Chapter 5

5. DISCUSSION

5.1 Hypothetico-Deductive Analysis

An essential component of hypothetico-deductive research is the comparison of the findings with the hypotheses deduced a priori from preliminary research. In this study, the preliminary study was an examination of secondary sources. Based upon these the hypotheses listed in Table 41 were deduced and were examined with respect to the findings described in Chapter 4.

It is again stressed that the measures used to determine the effectiveness of technology transfer when comparing different hypotheses were adherence to project start- time schedule, the project end-time schedule and to the project budget the composite value of the three, TTE (Technology Transfer Effectiveness). Absolute measures of cost like those measured by Teece (1977) are not used. The technology transfer effectiveness is not a true measure of the costs of transferring technology. It is a measure of how successfully the transfer was when compared to expectations, both schedule and budget.

Table 41. Hypotheses examined in the discussion of the research findings.

1. American firms are more effective at transferring technology than European firms.

2. Longer established subsidiaries transfer technology more effectively.

3. Firms with fewer employees transfer technology more effectively.

4. Subsidiaries with greater autonomy transfer technology more effectively.

5. Subsidiaries that have completed more previous transfers transfer technology more effectively.

6. Subsequent transfers of technology that has been previously transferred are more effective.

7. Technology transfer is less effective when more of the process, site and people are new.

8. More generic technology is transferred more effectively.

9. Companies that experience defections to competitors transfer technology more effectively.

10. Technology transfer is more effective when more Irish employees travel abroad to assist.

11. Technology transfer is more effective when more overseas employees travel to Ireland to assist.

12. Fewer trips abroad by Irish employees are needed for subsequent technology transfers.

13. Fewer trips to Ireland by overseas employees are needed for subsequent technology transfers.

14. Technology transfer is more effective when there is a key employee responsible.

15. Technology transfer is more effective when the links are strong.

16. Managers are more satisfied with the technology transfer process when it is more effective.

17. Technology transfer is more effective when both parties are committed to its success.

18. Technology transfer is more effective when employees of the transferee are more qualified and capable.

19. Technology transfer is more effective when training and documentation are better.

20. Technology transfer is more effective when companies apply business improvement initiatives.

21. Usage of the Internet improves technology transfer effectiveness.

22. Subsidiaries with longer established R&D facilities transfer technology more effectively.

23. Subsidiaries with more complex R&D transfer technology more effectively.

24. Companies which include employees of subsidiaries in product development teams transfer technology more effectively.

5.1.1. American firms are more effective at transferring technology than European firms.

The can-do attitude and American flair for innovation have seen American multinational corporations establish subsidiaries all over the world. They seem to be more effective at technology transfer that European companies. Some of the underlying reasons for this can be found in the studies of Hofstede (1993). Low power distance, high individualisation and low uncertainty avoidance have been associated with risk taking, a certain amount of which was considered necessary for the establishment of subsidiaries in countries with alien culture. However, in no studies available to the author was this theory examined. Ghoshal and Bartlett (1988) surveyed the diffusion of innovation by American and European multinational companies. However, they report no examination of the differences with respect to nationality.

The technology transfer effectiveness of American pharmaceutical and medical device companies was found to be significantly better than that of European companies (Section 4.3.1). Therefore the hypothesis was found to be true.

There was only data form one Japanese company available in the study. But the Japanese are not as strong in the pharmaceutical and medical device industry as they would be in other consumer goods. However, this one company did score very highly in the TTE. This is not enough information to state anything with certainty, although it is consistent with the good reputation that Japanese companies have for knowledge diffusion (Nonaka and Takeuchi, 1995).

5.1.2. Longer established subsidiaries transfer technology more effectively.

The role of duration on multinational investment strategy in the UK was found to be significant (Mudumba, 1998). Teece (1977) reports that older firms are likely to have lower costs. While neither of these points necessarily means that the technology transfer effectiveness will be better for longer established firms, it is felt that the balance of probabilities are that they will transfer technology more effectively.

The hypothesis was not found to be true. The TTE values follow no discernible pattern with respect to age (Section 4.3.2). There are several possible reasons for this. Firstly, the TTE value is a measure against expectations, not against absolute costs. Perhaps the newer companies are allowing for their relative inexperience by setting longer schedules and larger budgets than more established firms. Secondly, newer firms are in a technology transfer mode. They have probably hired personnel who will smooth the transfer process. Older companies may no longer employ all the key personnel who were responsible for the major technology transfers. The companies 6-10 years old had the worst TTE scores. Perhaps, the young companies benefit from being in start-up mode and the older companies form their experience and the companies of intermediate age are the ones that are less effective. However, the differences in the value of the TTE for the different age bands are not significant.

5.1.3. Firms with fewer employees transfer technology more effectively.

There are contradictory opinions as to whether large firms or small firms transfer technology more effectively. Teece (1977) reasons that larger firms will have a wider range of managerial talent and so will be better able to handle the demands of technology transfer. Smaller firms will be stretch managerially and the extra demands would be especially burdensome. However, Teece is considering the costs which may not be the same as technology transfer effectiveness. Smaller companies may make allowances for their lack of talent and allow for hiring consultant in the budget. The tacit barriers to the diffusion of knowledge that tend to be the main impediments to transfer of knowledge (Szulanski, 1996) would be a greater problem throughout a larger organisation. The author has experienced both the stretched resources of within a small organisation and the communications difficulties within a larger organisation. It is his opinion that the deleterious effect on communications within a large organisation would be a greater obstacle to technology transfer than the resource shortage in smaller organisations. The effect of TTE was measured against size measured in terms of number of employees and in terms of sales (Section 4.3.3. There is not a particularly strong trend in either set of data which would have any great statistical significance. The trends are that companies with higher sales transfer technology more effectively and companies with fewer employees transfer technology more effectively. The combination of trends may suggest that less labour intensive firms are more successful at transferring technology than more labour intensive industries. This is similar to the finding by Teece (1977) that the cost of technology transfer was lower in the less labour intensive chemicals and petroleum sectors than in the more labour intensive machinery sectors. It is not surprising when considering Szulanski's (1996) hypothesis that the major impediments to transfer of best practice are the people related stickiness factors. The result seems to be compatible with both Teece's and Szulanski's positions. Smaller companies with large sales will have sufficient money in the budget to smooth out transfer issues. Companies with fewer employees will diffuse knowledge more smoothly throughout the firm.

5.1.4. Subsidiaries with greater autonomy transfer technology more effectively.

Like the last hypothesis, an examination of the literature reveals findings that both support and contradict this hypothesis. Ghoshal and Bartlett (1988) proposed that high levels of local autonomy would facilitate creation and diffusion of innovation, but impede adoption. However, they did not find this in Phases II and III of their study. Teece (1977) found that a greater level of R&D at the subsidiary, which would suggest a higher degree of local autonomy, reduced the cost of technology transfer. The author opines that subsidiaries with a greater degree of local autonomy would be better able to react to difficulties with the transfer which will result in a better technology transfer effectiveness.

Subsidiaries with a greater degree of autonomy over human resources (pay and conditions, hiring and recruitment) and financial plans transfer technology more effectively (Section 4.3.4). There are no significant trends related to the other areas. Indeed, subsidiaries which have the least amount of autonomy over R&D and product specification have slightly more effective technology transfer, although the difference is not significant.

The trend with respect to HR and Finance matters supports the hypothesis that subsidiaries with a greater degree of local autonomy would be better able to react to difficulties with the transfer which will result in better technology transfer effectiveness. Because of the autonomy over decisions related to budgets, pay and recruitment they can react by hiring whoever or spending whatever to solve issues that arise during the course of the technology transfer process.

5.1.5. Subsidiaries that have completed more previous transfers transfer technology more effectively.

This hypothesis is well supported in the literature. Teece (1977) reports that the cost of technology transfer is reduced with experience. Zander and Kogut (1995) report that speed of transfer improves with experience.

However, this trend was not seen in the analysis (Section 4.4.1). The most and least experienced firms have the best TTE scores, especially due to scheduling components. However, there are only 4 firms with more than 20 transfers so the data are somewhat questionable. As the judgement criteria relate to performance against the planned schedule and budget and not the overall cost examined by Teece, it is not too surprising that start-up technology transfers are good when compared to later transfers. It is likely that more planning and resources were allocated to these technology transfers.

5.1.6. Subsequent transfers of technology that has been previously transferred are more effective.

This hypothesis is supported by the literature. The number of previous transfers of the technology was found to have an effect on the technology transfer performance (Zander and Kogut, 1995). Teece (1977) found that the cost decreases with subsequent transfer of technology.

The second transfer of a technology was found to be more likely to finish on schedule than the first transfer (Section 4.4.2). The overall effect on the TTE is improved although the result is not statistically significant.

This result that subsequent transfers are more likely to finish on schedule is consistent with the finding of Zander and Kogut (1995) that the number of previous implementations has an effect on the speed of transfer.

5.1.7. Technology transfer is less effective when more of the process, site and people are new.

Japanese firms will try to limit technology transfers so that at most only one of the process, site or people is new. For instance when establishing a car plant in Europe the company will install a process already developed and transfer as many people as necessary to ensure a smooth transfer. Effectively only the site is new.

The effect of this on TTE is shown in Table 16. The expected effect is not seen and the score was best for the cases where all three were new, although with only 5 data the score can be questioned. In any case the difference is not statistically significant. However, there were problems with the answers to the question in the questionnaires indicating that many respondent did not fully understand the question. Therefore, no conclusion of any value can be drawn.

5.1.8. More generic technology is transferred more effectively.

Zander and Kogut (1995) studied the relationship between the uniqueness of the technology and speed of transfer. Technology that is common throughout an industry is more easily transferred than unique technology. However, no apparent correlation was found in this study (Section 4.5.2).

5.1.9. Companies that experience defections to competitors transfer technology more effectively.

Zander and Kogut (1995) also examined defections to competitors expecting that firms in industries with a higher degree of mobility of employees between competitors were more likely to have speedy knowledge transfer. This is observed in this study (Section 4.5.3). A reason for this may be that the firms who have lost employees have better-trained employees who can move easily to other jobs. It would be interesting to compare the technology transfer effectiveness of firms who had hired employees from competitors with those who had not. There should also be a positive correlation. It is unfortunate that this question not asked.

Many companies are unwilling to train their employees because of fear of losing them once trained to competitors. This would seem to be an incorrect strategy. Companies who train their employees well not only gain the benefit of them while they are working for them. They are likely to be enthusiastic dynamic employees. They also receive the benefit of new blood with fresh ideas when employees who leave are replaced. While a high level of employee turnover is detrimental to morale and efficiency, some employee turnover is essential to maintain the vigour of an organisation (O'Keeffe, 2000). It is also noteworthy that only 14% of firms have lost employees to competitors. This indicates that there is not much mobility of employees. There has not been much change since Kelly (1985) studied technology transfer in the Irish chemical industry. He examined the effect of staff mobility and found that staff turnover was too slow to have the effect of diffusing technology from subsidiaries of foreign multinationals to the sector in Ireland as a whole.

5.1.10. Technology transfer is more effective when more Irish employees travel abroad to assist.

Travel and strong communications are important in the knowledge diffusion process. It is especially important for conveying tacit knowledge during product and process development. Nonaka and Takeuchi (1995) report that Nissan sent its entire product development team to Europe to experience European driving conditions when developing the Primera. Ghoshal and Bartlett (1988) found that strong communications between headquarters and subsidiary and between subsidiaries were important to aid the creation, adoption and diffusion of innovations within multinational corporations.

Therefore, it was expected that travel by more Irish employees as part of a technology transfer process would improve its effectiveness. However, only a marginal benefit was found (Section 4.6.1). In comments about recommended practices many respondents stressed the importance of strong communication (Section 4.10), although only 7 respondents used the word visit. Training which may or may not suggest foreign travel was also recommended.

A reason why travel by Irish people abroad may not be of great benefit to the pharmaceutical and medical device industries when compared to other industries is that the regulated nature of the industries do not allow changes to be made to processes without validation. All processes must be well documented. Therefore, the tacit component of the knowledge is not as important as in other industries where documentation requirements are slacker.

5.1.11. Technology transfer is more effective when more overseas employees travel to Ireland to assist.

Likewise, we would expect an improvement in technology transfer effectiveness with more travellers to Ireland by employees of headquarters, home-country manufacturing sites and employees of key equipment vendors. Indeed, in this case there is a marked improvement in TTE with an increasing number of travellers (Section 4.6.2). This effect is most especially marked on project start time. There is statistically significant benefit from sending five or more employees to Ireland when compared to sending fewer than five.

This benefit due to travel supports the findings made by Ghoshal and Bartlett (1988) that the positive benefits of normative integration through organisational socialisation improves a multinational company subsidiary's ability to contribute to different normative tasks.

But why is there a significant benefit to sending more travellers to Ireland and there is not when sending travellers from Ireland? The benefit was only significant on the project start schedule. There was some marginal benefit on the end schedule. The effect was marginally deleterious on adherence to the budget. Therefore, it is probable that the project starts more quickly because the travellers are more knowledgeable and can 'hit the ground running', so to speak. It is also possible that they are working primarily on the early stages of the process and the sooner they complete their tasks, the sooner they can go home.

5.1.12. Fewer trips abroad by Irish employees are needed for subsequent technology transfers.

This hypothesis is similar to the hypothesis that technology transfer effectiveness will improve with experience of conducting transfers (Kogut and Zander, 1995). That hypothesis was not found to hold in this study (Section 5.1.5) and this hypothesis does not hold either. Indeed, more respondents reported an increase in the amount of Irish employees travelling abroad when compared to previous transfers (Section 4.6.3). The reasons are hard to understand without more information. Perhaps, the scale and complexity of the technology projects being transferred to Ireland is increasing. This would require larger number of people to travel to absorb the tacit knowledge necessary for these more complex projects. None of the travel data takes the complexity of the projects into account. This is consistent with the findings of Mudumba (1998) concerning the duration on multinational investment strategies. Also, more Irish people travelling abroad may indicate that firms now employ more indigenous employees than before who are capable of transferring complex technology. It is a good thing that there are more Irish based employees travelling abroad to absorb the tacit knowledge and improve the value of the Irish operation. Nonaka and Takeuchi (1995) emphasise the importance of travel on the new product development process.

But the minority of companies that did report fewer travellers transferred. This could be because less money was spent on travel. It may also indicate that the firms with better organisational learning, both reduce the requirement for employees to take trips abroad and also learn better methods of transferring technology. Perhaps much of the travel requirements is to deal with problems that arise during the transfer process. Companies that require less travel, and indeed a reducing amount of travel are probably those who have the best developed transfer procedures.

5.1.13. Fewer trips to Ireland by overseas employees are needed for subsequent technology transfers.

This hypothesis does not hold either. The same number of companies report an increase in the amount of travel to Ireland by overseas employees as report a decrease (Section 4.6.4). So while the number of companies receiving more travellers in not increasing as in last case relating to Irish employees travelling abroad, it is not decreasing either. The explanation may well be that the complexity of projects being transferred is increasing and for that reason it is not possible to reduce the number of travellers significantly. The same trend is seen with respect to the TTE score for the companies that require fewer travellers from abroad. This supports the earlier contention that in some cases the Irish subsidiaries are becoming much more competent at transferring technology. As such they are transferring the technology more effectively and requiring fewer visitors form abroad to assist.

5.1.14. Technology transfer is more effective when there is a key employee responsible.

Nonaka and Takeuchi (1995) propose that their American Football model is the most effective for new product development. In this model an expert project manager calls on the requirements of the expert teams needed to complete various stages of the project. The same model may be applied to technology transfer projects.

However, companies that do employ a key individual to co-ordinate technology transfer projects are not more effective than those who do not (Section 4.6.5). Indeed, technology transfer is marginally more effective for companies that do not, although the result is not statistically significant. Again complexity of projects may be the explanation. Companies that require a technology transfer specialist are those that are transferring more complex technology. Or alternatively the explanation may be that companies that have a key employee are those who are starting to transfer technology. Once companies become more experienced they have several employees capable of managing a technology transfer project and no longer require a key employee to coordinate.

5.1.15. Technology transfer is more effective when the links are strong.

Hansen (1999) found that having weak inter-unit links speeds up projects when knowledge is not complex, but slows them down when the knowledge to be transferred is highly complex. He supports the weak-tie theory proposed by Granovetter (1973) that distant and infrequent relationships (weak ties) are efficient for knowledge sharing because they provide access to knowledge by bridging otherwise disconnected groups and individuals in the organisation. Strong ties on the other hand are likely to lead to redundant information because they tend to occur among the same small group and everyone knows what the others know. However, for complex information transfer strong links are needed.

So is the transfer of technology to Irish subsidiaries of multinational pharmaceutical and medical device companies generally not so complex that they would benefit from weak links or complex requiring strong links? Given the complications that can result from the validation procedures, the author feels that on balance most of the technologies transferred are complex requiring strong links with the main R&D site. Whether needed or not, it was found that the majority (75%) have strong links with the main R&D site (Section 4.6.6). Most of the remainder have weak links and one respondent reports no links at all. There was no difference in technology transfer effectiveness between the companies with strong and weak links. Therefore the hypothesis has not been found to be true. The information available is not sufficient without more information about the complexity of transferred technology.

5.1.16. Managers are more satisfied with the technology transfer process when it is more effective.

Szulanski (1996) found that the greatest impediments to internal knowledge transfer were factors such as the recipient's lack of absorptive capacity, causal ambiguity and arduous relationship between source and recipient. He used satisfaction to measure these factors. For this reason there should be a correlation between technology transfer effectiveness and satisfaction with the transferred technology. A comparison of satisfaction data with technology transfer effectiveness is reasonable well correlated (Section 4.7.1). The reason for this relationship is intuitive. If the process is transferred on time and on budget, the Irish management is more likely to be satisfied than dissatisfied.

5.1.17. Technology transfer is more effective when both parties are committed to its success.

Szulanski (1996) found that transfer of best practice within a firm is more effective when both sides are committed to the process. The was a slight benefit to the technology transfer effectiveness when there was a good working relationship between transferor and transferee and when the levels of commitment to the process were highest. (Section 4.7.2). Only 10% reported that the Irish management was anything less than very committed to the process. This level of commitment is not unexpected but is especially important for successful knowledge transfer (Szulanski, 1998; Ghoshal and Bartlett, 1988).

The level of commitment of the exporting site was not as high. This is not unexpected as in some cases the transfer of technology to Ireland may equate to a loss of jobs in the technology exporting country. Just under 50% were very committed and greater than 80% were committed or very committed. There is a slight difference in technology transfer effectiveness depending on whether the exporting site is very committed or just committed.

5.1.18. Technology transfer is more effective when employees of the transferee are more qualified and capable.

Szulanski (1996) found that the one of the impediments to internal knowledge transfer was the recipient's lack of absorptive capacity. To examine this effect respondents were questioned about the level of qualification of the Irish workforce and the managerial capability. All answers except three to each question were answered as strong or very strong. The technology transfer efficiency is marginally better for those expressing the highest level of qualifications and managerial capability, but the differences are not significent. Therefore there is insufficient information to determine whether the hypothesis is valid or not.

5.1.19. Technology transfer is more effective when training and documentation are better.

Another conclusion that can be drawn from the study by Szulanski (1996) is that technology transfer is more effective when the training and documentation are better. Training and documentation are essential components of knowledge transfer process and are the major modes of knowledge conversion. Training is a method both of socialisation, the mode by which tacit knowledge is transferred, and of internalisation, the conversion of explicit knowledge to tacit knowledge (Nonaka and Takeuchi, 1995). Documentation is a method both of externalisation, the conversion of tacit knowledge to explicit knowledge, and of combination, the transfer of explicit knowledge. Both are essential to effective technology transfer.

The ease of training and the quality of training are consistent with improved technology transfer effectiveness (Section 4.7.4). The improved technology transfer effectiveness is mainly the result of better scheduling, both start time and finish time. The latter can be intuitively understood because being able to complete training speedily will allow the project to completed in good time.

The correlation between quality of training and good adherence to the start schedule is less clear. Perhaps, when the project starts in good time, it allows more effort to be spent on training that might otherwise have to be spent on other matters. There is no obvious correlation between the technology transfer effectiveness and the quality of documentation. This would suggest that effective socialisation and internalisation are more important to the technology transfer process than effective externalisation and combination.

5.1.20. Technology transfer is more effective when companies apply business improvement initiatives.

In a recent study Quazi and Bartels (1998) studied the application of TQM principles to international technology transfer processes. They found that a combined ITT-TQM (International Technology Transfer - Total Quality Management) should ensure proper assimilation of imported technology and permit continuous improvement of the processes. It is hypothesised that those companies applying TQM to the technology transfer process will transfer technology more effectively. It is further surmised that the application of other business improvement method described in Section 2.4 are also likely to benefit the technology transfer process because of the objective of most is to improve the way business is done. Since technology transfer is an essential business function for multinational healthcare and pharmaceutical companies there is likely to be a benefit. In this study, those surveyed were asked whether they had applied TQM or 6?, Business Process Engineering, Benchmarking, Integrated Technologies or Lean Thinking to the technology transfer process. The effect on technology transfer effectiveness was examined (Section 4.8). In the analysis around 50% of firms had applied TQM or 6? and benchmarking to technology transfer. There was significant improvement to technology transfer effectiveness. Fewer firms applied the Integrated Technologies, Lean Thinking or Business Process Engineering and those who did have more effective technology transfer. An improvement in technology transfer effectiveness was most marked for those who applied Business Process Engineering and Integrated Technologies (WCM) although the effect is not that significant. Both methods involve re-examining the way things are done and implementing new improved methods. It is impossible to know if the firm's TTE score was better because these methods had been applied to technology transfer, or because the more innovative firms who try all new ideas are the ones that have better developed knowledge transfer techniques.

There was very slight benefit from the application of lean thinking. It is interesting to note that the improvement is all due to better adherence to the budget. There is insufficient evidence to be completely convinced about the benefits of application of the various business improvement initiatives.

5.1.21. Usage of the Internet improves technology transfer effectiveness.

The use of the World Wide Web for international technology transfer was studied Hoetker (1997). He highlighted the fact that some difficulties associated with international technology transfer are (1) the lack of familiarity and contact makes it more difficult to find technologies and people; (2) language differences; (3) logistics of communication; and (4) cultural differences. Many of these difficulties are overcome by the usage of USENET and WWW on the Internet.

Therefore the effect of usage of the Internet on technology transfer effectiveness was assessed (Section 4.8.2). The responses were rated extensive use, limited use or no use. There is no apparent effect due to Internet usage on technology transfer performance, TTE. Firms that made limited usage of the Internet actually transferred technology slightly better than those that made either extensive usage or no use. However, the differences are not that significant. The reason that more benefit was not seen may be that companies in industries other than computers and electronics have still not completely embraced the benefits of this new technology. Also, the proprietary nature of a lot of technology transferred with the pharmaceutical and medical device industries may prevent there being much information of use on the Internet that can assist with technology transfer.

5.1.22. Subsidiaries with longer established R&D facilities transfer technology more effectively.

Teece (1977) reports that subsidiaries of multinational companies which spend a greater amount as a proportion of sales on research and development have lower technology transfer costs. A greater amount of R&D conducted is consistent with have a more knowledgeable workforce better capable of absorbing new technology. On the other hand Ghoshal and Bartlett (1988) proposed that high levels of local autonomy would impede adoption of innovation. However, they did not find this in Phases II and III of their study. The effects of two factors concerning R&D on technology transfer effectiveness were examined in this study. The first was the effect of the presence of R&D facilities and the duration since their establishment (Section 4.9.1). There is no effect so the hypothesis does not hold.

However, what is interesting is that over half of the companies now have R&D facilities in Ireland and the majority of these have been established in the last five years. This indicates that the abilities of personnel at Irish manufacturing operations is being recognised by their parents. This may be the result of many talented returning emigrant and immigrant who have joined the Irish workforce in recent years. The author is aware that his own employer (not included in the survey) has established an R&D facility that has developed products for worldwide sale in the last five years. While anecdotal this observation supports that result that 54% of multinational healthcare and pharmaceutical companies have now established R&D facilities in Ireland, many of them within the last five years.

5.1.23. Subsidiaries with more complex R&D transfer technology more effectively.

The second question asked about the effect of complexity of R&D on technology transfer (Section 4.9.2). Again there is no apparent benefit to conducting R&D on technology transfer performance. And again the most interested findings relate to the extent of R&D in Ireland.

The R&D complexity gradings used by Taggart (1998) in his study of the complexity of R&D in affiliates of multinational firms in the UK are used and compared with technology transfer effectiveness. The ability to adapt manufacturing processes to local needs (2) and capability to develop new and improved products and processes for worldwide use (4) are the most popular categories for those conducting firms who conduct R&D. Because the vast majority of multinational firms operating in Ireland export most of their production the categories basic customer technical services (1) and capability to develop new and improved products and processes for domestic use (3) are not important. Only four firms carry out the most advanced R&D (5). The next category down (4) had the poorest TTE score and it is considerably worse than companies conducting no R&D. Therefore, the hypothesis that firms conducting more complex R&D have more effective technology transfer processes does not hold.

Fewer firms do no R&D (18) than firms that have no R&D facilities (26). This indicates that some firms conduct research without specific facilities. In all 68% of firms conduct some R&D. Of these 44% of companies carry out R&D in the top three categories.

5.1.24. Companies which include employees of subsidiaries in product development teams transfer technology more effectively.

Subramaniam et al. (1998) studied the usage of global product development teams. They found that firms were more likely to employ cross-national teams and include overseas subsidiaries as sources of new product concepts when they addressed tacit differences among overseas markets and plants for their overall product design. Because of this emphasis on tacit knowledge it is hypothesised that subsidiaries with personnel involved in global product development will have better tacit links with R&D sites, and as a result will transfer technology more effectively.

The majority of firms use such teams and there are Irish members on most of these teams. The technology transfer effectiveness is slightly better for companies that use global teams and are better again when there is Irish participation. However, the results are not statistically significant. This demonstrates that the inclusion of Irish based employees in product and process development may marginally improve the efficiency of the technology transfer process, although the relationship is not strong enough to be certain that the hypothesis is true.

Technology transfer effectiveness was compared with the four different product development processes proposed by Nonaka and Takeuchi (1995) in their study of knowledge creating companies. The rugby approach is most popular in the medical device and pharmaceutical industries. The relay approach is not popular in the industries studied where more overlap of different stages is required. The one result here is not enough to draw a meaningful conclusion. Companies using the more modern American Football method in which a project manager co-ordinates the various 'special teams' score slightly better than companies using the more traditional rugby approach. This approach is recommended by Nonaka and Takeuchi (1995) because it combines the best aspects of the rugby and relay approaches, i.e. the continuity of having a core individual overseeing all stages combined with the usage of specialised teams to perform all the actions.

5.2. Qualitative analysis

A limited qualitative analysis was conducted by asking participants to list practices and methods which they found aided the technology transfer process. They were also asked to list practices, methods, or the absence of practices which hindered the process.

5.2.1. Recommended practices and methods

Forty-eight respondents provided an answer to the question which asked for a list of specific practices or methods which aided the technology transfer process. Twenty-eight answered the questions which asked for a list of specific practices or methods, or absence of practices, which hindered the technology transfer process. The results are summarised in Section 4.10 and the full list of responses is given in Appendix C.

The practices that were considered most important overall by the respondents are project management/planning, teamwork, communications (also meetings and visits), training, clear procedures and documentation.

Other practices of interest mentioned a few times are conferences calls, validation procedures, benchmarking, information sharing, the use of existing production methods, contract with specialist equipment companies, supplier management, risk analysis, ISO9002 and good on-site management.

5.2.2. Practices and methods not recommended

Fewer respondents (28) noted practices that hindered and the string analysis used to analyse the best practices was not useful. There was a greater variety of problems mentioned than good practices, even with around half the number of responses. Validation was the issue that received most mentions, but this was only in 10% of repsonse. In general the problems can be grouped together and are opposite to the good practices most frequently mentioned. Good planning, communications, teamwork, training, procedures and documentation would prevent most of the problems mentioned from occurring.