3
Issues in Initiating, Organizing, and Planning International Collaborations
The specific challenges involved in initiating and organizing collaborations constituted the second major set of issues that the workshop examined, with participants presenting both European and U.S. perspectives. These issues arise because international collaborators necessarily come from different cultural backgrounds and intellectual traditions, work within differing national bureaucratic and legal contexts, and often belong to different academic disciplines, which can greatly complicate the task of agreeing on research concepts and methods. Obtaining funding for international research can also be extremely complicated, both because of differing national requirements and policies and because working in an international context generally takes more time and therefore more money than research within a single country. In addition, the many sources of difference among international collaborators can create countless opportunities for misunderstanding, disagreement, confusion, and even conflict among colleagues. Session 2 focused on practical methods of dealing with these problems that might be adapted for particular settings.
Several of the workshop speakers and many other participants agreed that, because of the many potential areas of misunderstanding, international collaborations need to work out detailed agreements at the outset of each project on as wide a range of scientific and operational questions as possible. Formal agreements, however, do not suffice to assure the success of international projects. Strong, trusting, and resilient personal relationships form a more reliable glue for international research groups, speakers and other participants emphasized, advising that collaborators work constantly and systematically to foster them.
Even with apparently comprehensive formal agreements and strong personal bonds, however, cultural differences still remain that can influence both the scientific and personal aspects of the collaboration. Attention and sensitivity toward cultural issues must continue throughout the research, publication, and dissemination process. This chapter covers each of these issues in greater detail.
AGREEING AT THE OUTSET
The need for collaborators to reach explicit agreements on a number of scientific and operational issues received major attention at the workshop. Because of the need for highly detailed agreements, collaborative international projects in Europe can become extremely bureaucratic, said Martyn Barrett, emeritus professor of psychology at the University of Surrey in the United Kingdom. His remarks drew on his experience leading three such projects, especially one entitled “Processes Influencing Democratic Ownership and Participation” (PIDOP) that was carried out between 2009 and 2012 by teams from nine countries with funding from the European Commission (EC). Research done under EC auspices requires a formal consortium agreement drawn up by university lawyers and completed before any money is released, he said. This legally binding document, prepared by the coordinating institution and signed by the partner institutions, controls many aspects of the research project and how it is run.
Topics covered in such agreements include the internal organization and governance of the consortium, how the EC’s financial contribution is going to be distributed, management of intellectual property rights and access rights, and the liability and confidentiality arrangements between the partners. All research projects are highly directed; they must be multidisciplinary and policy relevant, supporting policy development at multiple levels. Beyond that, projects must have a European dimension that adds value over and above what a national project could achieve.
Although the process of organizing such agreements can be less formal in the United States, the need for clarity concerning such issues as ownership of data, procedures for sharing it, and the funding of project elements such as data collection, analysis, conferences, and meetings of collaborators is every bit as great, Charles Super noted.
Key points in such agreements should, from the start, make the project truly international and make it clear whether the project has one director or is a federation of equal partners, suggested Joseph Tobin, Elizabeth Garrard Hall Professor of Early Childhood Education, University of Georgia, in written remarks read to the workshop. “Too often,” he continued, “the international project begins with a research team from one country who makes up the research questions and then gets other countries to sign on. This may give the project greater consistency, but it’s at a cost because the variables to be considered are very often those assumed to be important in just one or two of the countries. A better approach is for a small group of researchers from several countries to collaboratively develop the research questions, conceptual framework, and methods, including the variables to be studied. This allows for more potential to challenge taken-for-granted assumptions, and to provide fresh insights.”
As to leadership, the single-leader and federated leadership models each has its advantages. A single leader simplifies coordination, but at the risk that one researcher’s perspective will dominate the project. Overall, “international research requires compromises on framing questions, site selection, participant recruitment, and research methods, for example, culturally responsive modifications in how interviews are conducted,” Tobin noted. “The group needs to agree from the start what to do when significant differences of opinion arise.”
Choosing teams with complementary capabilities contributes significantly to success. Barrett noted that EC practices require justifying the inclusion of each team and specifying the significant contribution it will make. Beyond that, he explained, the EC system also requires dividing the research into distinct work packages and specifying a verifiable endpoint for each one, generally a work product with an established deadline. Though complex, the process of dividing work into packages protects the overall project by limiting the potential damage should any partner fail to fulfill the agreement. Barrett observed that international research, already difficult, can be more so when collaborators have different levels of ability and infrastructure. These detailed agreements with their distinct work packages can allow one partner to move forward with their tasks if another collaborator runs overtime or fails. Beyond that, some partners may choose to participate in only those aspects of the project best suited to their expertise and circumstances.
To assure that each collaborating group can carry out its assigned tasks, Barrett continued, assembling a collaboration also calls for an audit to evaluate each team’s expertise and the ethics procedures that govern its work. A separate audit checks the resources available to each team. One project, he noted, had to budget for an electrical generator for a partner in the Republic of Georgia, where frequent power cuts presented the risk of losing work and data.
Budgets may also have to compensate for problems arising from particular funding practices, such as the cost-reimbursable systems used by some funders, said Jennifer Lansford, research professor at the Social Science Research Institute of Duke University. She described the difficult situation created when institutions are expected to spend funds to get projects started and then wait for reimbursement. U.S. institutions often have enough financial backing to provide for reimbursable start-up funding, a luxury that universities in less affluent countries may not enjoy.
In addition to the collaborating teams of researchers, Barrett noted, nonacademic stakeholders often also play crucial roles in policy-relevant research projects and therefore also need to be involved from the outset. His PIDOP project used several methods to accomplish this, he continued, including compiling a directory of stakeholders early in the project and keeping them informed through policy briefing papers and periodic newsletters as work progressed. The researchers also nurtured close relations with a small group of
key stakeholders, seeking their opinions on issues during the design phase and on policy recommendations at the end.
THE ROLE OF RELATIONSHIPS
Despite the high level of formality required for EC collaborations, the actual process of recruiting the participants in research consortia very often rests on unofficial considerations, Barrett continued. Especially central is the need for positive preexisting relationships among the members of the different teams. For that reason, he has always based his collaborations on relationships established through networking. Communication needs to remain open and positive throughout the project, and it is helpful if all participants are comfortable with their assigned duties.
Many participants concurred on the centrality of personal ties and connections in finding and enlisting collaborators, as well as in fostering any project’s success, although some also mentioned that cultural differences among collaborators may at times present challenges. Nor have researchers in more affluent countries always recognized the importance of cultivating relationships. Super observed that in the post-World War II period cross-cultural research generally followed the lines of the dissolving empires, particularly the British Empire, with “safari research” in which investigators went abroad, collected their data, and returned. Attitudes toward collaborating with colleagues in less affluent countries have significantly improved in recent times, he noted.
Over time, the importance of working with a local collaborator who could serve as a cultural guide became increasingly evident, especially under the influence of anthropological methods, Super continued. Researchers became aware that the inside perspective tells something different from the outside perspective. The cultural guide could be anyone from a local professional to a local person with minimal education. Graduate students from the culture in question have also often served in that role. More recently, however, research has arrived at “an interesting phase where your collaborators really are [of] equal status in many cases, and in some cases senior status to you,” opening up the possibility of long-term collaborations, Super noted.
Researchers with bicultural backgrounds who are actually participants in both cultures may face special issues, such as navigating two different cultures while maintaining relationships in both, noted Sonia Suchday. Although one may understand the rules of both cultures in their contexts, such researchers have the unique challenge of representing one culture to the other, as an insider to both. This presents many additional difficulties and challenges.
FINDING COLLABORATORS
A major challenge of collaborating across national lines is the potentially daunting task of identifying potential collaborators and establishing relationships
with them. Beyond the methods already mentioned in this report, George Alter, director of the Interuniversity Consortium for Political and Social Research (ICPSR) at the University of Michigan, suggested that working with data can in itself provide fruitful opportunities for finding international collaborators. He drew on his experience with the ICPSR, a data archive whose mission is to acquire, archive, distribute, and preserve data for future generations. It also provides training in quantitative methods. Founded in 1962, it has grown from 21 member universities to more than 700 today, 400 of them in the United States. It also has more than 40 federation or national memberships, which bring in other international members as well. ICPSR also collaborates with several data sharing initiatives that cross national borders.
Combining data with training is a proven method of building collaboration around data, he said. A partnership between the University of Michigan and the University of Cape Town in South Africa, for example, provides a 2-week course on analyzing data given at Cape Town and taught by instructors from both institutions. In addition to training researchers in South Africa, the program identifies people who could benefit from advanced training and invites them to Ann Arbor as graduate students, to the ICPSR summer program, or to the Survey Research Center. This has resulted in an ongoing source of potential collaborators for joint projects. Starting in 2011, the course is also offered at a second site in Ghana.
A second strategy for building collaborations around data involves projects that coordinate data collection by international collaborators. Alter gave the example of the Barometer studies of public attitudes toward democracy and government that are designed to collect data and also build local capacity in Africa, Europe, the Arab Countries, Asia, and Latin America.
Online technology, especially cloud computing, forms the basis for a third strategy for creating opportunities to work with colleagues in less developed countries. Such partnerships can succeed even in places with less technological abilities, Alter said. One example involves a university in South Africa where a political science instructor had his students do an exercise on the web using data that was provided by the University of Michigan.
International collaborations built around data can, however, also harbor pitfalls, such as the temptation to value skills over local content. In addition, a kind of “data imperialism” may arise if researchers from developed countries tell colleagues in developing countries, “Let’s work together. First of all, you send us all your data.” To overcome this attitude, Alter observed, would-be collaborators must remember that many very good researchers live outside the United States, and we need to work together and collaborate in a true sense.
THE REALITY OF CULTURE
The need to take cultural differences into account not only in designing research but also in working with colleagues during the project was another major concern of the workshop participants.1 Doing this can be difficult to accomplish, however, because “people still tend to think culture is just a veneer, something that is not about basic psychological processes,” said Michele Gelfand, professor of psychology at the University of Maryland. Her own work emphasizes that this is not the case, but still, “theoretically speaking, we haven’t convinced psychologists that [culture] is a very important topic from a theoretical point of view.” Many people, furthermore, lack the training needed to do cross-cultural work. Cross-cultural research methods must be taught to make researchers more fluent in the methodology.
Her own experience in international research shows that culture infiltrates the research process. If one does not understand that culture affects behavior in the laboratory or field, data can be misinterpreted. Gelfand noted that culture shapes the questions researchers ask and the methods they use. Even those most aware in cross-cultural psychology in the United States still may ask questions that are biased in their orientation.
Suchday gave the example of cultural difficulties she has faced in dealing with colleagues in India while coordinating plans for U.S. students to travel there. Administrators in the two countries have different academic calendars, different sensitivities to time, and different feelings of urgency about when decisions must be made, often resulting in annoyance and confusion on both sides.
As Gelfand observed, it is important to recognize that all of us are cultural novices in regard to methodology each time we do cross-cultural research. To accomplish this preparatory work, one must partner with and rely on local collaborators as equals and test the appropriateness of the question and methods to the particular cultural context by first doing a lot of pilots and focus groups. Working with local collaborators also helps avoid making embarrassing mistakes and wasting valuable time.
In Arab countries, Gelfand noted, local collaborators pointed out that using money as an incentive or reward for subjects, although routine in the United States, is inappropriate because people see something fundamentally dishonorable about accepting money to participate in research. Instead, she and her colleagues give out blankets or gift certificates.
_________________
1 The Government-University-Industry Research Roundtable (GUIRR) at the National Academy of Sciences held a workshop that focused specifically on the role of culture in international research collaboration: National Academy of Sciences, National Academy of Engineering, and Institute of Medicine. Culture Matters: International Research Collaboration in a Changing World—Summary of a Workshop (Washington, DC: The National Academies Press, 2014). Available at http://www.nap.edu/catalog.php?record_id=18849.
Americans are well known for being decontextualized and answering questions about relationships in the abstract, she continued. In Taiwan, on the other hand, research subjects often say that they need more information about relationships and the social context of interactions before they can answer questions or participate in experiments concerning them. Americans accept the notion of participating in role-playing exercises and then leaving that behind in the lab. In the Middle East, Gelfand has found that if you violate someone’s trust, even in the lab, it is not as readily forgotten. In planning cross-cultural research, it is therefore crucial to triangulate research methods—that is, to check them against other methods to see if they produce comparable results—because every research method has its flaws and cultural baggage. Finally, researchers need to remember that culture is operating in all of our collaborations. Even when the questions and training are similar, there are cultural differences among researchers. Gelfand urged collaborators to talk often, truly listening to each other, to manage expectations about research projects.
Barrett observed, however, that he has found multidisciplinary issues to be more problematic than cross-cultural and cross-national ones. Different disciplines have different cultures and modes of reasoning and levels of explanation or terminology.
In negotiations with collaborators across cultural boundaries, whether those of nations or disciplines, the ideal is to come together to create a research project that everyone can agree upon, Gelfand noted. She acknowledged, however, that sometimes you must compromise. She errs on the side of making sure that her collaborators are comfortable with what is happening.
AUTHORSHIP QUESTIONS
Reaching agreement in advance on authorship and dissemination of research results is particularly crucial but can be tricky, participants observed. In many collaborations, it is expected that coauthoring within national teams will follow local traditions. But, Super noted, the details can still be difficult to deal with. Many practices differ among countries, such as those for acknowledging graduate students. American researchers generally grant them authorship if they made a real contribution and share some responsibility for the scientific integrity of the publication, but some other countries do not. Box 3-1, Some Authorship Guidelines, contains two methods of dealing with this issue as presented by Super and by Lansford. In both approaches, the rules for sharing and use of data and for authorship are outlined and agreed upon by collaborators early in the process.
Journals’ authorship policies can also complicate the situation, as mentioned earlier. In addition, Merry Bullock noted, issues involving dissemination of research can arise because the broader incentives in many countries may be different from those in the United States. In many countries,
Box 3-1
Examples of Authorship Guidelines
Below are the authorship policies used by two international collaborative projects, “Families for the 21st Century” and “Parenting Across Cultures,” as presented at the workshop.
Families for the 21st Century:
An International Study of Parents, Children, and Schools
Guidelines on publications
Sara Harkness, Charles M. Super, Giovana Axia, Harry McGurk, Jesús Palacios, Barbara Welles-Nystrom, Andrzej Eliasz
If issues of authorship arise during the course of the project, colleagues should feel free to address them in an appropriate manner. It is expected that our spirit of good faith will carry us through any such problems that might arise, and that we will be able to resolve them courteously. The following eight principles represent a group consensus and will be used until such time as they may be altered by the group:
Principle 1. Every lead investigator (that is, the senior investigator in each country) has the right to publish (with local team members, as appropriate) his or her national data in any way he or she wishes. In most cases the two project directors (Harkness and Super), and possibly other lead investigators, should be considered for co-authorship, but that may not always be appropriate. In any event, it is appropriate for Harkness and Super, and the lead investigators, to be kept informed about planned single-country publications, and to receive copies in advance.
Principle 2. Publication of data pooled from all the samples will be coordinated (though not necessarily carried out) by Harkness and Super, in order to prevent simultaneous and overlapping efforts.
Principle 3: Co-authorship of any publication using data from a specific country will normally include at least the lead investigator for that country.
Principle 4: No investigator will be included as co-author without consent.
Principle 5: Order of authorship for any report from this project will be discussed by all concerned prior to completion. In general, the person who takes the lead in organizing and writing a report will be the first author. It is assumed that Harkness and Super will play this role on the initial major publication of multi-site data. When co-authors’ contributions are essentially equal, the sequence of authorship will be systematically varied.
Principle 6: Every publication, national or international, will include a standard project acknowledgment. A partial draft follows, but this may be amended from time to time: “The International Study of Parents, Children, and Schools is a collaborative project funded in part by the Spencer Foundation (Chicago, USA), the Australian Institute for Family Studies, and ... [SEND ADDITIONS to Harkness and Super]. All statements made and opinions expressed are the sole responsibility of the authors.”
Principle 7: It is expected that co-authorship within national teams will follow local traditions and guidelines; however, national teams will attempt to achieve a common standard of equity for multi-team publications.
Principle 8: Persons who leave the project will still be considered for co-authorship under these guidelines, to the extent that this is practical.
Parenting Across Cultures Authorship Policy
Jennifer Lansford
(As presented at the workshop)
For papers that use data from multiple countries, our group has an authorship arrangement that includes the naming of the lead investigator in each participating country. For any given paper, the primary authors do the actual writing and analyzing, and the remaining authors provide critical feedback. The primary authors are listed in order of their contribution to the specific paper; the remaining authors are listed in a revolving alphabetical order.
The first step in our writing process is to send a one-page data use proposal to the entire group outlining the research questions, data to be used, analysis plan, and intended product. The main purpose of the data use proposal is to be sure a researcher in one site is not duplicating the efforts of a different researcher working at a different site.
Although we have geared our collaborative papers primarily toward English-language international journals (as these are given more scientific weight for all of our collaborators), collaborators have sometimes published papers in local or regional journals and in languages other than English. Graduate students at the international sites have tended to take the lead role on those papers, and these outlets have provided an opportunity for students to gain publication experience without the pressure of writing in a nonnative language. Of course, one of the benefits to the international collaborators in the group is that they can take advantage of native English speakers as coauthors when they do publish in English-language journals.
researchers may need to convey their results back to the communities with which they have been working because the process is intended to help address societal problems. This requirement, Bullock continued, is often at odds with academic pressure for high-quality publications and also with some journals’ requirement that studies not be published anywhere before appearing in a journal. Barrett responded that one strategy to avoid this conflict can be to present research to affected communities in reports written in lay language and disseminated through local organizations. Another is to convey research through meetings with local communities.
FUNDING FOR INTERNATIONAL COLLABORATIONS
Two representatives of U.S. funding agencies addressed the workshop in hopes of “demystifying” donors, in the words of Janet Shriberg, senior evaluation advisor at the United States Agency for International Development (USAID). “U.S. reliance on international collaboration for high-quality science is growing,” added Anne Emig of the National Science Foundation (NSF). Emig noted that Science and Engineering Indicators, published by the National Science Board, shows that a growing percentage of highly cited articles have international coauthors.
Nonetheless, Emig pointed out that the fact that collaboration requires scientists and engineers to work together, often each seeking funding from their own national funding agency, creates a coordination problem. To address this issue, in September 2013, the Social, Behavioral, and Economic Sciences Directorate (SBE) of the NSF and the Research Councils of the United Kingdom (RCUK) announced the SBE-RCUK Lead Agency Agreement to facilitate collaborations by the two nations’ scientists. The agreement permits collaborators to submit a single proposal to both agencies, which then undergoes a single review process and receives a single decision. The agency providing the larger share of funding serves as the lead.
Some newer NSF approaches that Emig noted also provide opportunities for international work and experience. These include a solicitation entitled Catalyzing New International Collaborations, which will cover start-up costs of establishing collaboration, and Partnerships for Enhanced Engagement in Research (PEER), a USAID program to fund developing-country researchers collaborating with NSF-funded U.S. scientists.
For NSF, Emig noted, major drivers of support for international collaborations include “gaining access to unique facilities, equipment, research environments [and] sharing the costs and risks of scientific investments,” as well as “exchanging expertise on techniques and other insights” and building capacity. True intellectual partnership is a core value for NSF collaborations. While some scientists think it acceptable to go abroad, pick up specimens, and come back, that is not true intellectual partnership. Emig stated that NSF
believes “that international engagement should have mutual benefit from synergistic expertise and resources.”
Offering suggestions on how to work successfully with her agency, and, by extension, all funding agencies, Shriberg noted the importance of reading the evaluation criteria for the proposal. There is good research that funders do not look at because proposals do not meet the specific evaluation requirements of the particular program. Beyond that, context and partnerships are key. Funders look at who the local collaborators are when evaluating proposals containing international work. Also crucial is demonstrating rigor in quantitative and qualitative methods.
This page intentionally left blank.
