PART 1 – The Empirical Picture
6.4 Involvement of respondents in e-infrastructures
6.4.5 Catalysts of and barriers to involvement
Catalysts of and barriers to involvement were asked in open-ended questions which permitted that the respondents themselves assessed and prioritized the most important influences on the adoption process. These responses were then coded independently by experts from the project team. Examples for the answers and the respective codes are shown in Table 6-25.
Table 6-25: Examples for answers on catalysts and barriers
Catalysts Barriers
Access to resources
− Access to a larger distributed network than available locally
− Sharing of data across multiple institutions
− Additional resources available
− Computer resources assigned to DEISA
− Reasonable existing local resources
− Already have access to other (much larger) resources in the US
Organizational − Enthusiasm of most stakeholders
− Collaboration among scientists
− Job requirement
− Developing high level analysis services for research that requires industrial-strength organization of computation flows
− Good infrastructure and organization
− Support from colleagues at UCSD, UvA, USC, Lucasfilm, Sony, Keio University, NTT Labs
− No support for radio astronomical data
− Grid infrastructure changed often, and a few changes to my application were needed as a result sometimes
− EU legal constraints not compliant with my institution's requirements
− Lack of support from my institution
− Low administrative pressure to stimulate the use of these tools
− Bureaucracy Technical
capabilities
− Need to bridge interoperability gaps among communities of practices
− Reporting tool
− Computing Power and Fault Tolerance capability
− Possibility to use state of the art technology
− Research interest on grid technology and remote instrumentation
− MT may only improve by having machines learning differently from humans
− It is not easy, in basic research, to make detailed statements on how much CPU time will be needed to complete a project.
− Time required to adapt usual workflows to DEISA
− Lack of structure to support anonymous access
− Download and Installation of applications
Ease of use − User-friendliness
− Easy application process
− Availability & reliability
− Easy writing and uploading project
− Interface
− Slow to get to compared to other resources
− Difficult to use in the beginning Funding-related − Funding
− Continuous funds to guarantee continuous research
− Outsourcing infrastructure
management and maintenance costs
− Developing fundraising and governance structure
− Securing national (matching) funding
− Cost of network infrastructure
− Insufficient funds
Page 165
Catalysts Barriers
− The grant of the financing institution
Training-related
− Technical support and training
− Need of HEP communities in Latin America to create support infrastructure
− Time spent to get the application compiled and running
− Learning curve
− Lack of background in grid computing
− Not known by individual researchers
− Learning material is good, but sparsely distributed through the web Other
catalysts/
barriers
− Need for European A&H e-infrastructure
− Personal interest
− Desire to help the researchers
− Part of my work
−
As our sample is not representative in any way, these results do not allow for any conclusions on the influences on adoption in general. We see that access to resources was the most important catalyst, mentioned by 28% of the respondents (see Figure 6-25); organizational factors and the technical capabilities of the e-infrastructure were also listed frequently among the catalysts. The latter were also the most important barriers that had to be mastered in the adoption process, together with organizational barriers and low usability of the e-infrastructure.
Figure 6-25: Respondents by catalysts and barriers (in %)
28.0%
Distinguishing the catalysts and barriers by continents of the respondents we get some notable differences (see Figure 6-26 and Figure 6-27): North-Americans rated access to resources as the most important catalyst, Latin Americans organization-related issues and Asians technical capabilities. Organizational, training- and funding-related barriers were particularly important among Latin-American respondents. Among North-Americans technical capabilities, usability and organizational barriers were the most important. European and Asian respondents mentioned barriers less often than respondents from the American continent.
Page 166 Figure 6-26: Catalysts by continent (in %)
0%
25%
50%
Access to resources
Organizational catalysts
Technical capabilities
Ease of use Funding-related catalysts
Training-related catalysts Other catalysts
Europe North-America Latin America Asia
Figure 6-27: Barriers by continent (in %)
0%
25%
50%
Access to resources
Organizational barriers
Technical capabilities
Ease of use Funding-related barriers
Training-related barriers Other barriers
Europe North-America Latin America Asia
If we regroup respondents by the development level of their country we also get some interesting differences (see Table 6-26): Organizational, funding and training issues are mentioned both as catalysts and barriers more often among the developing countries.
Technical capabilities and usability are more often barriers in developed countries than in the developing world.
Page 167 Table 6-26: Catalysts and barriers by development level of the country (in %)
Catalysts Barriers
Technical capabilities 20.9% 23.3% 11.8% 23.3%
Ease of use 4.5% 7.1% 11.8% 18.9%
Funding-related 8.2% 5.7% 14.5% 5.7%
Training-related 24.5% 10.8% 19.1% 8.1%
Other
catalysts/barriers 6.4% 4.4% 12.7% 9.5%
Next we also find some differences between the responses on catalysts and barriers if we categorize the respondents according to their institutional affiliation (see Table 6-27). The differences are particularly strong for technical capabilities and usability. These appear – both as barriers and catalysts - as much more important for respondents from government and international organizations. Training-related catalysts and barriers were more often mentioned by respondents from academia.
Table 6-27: Catalysts and barriers by institutional affiliation (in %)
Catalysts Barriers
Catalysts and barriers
Academia Government and international org.
Comparing the catalysts and barriers between the types of e-infrastructure which we
distinguished – national versus international, disciplinary versus multidisciplinary, computing versus data e-infrastructures, and developer- versus community-driven – there are only few differences (see Table 6-28 in the annex). In particular for technical capabilities we get some notable differences: They were mentioned more often as catalysts in connection with
national, multidisciplinary and computing infrastructures. As barriers they were mentioned more often in connection with national, disciplinary, data and community-driven
infrastructures. Organizational catalysts were the dominant catalyst fro respondents involved in developer-driven e-infrastructures. Usability and training were more often mentioned as problems for becoming involved with a computing infrastructure.
Several catalysts have become more important over time and for those who joined an e-infrastructure rather late (see Figure 6-28): Access to resources, organizational catalysts, usability and funding issues were more often mentioned by those who became involved three or more years after the project had started. For barriers we would expect an opposite trend,
Page 168 namely that they become less important over time and those who become involved in later phases encounter fewer barriers. However, this is not supported by the responses (see Figure 6-29). It is not possible to identify any clear trends; only technical capabilities were clearly mentioned more often as barriers by newcomers to the infrastructure than by those involved from the start.
Figure 6-28: Catalysts by start of involvement with the selected e-infrastructure (in %)
0%
20%
40%
Access to resources
Organizational catalysts
Technical capabilities
Ease of use Funding-related catalysts
Training-related catalysts Other catalysts
Involvement from the start Involvement 1-2 years after project start Involvement 3-5 years after project start Involvement > 5 years after project start
Figure 6-29: Barriers by start of involvement with the selected e-infrastructure (in %)
0%
20%
40%
Access to resources
Organizational barriers
Technical capabilities
Ease of use Funding-related barriers
Training-related barriers Other barriers
Involvement from the start Involvement 1-2 years after project start Involvement 3-5 years after project start Involvement > 5 years after project start
Page 169