as shown in Table 5. The clear contract conditions factor was voted the most important factor by both clients and consultants with an RII of 0.866 and was ranked 10th overall. It was also ranked first separately by clients and consultants with RIIs of 0.858 and 0.875, respectively. These results agree with Dysert (2006), who stated that clear contract conditions and specifications ensure the proper delivery of services by the contractors and protect the organisation from losses or damages.

Tender selection method was assessed as the second most important factor by both clients and consultants with an RII of 0.814 and it was 20th in the overall ranking. Advanced payment and content of disputes resolution methods clause factors were placed in the last two positions by both clients and consultants, with RIIs of 0.703 and 0.674 and overall rankings of 49th and 56th, respectively. These results were supported by Madi (2003), who argued that these two factors have moderate effects. Moreover, these results are in complete agreement with Elhag et al. (2005), whose results placed this factor 55th out of 67 factors.

For the Group 4 factors, W was 0.800 with a p-value (Sig.) less than the level of significance, α = 0.05, which leads to a rejection of the null hypothesis. Hence, there is a significant degree of agreement among the clients and consultants in this group and the alternative hypothesis is justified.

From the Mann-Whitney test, the obtained p-value (Sig.) was 0.662, which is greater than the level of significance, α= 0.05, for this group. Therefore, there is an insignificant difference between the client and consultant respondents toward the factors related to contract requirements and procurement methods.

Group 5: External Factors and Market Conditions

As shown in Table 6, the 14 factors analysed in this group are related to the external factors and market conditions.

Both clients and consultants evaluated material as the most important factor, with an RII of 0.954 and 1st overall ranking. Separate voting of clients and consultants also placed this factor first, with RIIs of 0.947 and 0.967, respectively. This result came as no surprise because in the case of a border closure, available stocks of construction materials would soon run out, leading to price escalation. Enshassi, Mohamed and Madi (2007) reported that a continuous increase in materials consumption may be justified by the repeated border closures and the instability of local markets. During any closure, construction activities would usually be suspended, so this factor has to be considered very carefully by estimators when preparing their pre-tender estimates. The variation between this research and others can be attributed to the different environmental conditions between the Gaza Strip, the UK and Nigeria.

Closure and blockade was ranked second by both the clients and consultants with an RII of 0.951 and second overall. The clients ranked this factor first with an RII of 0.947 and the consultants ranked it second with an RII of 0.954. This result is in line with findings reported by Al-Shanti (2003), who noted that a closure has a large effect on the prices of basic construction materials such as cement, steel and aggregate, resulting in substantially increased project costs.

Impact of government regulations requirement and Social and cultural impact were assessed as the two least important factors by both clients and consultants, with RIIs of 0.649 and 0.540, respectively and overall rankings of 60th and 64th, respectively. These results were supported by Al-Khaldi (1990), who argued that these two factors have little influence on the estimating process and scored 50% and 61%, respectively.

The W was equal to 0.842 with a p-value (Sig.) less than the level of significance, α = 0.05, so the null hypothesis is rejected and the alternative hypothesis is justified. That is, there is a significant degree of agreement among the clients and consultants in this group.

From the Mann-Whitney test, the p-value (Sig.) was 0.058, which is slightly greater than the level of significance, α= 0.05, for this group. Therefore, there is an insignificant difference between the client and consultant respondents toward the factors related to external factors and market conditions.

Group Comparison

The comparison of the five groups of factors affecting the level of accuracy of pre-tender cost estimates in the Gaza Strip is shown in Table 7.

The factors related to consultants, design parameters and information group was ranked first by clients and consultants, with RIIs of 0.809 and 0.788, respectively. The presence of six of the ten highest ranked factors from this group of factors indicates that this is the most important group of factors affecting cost estimate accuracy of pre-tender. This result agrees with Trost and Oberlender (2003) and Elhag, Boussabaine and Ballal (2005).

Factors related to external factors and market conditions was ranked second by clients and consultants, with RIIs of 0.794 and 0.740, respectively. Three factors from this group were within the 10 highest ranked factors and two of these three, material and closure and blockade, were ranked as the two most important factors by both clients and consultants.

While factors related to contract requirements and procurement methods group was ranked third by both the clients and consultants, the clear contract conditions factor of this group achieved the 10th highest score overall. This result indicates that the clear contact conditions have a clear effect on tender prices.

Factors related to clients' characteristics and factors related to project characteristics were ranked fourth and fifth, respectively, by clients and consultants. It is evident that both clients and consultants regarded these two groups as having relatively little effect on the accuracy of pre-tender cost estimates in the Gaza Strip.

CONCLUSIONS

An exploratory study of factors affecting the accuracy of pre-tender cost estimates was conducted to determine the relative level of influence of each factor. The ranking of 64 factors revealed that the ten most influential factors affecting cost estimate accuracy are as follows:

1. Material (prices/availability/supply/quality/imports)

2. Borders closure and blockade

3. Project team's experience in the type of construction

4. Experience and skill level of the consultant

5. Having clear and detailed drawings and specifications

6. Quality of information flow

7. Completeness of cost information

8. Accuracy and reliability of cost information

9. Currency exchange fluctuation

10. Clear contract conditions

The five least influential factors, as evaluated by clients and consultants, are impact of government regulations requirement, project location, type of structures, size of estimating team and social and cultural impact.

From the above results, it was concluded that both clients and consultants generally agree on the ranking order of the factors affecting cost estimate accuracy. This agreement confirms the influential effects of those factors on the accuracy of cost estimates and provides a level of validation for this research. This validation was confirmed by the high values of the Kendall's coefficients of concordance achieved within each group. According to the Mann-Whitney test, there is no difference of opinion between clients and consultants in the factors affecting accuracy of pre-tender cost estimates at the significance level of 0.05.

It is recommended that clients and consultants give more attention to the most important factors that affect the accuracy of pre-tender cost estimates in order to achieve more reliable and realistic estimates. They should monitor the performance of their estimates in terms of accuracy and hire a qualified technical staff to obtain accurate estimates. Clear identification of project requirements is essential before the start of the estimating process. Clients and consultants should also obtain as much accurate information as possible from manufacturers and suppliers pertaining to the costs of procured materials and/or systems. If clients and consultants have a poor understanding of materials (prices/availability/supply/quality/imports), this would undoubtedly affect the accuracy of cost estimates. Clients and consultants should make sure that contract conditions are very clear to both parties.

Finally, it is also recommended that training courses on factors affecting the accuracy of cost estimates should be conducted. These activities would improve the local practice of cost estimating and increase the capabilities of estimators by using estimating software packages. A number of case studies from real life projects are being conducted to gather empirical data.

The findings of this study will help clients and consultants focus on the main causes affecting the accuracy of pre-tender cost estimating and develop effective strategies for accurate cost estimating.

References

REFERENCES

Ahuja, H., Dozzi, S. and Abou Rizk, S. (1994). Project Management Techniques in Planning and Controlling Construction Projects. 2nd Ed. New York: John Wiley and Sons Inc.

Aibinu, A.A. and Pasco, T. (2008). The accuracy of pre-tender building cost estimates in Australia. Construction Management and Economics, 26(12): 1257-1269.

Akintoye, A. (2000). Analysis of factors influencing project cost estimating practice. Construction Management and Economics, 18(1): 77-89.

Al-Khaldi, Z.S. (1990). Factors affecting the accuracy of construction costs estimating in Saudi Arabia. MSc diss. King Fahd University of Petroleum and Minerals.

Al-Shanti, Y. (2003). A cost estimate system for Gaza Strip construction contractors. MSc diss. The Islamic University of Gaza.

Al-Thunaian, S. (1996). Cost estimation practices for buildings by A/E firms in the eastern province, Saudi Arabia. MSc diss. King Fahd University of Petroleum and Minerals.

Babalola, O. and Aladegbaiye, O. (2006). Determination of appropriate contingency sum for building projects. Paper presented at the Annual Research Conference of the Royal Institution of Chartered Surveyors. University College London, 7-8 September.

Dysert, L.R. (2006). Is "estimate accurate" an oxymoron? AACE International Transactions, EST.01: EST01.1-01.5.

_____. (2003). Sharpen your cost estimating skills. Cost Engineering, 45(6): 1-6.

Elhag, T.M.S., Boussabaine, A.H. and Ballal, T.M.A. (2005). Critical determinants of construction tendering costs: Quantity surveyors standpoint. International Journal of Project Management, 23(7): 538-545.

Enshassi, A., Mohamed, S. and Madi, I. (2007). Cost estimation practice in the Gaza Strip: A case study. The Islamic University Journal, 15(2): 153-176.

Flyvbjerg, B., Holm, M.S. and Buhl, S. (2002). Underestimating costs in public works projects: Error or lie? Journal of the American Planning Association, 68(3): 279-295.

George, D. and Mallery, P. (2003). SPSS for Window Step by Step. 4th Ed. Boston, MA: Allyn and Bacon.

Gunner, J. and Skitmore, M. (1999). Comparative analysis of pre-bid forecasting of building prices based on Singapore data. Construction Management and Economics, 17(5): 635-646.

Holm, L., Schaufelberger, J.E., Griffin, D. and Cole, T. (2005). Construction Cost Estimating: Process and Practice. Upper Saddle River, NJ: Pearson Prentice Hall, 3-12.

Le, K.N. and Tam, V.W.Y. (2007). A survey on effective assessment methods to enhance student learning. Australasian Journal of Engineering Education, 13(2): 13-20.

Ling, Y.Y. and Boo, J.H.S. (2001). Improving the accuracy of approximate estimates of building estimates. Building Research and Information, 29(4): 312-318.

Liu, L. and Zhu, K. (2007). Improving cost estimates of construction projects using phased cost factors. Journal of Construction Engineering and Management, 133(1): 91-95.

Madi, I. (2003). Essential factors affecting accuracy of cost estimation of building contractors. MSc diss. The Islamic University of Gaza.

Marjuki, M. (2006). Computerized building cost estimating system. MSc diss. University Teknologi Malaysia.

Naoum, S. (1998). Dissertation Research and Writing for Construction Students. Oxford: Butterworth-Heinemann.

Oberlender, G. (1993). Project Management for Engineering and Construction. New York: McGraw-Hill.

Odusami, K.T. and Onukwube, H.N. (2008). Factors affecting accuracy of pre- tender cost estimate in Nigeria. Cost Engineering, 50(9): 32-35.

Polit, D. and Hungler, B. (1985). Essentials of Nursing Research: Methods and Applications. Philadelphia: JB Lippincott.

Shash, A.A. and Ibrahim, A.D. (2005). Survey of procedures adopted by A/E firms in accounting for design variables in early cost estimates. Journal of King Saud University, 18(1): 1-17.

Siegel, S. and Castellan, N.J. (1988). Nonparametric Statistics for the Behavioral Sciences. Singapore: McGraw-Hill International.

Skitmore, M. and Picken, D. (2000). The accuracy of pre-tender building price forecasts: an analysis of USA data. Australian Institute of Quantity Surveyors Refereed Journal, 4(1): 33-39.

Trost, S.M. and Oberlender, G.D. (2003). Predicting accuracy of early cost estimates using factor analysis and multivariate regression. Journal of Construction Engineering and Management, 129(2): 198-204.