Chunnian Liu1

1Beijing Municipal Key Laboratory of Multimedia and Intelligent Software Technology,

College of Computer Science and Technology,

Beijing University of Technology, Beijing 100022, China

{jianqiang.ma, weibingchen}@gmail.com

2Department of Computer and Information Science,

Norwegian University of Science and Technology (NTNU)

NO-7491 Trondheim, Norway

{jingyue, conradi}@idi.ntnu.no

Abstract.Most studies of software outsourcing focus on how to manage the

outsourcing from the perspective of the outsourcer, i.e., a company issuing a

subcontract. There are few studies of outsourcing presented from the viewpoint

of the supplier, i.e., a company receiving a subcontract. Since more and more

Chinese software companies are getting software outsourcing subcontracts from

all over the world, it is important to investigate how software outsourcing

projects are actually performed in China, and to identify possible enhancements.

Our study has collected data by a questionnaire-based survey from 53 finished

projects in 41 Chinese software suppliers. The results show that: 1) Differences

in natural languages may not be the barrier of Chinese software suppliers. 2)

Email is most used to discuss development related issues, while face-to-face

meetings are mainly used to discuss management and requirements issues. 3)

The main reasons for overtime work at the Chinese suppliers are design or

requirements changes initiated by the outsourcers and the suppliers’ initial

underestimation of the effort.

Keywords: Software Outsourcing, Empirical Study

1 Introduction

Software outsourcing is gaining more and more attentions. It can help software companies to save costs and to focus on their core businesses [1, 5]. Along with the China’s policy of opening to the outside world, China is becoming one of the biggest software supplying countries, together with India, Ireland, Russia, and so on. Although the volume of Chinese software exports in 2005 is $3590 million, which has grown from $720 million in 2001, it is still only 1/6 of India’s and is 0.5% of the total volume of Chinese exports [2, 15].

Comparing to in-house software development, there are several factors that may impact the effectiveness of software outsourcing, such as infrequent and ineffective communications between decentralized teams [6], time zone differences [4], andcultural differences [7, 8, 13]. Many studies have been performed on these issues, but mainly from the outsourcers’ perspective. Since China is emerging as a major player in software outsourcing, it is important to study and make guidelines on such issues, in order to help Chinese software suppliers to improve their businesses.

Our questionnaire-based survey has investigated seven software engineering issues in Chinese software suppliers. Due to page limitation, we present only three of them in this paper, namely differences in natural languages, effective communication, and overtime work. We have used membership lists from a national Chinese Software Organization (called CSO*in this paper) to achieve a representative subset of software companies. We have gathered information from 53 finished projects in 41 Chinese suppliers. The results show that differences in natural languages may not be the critical factor to affect the success of Chinese software suppliers. Email is the most common mean of communication in development related issues between the outsourcers and suppliers, while face-to-face meetings are mainly used to discuss management and requirements issues. Design or requirements changes initiated by the outsourcers and the suppliers’ initial underestimation of the effort are the main reasons for suppliers’ overtime work.

The remainder of this paper is structured as follows: Section 2 presents the related work and research questions. Section 3 describes the research design. Section 4 presents results and discussions. Section 5 contains a general discussion. Conclusion and ideas for future work are presented in Section 6.

2 Related work and research questions

There are two main participators in software outsourcing: the outsourcer and the supplier. Software outsourcers are the organizations that give software development subcontracts to other organizations, called the suppliers. Most previous studies focus on managing the software outsourcing from the outsourcers’ point of view, such as how to manage the cultural differences [7, 8, 13], how to evaluate and select capable suppliers [10], and how to manage contractual related issues [3]. However, only few studies [11, 12] focus on facilitating the efficiency of software outsourcing from the suppliers’ perspective, such as how to communicate and cooperate with outsourcers.

Since China has one of the largest software supplier industries in the world, it is opportune to investigate how Chinese software suppliers can improve their practices to communicate and cooperate with software outsourcers all over the world. The possible lack of English skill can pose constraints for Chinese software suppliers in communicating with outsourcers from Western countries [7]. As it is much harder to follow the plans in projects across organizational and geographic boundaries than projects within the same place and organization [16], the geographical differences between China and Western countries may also introduce additional project overruns [11, 14].

* The name of this organization was omitted for confidential reasons2.1 Differences in Natural Languages

In software outsourcing projects, the natural language being used in the documents is an important factor, because language differences may cause misunderstandings. For instance, word “more than” in Japanese means equal and more (>=), but in Chinese it means more (>). The misunderstandings in requirements specifications or other development documents may cause project overruns.

A previous study [4] proposed that “the language factor is one of the reasons for the success of software outsourcing in countries with strong English language capabilities, such as Philippines and Singapore”. The study [9] concluded that the major Chinese market for software outsourcing is Japan, because of a similar culture. On the other hand, India’s major market is in US and UK because of a shared English language. So, the biggest barrier to Chinese software suppliers towards Western companies seems to be the English language. However, Japan and China also speak different natural languages, but the outsourcing projects between them seem to work well. So our first research question is:

RQ1: Are differences in natural languages the barrier to Chinese software suppliers?

2.2 Effective Communication

Compared with in-house software development, the communication shared in software outsourcing is affected by distribution of both space and time. The study [4] concluded that synchronous communication, such as telephone meetings or video conferences, can resolve misunderstandings and small problems before they become bigger problems. On the other hand, asynchronous communication like email often delays or complicates problem resolution. However, the results of another study [11] show that email is used much more than synchronous channels in a Latvian subcontractor company. To improve suppliers’ current practices, it is important to know which communication channel is mostly used between foreign outsourcers and Chinese suppliers. So our second research question is:

RQ2: What is the most common mean of communication between foreign outsourcers and Chinese suppliers?

2.3 Overtime Work

According to a survey [18], around 60% to 80% software development encountered effort and/or schedule overruns. The average overrun is about 30% to 40%. Schedule and budget risks are the highlighted risk factors in software outsourcing [12]. Comparing with in-house development, more issues, such as time zone differences and organizational differences, may also cause project overruns [11]. To keep the project schedule, overtime work is often used as a remedy. Most Chinese software suppliers get contracts from Japan, which has only one hour time difference with China. It seems that time zone difference is not the major reason of overtime work ofChinese suppliers. We wonder there are other reasons that cause the project overrun. So, our third research question is:

RQ3: What are the main reasons for overtime work at Chinese software suppliers?

3 Research design

To answer the research questions, we have used a questionnaire-based survey to collect data. First, a preliminary questionnaire with both open-ended and close-ended questions was designed by reading literatures. Second, a pre-study was performed to verify the quality of questions in the preliminary questionnaire, and to get answers on the open-ended questions. Based on the results of the pre-study, most open-ended questions in the preliminary questionnaire were redesigned into close-ended questions. In addition, the problematic questions in the preliminary questionnaire were revised. Then, the revised questionnaire was used to collect data in a main study.

3.1 The preliminary questionnaire

The preliminary questionnaire was designed to study seven issues of software outsourcing from the suppliers’ perspective and had 10 sections. Sections 1 and 10 contain questions to collect background information of projects, companies and respondents. Questions in sections 2, 5 and 7 are related to the research questions in this paper. The remaining sections investigate four issues, which are related to technology management (e.g., how is source code integrated and updated) and business relationship management (e.g., how did the outsourcer and supplier get to know each other), are not reported in this paper.

3.2 The pre-study to verify and refine the preliminary questionnaire

The pre-study included two steps, where individual interviews were followed by a workshop.

Step 1 – Individual interviews. We interviewed 5 project managers from 5 different software suppliers. All the interviewees had solid experiences in software outsourcing. Each interview was conducted by two authors of this paper. One was responsible for conducting the interview and the other recorded answers and asked for clarification if needed. The interviews were also taped for later use.

Step 2 – A workshop discussion. After the individual interviews, we revised most open-ended questions in the preliminary questionnaire into close-ended questions and made a second version of the preliminary questionnaire. We then organized a group discussion (a workshop) with more than 30 industrial experts to verify and comment the second version of the questionnaire. Based on inputs from the workshop, we revised the questionnaire into a final version. The final questionnaire includes 66 questions and takes about forty minutes to be filled in.3.3 The main study to collect data

In the main study, the data was collected by cooperating with the CSO. As mentioned, we got 53 questionnaires from 41 companies. The sample selection and data collection process were as follows:

1.Establish the target population. We randomly selected 2,000 companies from a

database of the CSO, which included about 6,000 Chinese software companies.

2.Send invitation letters by email to obtain possible participants.We sent

invitation letters by email to the 2,000 selected companies. The invitation letter introduces the survey and specifies that the survey participant will be rewarded with either the final report of this survey or an annual membership of the CSO worth 500 Chinese Yuan. We got responses from 300 software outsourcing companies. These companies were later used as the original contact list.

3.Send questionnaires by email to possible participants. The unit of our study is

defined as a finished software outsourcing project. We sent questionnaires (as word files) by email to the 300 companies and asked them to select one or more projects to answer the questionnaire. Since we cannot get the complete list of relevant projects in each company, the projects selection within the companies were decided by the respondents themselves, i.e., convenience sample.

4.Collect filled-in questionnaires with follow up.From the 300 companies, we

first got 40 questionnaires back. To ensure the quality of the data, we excluded 10 questionnaires answered by respondents with less than three years working experiences. For the remaining 30 questionnaires, we contacted the respondents again by telephone to clarify possible misunderstandings and to fill-in the missing data. At the same time, we contacted the rest of 260 companies through telephone to persuade them to fill in the questionnaire. By doing this, we got 23 more questionnaires back.

4 Results and discussions of research questions

In this section, we first present an overview of the collected questionnaires. After that, we show the results of each research questions followed by detailed discussions.

4.1 Overview of collected questionnaires

Participating companies. According to the number of employees, the participating companies include 7 small, 22 medium, 8 large, and 4 super large software companies, as shown in Fig. 1. Comparing with the profile of the number of employees in Chinese software industry [2], it shows that most of the participating companies in our survey are medium and large companies.

Fig. 1. The distribution of companies

Human respondents.Most respondents are IT managers, project managers, or software architects. More than 70% of the respondents have at least 5 years experiences of software development. All of them have at least a Bachelor’s degree in computer science or telecommunication.

Participating projects.All of the 53 projects are finished software outsourcing projects. The mean duration of these projects (before the first delivery) is 161 days, and mean project effort is 1758 person-days. There are 21 projects outsourced from the US or European companies. One of these 21 projects used French in the document, and the other ones used English. There are 32 projects outsourced from Japanese companies. One of the 32 projects used Chinese in the documents, 8 of them used both English and Japanese, and the other 23 projects used Japanese.

4.2 Results and Discussions

4.2.1 Investigating RQ1: Are differences in natural languages the barrier to Chinese software suppliers?

Results on RQ1. For this research question, we investigate how well the suppliers’ employees understand the non-native natural languages used in the documents (e.g., business agreements, requirements specifications, and design specifications) provided by the outsourcers. The scales used to measure people’s language skills are: master (near native), fluent, competent (can speak, read, and write, but not fluently), read&write only, unfamiliar, and do not know. The results of RQ1 are presented in Fig. 2 and Fig. 3.

−Fig. 2(a) shows that respondents working on projects outsourced from American and European companies (with only English in documents) have good knowledge of English. However, their Japanese language skills are poor.

−Fig. 2(b) shows that respondents working on projects outsourced from Japanese companies with only Japanese in the documents are good at Japanese. Some of them have solid knowledge about English, while the others have poor English

skills.

−Fig. 2(c) shows that respondents working on projects outsourced from Japanese companies with both English and Japanese in documents have much better skills in English than in Japanese.

−Fig. 3shows that employees of the suppliers on average have better language skills in English than in Japanese.

Fig. 2(a).

Fig. 2(b).

Fig. 2(c).

The respondents’ knowledge of the different natural languages used in documents.

Fig. 2.

Discussion on RQ1.Kshetri [9] concluded that “The US and the UK are India’s major markets, thanks to its English competence. On the other hand, China’s cultural similarity to Japan and South Korea has facilitated its software export.”However, our results in Fig. 2(c) and 3 show that the employees of Chinese software suppliers on average have better knowledge in English than in Japanese. Moreover, results of Fig. 2(b) show that some respondents working with documents written in Japanese also have good English language skills. The possible reason is that most employees in the investigated Chinese software suppliers have a Bachelor’s or higher degree. In order to get such degrees, they must pass an English certification called CET-4 (College English Test band 4) or CET-6 to qualify their competence of English. Since Chinese software suppliers have successfully developed large software systems for Japanese outsourcers, their equivalent or better English skills indicate a big potentialto develop software for outsourcers from English speaking countries. In addition, our talks with interviewees in the pre-study reveal that some companies are using a steering group with languages competent people as a buffer between developers and outsourcers (as shown in Fig. 4). In this case, the Chinese software suppliers managed to resolve the confusions caused by the possible poor language skills of the developers.

Fig. 4. The steering group process

Thus, our conclusion on RQ1 is: Differences in natural languages may not be the barrier for Chinese software suppliers to develop software for English speaking outsourcers. Having a steering group with good English language skills may help to reduce the language impact.

4.2.2 Investigating RQ2: What is the most common mean of communication between foreign outsourcers and Chinese suppliers?

Results on RQ2. To study this research question, we listed several possible channels of communication, such as email, video conferences, and face-to-face meetings, which could be used to discuss issues between outsourcers and suppliers. For each communication type, we asked whether it has been mainly used to specify, clarify, or negotiate issues, such as requirements, business, and schedule. If the communication type has been used for more than one issue, the respondents can mark several alternatives.

We calculated the number of projects using a certain communication type to a specific issue vs. the total number of projects (i.e., 53). For example, 28 projects mainly used email to discuss requirements related issues. Thus, the result is 28/53 or 53%. The summary of all results stand in Table 1. Since the maximum number value in Table 1 is 60%, we divided them into four categories to simplify the presentation. We excluded numbers less than 30%. Numbers more than 30% and 40% are replacedwith one plus. Numbers between 40% and 50% are replaced with two plusses. Numbers more than 50% are replaced with three plusses.

The data in Table 1 show that email correspondence was used on almost all issues, such as discussing requirements, harmonizing the schedule, handling changes, development, and testing. The popularly used synchronous communication types are telephone meetings and face-to-face meetings, where the face-to-face meetings are most used to discuss the requirements and business related topics. On the other hand, net meetings and video conferences are less used than others.

Discussion on RQ2.The previous study [4] proposed that synchronous communication means are more suitable to be used in the outsourcing contexts than asynchronous communication means.However, few empirical studies have investigated how the synchronous and asynchronous communication types are used in practice. Our results show that both the synchronous, such as telephone meetings and face-to-face meetings, and the asynchronous communication type, such as email, have been used all over. However, different communication types have been used to solve different issues. The advantage of email is that it is much cheaper than synchronous communication channels, and is not affected by the distribution of both space and time. Chinese software suppliers have used email to discuss almost everything with outsourcers. On the other hand, face-to-face meetings are mainly used in requirements and business related topics. A case study in one Latvian software supplier [11] shows that email is used much more than other communication channels, while net meetings and video conferences are almost never used. Our results from Chinese software suppliers show the same trend. However, different time zones and holidays between the suppliers and outsourcers may affect their available communication types. If we divide these projects into two groups based on having or not having time zone problems (i.e., Japanese vs. Western), the two groups show similar results as in Table 1. Thus, the different time zones and holidays between the suppliers and outsourcers may not affect the results of our study.

Our conclusion on RQ2is: Email is mainly used for development issues, while face-to-face meetings are popularly used for management and requirements issues. Net meetings and video conferences are the least used communication types.

Table 1. Means of communications and issues to be discussed (N = 53, Multiple Choice)

Asynch

ronous

Synchronous

Email Net

meeting Phone

meetin

g

Video

conf.

Face-to-

face

meeting

Requirements +++ + ++ + +++ Development ++ + ++ + + Testing ++

Maintenance

Staff arrangement +

Harmonizing schedule +++ + ++ ++ ++ Handling changes +++ ++ + ++ Business related issues +++4.2.3 Investigating RQ3: What are the main reasons for overtime work at Chinese software suppliers?

Results on RQ3. To investigate this research question, we asked whether there was overtime work in Chinese software suppliers. Results show that 51 out of the total 53 projects had overtime work, usually unpaid. We also asked the respondents their average working hours per week, the percentage of employees in their companies involved in overtime work, and their outputs during overtime work vs. their total outputs. The results show that, the average work hours per week are about 48, meaning the average overtime is 8 hours per week (20% more). Since 70% of the full-time employees are involved in overtime work, the expected relevant outputs from the overtime hours should be 14% (20% * 70%) of that from normal hours. However, the real “output” production from overtime work is 35%/65% (35% outputs are from the overwork hours and 65% outputs are from the normal working hours), or 54%, which reveals that overtime work is 3 times more efficient than work during normal working time.

To know the reasons for overtime work, we listed several possible alternatives from the literature [12] and from the feedbacks of the pre-study. The respondents were asked to select one or more of the alternatives. More than 70% of the respondents point out that design or requirements changes are the major reason for working overtime (see Table 2). In addition, nearly 60% of the respondents think overtime work is caused by initial underestimation of project effort and duration. Table 2. Reasons for overtime work (N=51, Multiple Choice)

Number of projects Percentage Different time zones 7 14% Different holidays 10 20% Design or requirements changes 37 73% Required by the outsourcers 15 29% Insufficient local competence 16 31% Over-optimism of schedule or effort 30 59% Resolve some specific problems 1 2% Discussion on RQ3. We found that the major reason for overtime work in Chinese software suppliers is design or requirements changes, rather than time zones differences. Furthermore, initial underestimation of effort and duration is also an important reason for overtime work. Some of the projects were underestimated because of lack of experience, but many were also victim of tactical bidding [12]. That is, to get a contract, a supplier indicated an unrealistic time schedule or effort to the outsourcer. This then leads to later overtime work for keeping the deadline or budget. If the overtime work (20% more) in our investigated projects is paid, which is usually 1.5 time of normal payment, the real budgets of our investigated software outsourcing projects should be increased by 30%. This means that software outsourcing projects suffer from similar under-budgeting, which is about 30% to 40% percentage [18], as other software projects.

In our pre-study, some of the interviewees attributed the frequent and unpaid overtime work in their companies to companies’ encouragement on overworking. As a result, the employees may not have tried to work as effectively as they could duringnormal hours and may have postponed certain work into the overtime hours. This may explain our finding that the overtime work is more productive than normal time work. Also the less interruption from other employees could be another reason for more productive overtime work. Since our results show about 70% full-time employees worked overtime, it is hard to conclude that the interruptions in the overtime are less than in the normal time.

Our conclusion on RQ3is: The main reasons for overtime work at Chinese software suppliers are design or requirements changes and initial over-optimism of the budget. The cultural encouragement of overtime work may also reduce the efficiency in the normal work hours.

5 Final discussion

5.1 General discussion

Based on our results of RQ1 to RQ3, we suggest the following strategies for Chinese software suppliers to improve their efficiency.

Manage the possible language differences with a steering group

The language factor is one of the reasons for the success of software outsourcing. The successes of software industries in India, Philippines and Singapore are accepted examples [4, 9]. Before our study, few studies have systematically examined the language skills of employees of Chinese software suppliers. The outsourcers from English speaking countries may still be skeptical to the English language skills of their Chinese suppliers. However, our results show that the differences in natural languages are not the barrier to Chinese software suppliers. For the Chinese suppliers, one possible method to manage the language effects is to build a steering group that is comprised by language competent project managers and senior developers. The members of the steering groups can work as a bridge between the external outsourcers and internal developers. Since most communication between the internal developers and the external outsourcers need to go through such a steering group, we later need to study how to improve the performance of such a group.

Do not overuse synchronous communication methods

A previous study [4] shows that synchronous communication channels usually have more advantages than asynchronous communication channels. However, our results and another previous study [11] show that email is still used frequently by the outsourcers and suppliers. On the other hand, synchronous channels, such as net meetings and video conferences are rarely used. To improve the efficiency of communications and to reduce unnecessary communication costs, we recommend that suppliers should avoid excessive synchronous communications. Face-to-face meetings and telephone meetings should be mainly used to discuss the business related issues and requirements. For issues related to design, development, and testing, email may work as an efficient communication type.

Manage the requirements changes and avoid unnecessary encouragement on overtime work

In software development, the requirements evolution can be substantial, e.g., 1% per month or 50% over 3 years [17]. Our results show that the main reason for overtime work in Chinese software suppliers are design or requirements changes initiated by the outsourcer. Another reason is the initial underestimation of effort, which is often caused by tactical bidding. In addition, we observed that there is an unnecessary and perhaps cultural encouragement for overtime work, which may cause employees to postpone their work from normal work hours to overtime hours. We recommend the Chinese software suppliers to appoint a change management group to systematically discuss and negotiate changes with outsourcers. Moreover, the Chinese suppliers need to improve their work efficiency in normal hours and try to avoid tactical bidding. An overall remedy to most of these issues is incremental development to reduce the risks for overruns, but this may be hard to combine with the existing outsourcer-supplier contracting process.

5.2 Threats to validity

Construct validity. In this study, most variables and alternatives are taken directly, or with little modification from existing literature. We did a pre-study to ensure the quality of the questionnaire, and nearly 10% of the questions and alternatives in the final questionnaire were revised based on the pre-study. With respect to RQ2, one possible threat to construct validity is that we asked only the respondents whether they used some types of communication on each topic or not (i.e., binary scale) without asking for the frequency of usage (e.g., how often do they use a certain communication channel). To investigate the RQ3, we have tried to use quantitative metrics, such as lines-of-code, to measure project outputs in the overtime hours. However, results of the pre-study show that most companies did not record such data. Our currently used subjective measurements may bring threats to conclusions of this research question.

Internal validity. We promised respondents in this study a final report or the annual membership of the CSO which worth of 500 Chinese Yuan. Most respondents took part in this survey as volunteers and selected the report as reward. We therefore generally believe that the respondents answered the questionnaire truthfully. However, a possible threat is that the respondents might have failing memory on past projects, because our unit of study is a finished project.

External validity.There were more than 11,550 software companies registered in China in 2005 [2]. However, the initial database we used contained only about a half of them. Although we have put a lot of effort on collecting data, we only get from 41 companies out of randomly selected 2000 companies. For the remaining 1959 companies, we could not know their reasons for not participating. The respondents answered the questionnaires based on finished projects, which were selected based on convenience. All the above issues may bring external threats to the conclusion of this study.6 Conclusion and future work

This paper has presented results of a state-of-the-practice survey on Chinese software outsourcing in industrial projects. The main conclusions of our survey are: •Differences in natural languages are not the critical factor for the success of Chinese software suppliers. Chinese suppliers have sufficient English skills and routines (e.g., steering groups) to resolve languages differences.

•Email is the most used communication channel for resolving development issues, while face-to-face meetings are popularly used for management and requirements issues. Net meetings and video conferences are not much used.

•The main reasons for overtime work at Chinese software suppliers are design or requirements changes and initial over-optimism of the effort or duration.

The current solution, i.e., unpaid overtime work, however, may reduce the efficiency in the normal work hours.

•Finally, since China has no comprehensive, national database of IT companies, it is difficult to select a random sample of participants in such surveys.

The results of this study discover several important issues that we are going to investigate in the future.

•We illustrate the most common communication channels without knowing their efficiency for discussing a specific topic. Further studies on the efficiency of different communication means are needed.

•We discover that using a steering group may be a proper strategy to mitigate the effect of languages or cultural differences. However, how to improve the performance of such steering group needs to be further studied.

•Although we find that overtime work is more efficient than work in normal hours, a further case study to verify our conclusion with a more reliable quantitative metrics is needed.

•Our results show that it is important to control requirements changes to avoid overruns of the outsourcing projects. How to use alternative processes, such as incremental or agile ones, in projects across geographic and company boundaries needs future investigation.

Acknowledgements

This study was a joint research effort between BJUT and NTNU, partially funded by the Norwegian SEVO project with grant 159916/V30. We would like to thank the CSO for data sampling and questionnaire collection. We also thank all participants in the survey.

References

1. Ahmed, R. E.: Software Maintenance Outsourcing: Issues and Strategies. Computers andElectrical Engineering, November (2006), 32(6): 449-453.

2. Ministry of Information of the People’s Republic of China & Chinese Software Industry

Association: Annual Report of China Software Industry. (2006): http://www.soft6.com/news/detail.asp?id=15759.

3. Aubert, B. A., Rivard, S. and Patry, M.: A Transaction Cost Model of IT Outsourcing.

Information and Management, September (2004), 41(7): 921-932.

4. Carmel, E. and Agarwal, R.: Tactical Approaches for Alleviating Distance in Global Software

Development. IEEE Software, March/April (2001), 18(2): 22-29.

5. Donahoe, D. N. and Pecht, M.: Are U.S. Jobs Moving to China? IEEE Transactions on

Components and Packaging Technologies, September (2003), 26(3): 682-686.

6. Ferguson, E., Kussmaul, C., McCracken, D. D. and Robbert, M. A.: Offshore Outsourcing:

Current Conditions and Diagnosis. Proceedings of the Thirty-Fifth SIGCSE Technical Symposium on Computer Science Education, Norfolk, Virginia, USA, March (2004), 36(1): 330-331.

7. Kankanhalli, A., Tan, B. C. Y., Wei, K. and Holmes, M. C.: Cross-cultural Differences and

Information Systems Developer Values. Decision Support Systems, November (2004), 38(2): 183-195.

8. Krishna, S., Sahay, S. and Walsham, G.: Managing Cross-cultural Issues in Global Software

Outsourcing. Communications of the ACM, April (2004), 47(4): 62-66.

9. Kshetri, N.: Structural Shifts in the Chinese Software Industry. IEEE Software, July/August

(2005), 22(4): 86-93.

10. Lacity, M.: Lessons in Global Information Technology Sourcing. Computer, August (2002),

35(8): 26-33.

11. Smite, D.: Global Software Development Projects in One of the Biggest Companies in

Latvia: Is Geographical Distribution a Problem? Software Process Improvement and Practice, January/February (2006), 11(1): 61-76.

12. Taylor, H.: Critical Risks in Outsourced IT Projects: The Intractable and the Unforeseen.

Communications of the ACM, November (2006), 49(11): 74-79.

13. Walsham, G.: Cross-cultural Software Production and Use: A Structurational Analysis. MIS

Quarterly, December (2002), 26(4): 359-380.

14. Herbsleb, J. D. and Grinter, R. E.: Splitting the Organization and Integrating the Code:

Conway’s Law Revisited. Proceedings of the 21st International Conference on Software Engineering, Los Angeles, CA, USA, May (1999): 85-96.

15. Wong, J.: China’s Economy in 2005: At a New Turning Point and Need to Fix Its

Development Problems. China & World Economy, March (2006), 14(2): 1.

16. Herbsleb, J. D., Paulish, D. J. and Bass, M.: Global Software Development at Siemens:

Experience from Nine Projects. Proceeding of the 27th International Conference on Software Engineering, St. Louis, Missouri, USA, May (2005): 524-533.

17. Madhavji, N. H., Fernandez-Ramil, J. and Perry, D.: Software Evolution and Feedback:

Theory and Practice. John Wiley & Sons, April (2006).

18. Moløkken, K. and Jørgensen M.: A Review of Surveys on Software Effort Estimation.

Proceedings of the 2003 International Symposium on Empirical Software Engineering, Rome, Italy, September/October (2003): 223-231.下载本文