Assessing the Scholarly Communication Attitudes and Practices of Faculty: Lessons from a "Failed" Survey

INTRODUCTION This paper reports on a survey administered to faculty at Chapman University to assess their knowledge, attitudes, and practices with regard to scholarly communications, in order to help the new scholarly communications librarian plan appropriate library programs and services to meet faculty needs. DESCRIPTION OF PROGRAM The survey was adapted from the Institute on Scholarly Communications’ “Faculty Involvement in Scholarly Communications Opportunity Assessment Instrument” for a faculty audience in early fall 2013. It “failed” in that it faced long administrative delays and was met with a low response rate when finally published in December 2013. However, the responses received were enough to deduce general trends and gaps in faculty knowledge about scholarly communications, including a misunderstanding of the meaning of open access, misconceptions about its quality, concern with how publicly accessible research and data could be used by others, and a desire for information on how to manage, preserve, and share data. NEXT STEPS Both the survey results and the obstacles encountered in the survey’s administration provided important lessons in how to structure, market, and assess the impact of future scholarly communications discussions, such as those surrounding the university’s upcoming institutional repository. While the survey itself might have “failed,” these lessons can be applied to future endeavors in order to contribute to the long-term success of the faculty and the university as a whole. © 2014 Laughtin-Dunker. This open access article is distributed under a Creative Commons Attribution 4.0 Unported License, which allows unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. PRACTICE jlsc-pub.org | Journal of Librarianship and Scholarly Communication Received: 05/02/2014 Accepted: 06/24/2014


INTRODUCTION
Chapman University is a medium-sized, historically liberal arts institution in Orange County, California. Though it is over 150 years old, it has expanded greatly over the last decade, broadening its focus to incorporate the sciences (particularly the health sciences) and research across all disciplines. At the end of the 2012-2013 school year, the university hired a new scholarly communications librarian to help faculty navigate the changing world of scholarship. Because anecdotal data was spotty and no formal study of the faculty's scholarly communication knowledge, attitudes, or practices had been previously performed by either the library or another unit on campus, it was soon evident that the scholarly communications librarian would need to perform an assessment in order to plan appropriate programs and services to meet the faculty's needs. When the faculty returned to campus in fall 2013, a survey to collect this information was designed and administered.
The survey "failed" in several senses. It faced a series of administrative delays and solicited a low response rate from faculty, leaving a large number of faculty unrepresented in the results and any calculations from the data vulnerable to large margins of error. However, by using the data collected to deduce general trends and identify faculty knowledge gaps and needs, as well as learning how to conduct similar processes more efficiently in the future, this "failed" survey was transformed into a potential stepping stone for the university's long-term success.

LITERATURE REVIEW
Librarians have long been challenged to effectively promote scholarly communication concepts to their faculty, particularly when there are concerns about quality or going against established publishing models. In 2008, open access advocate Peter Suber noted that faculty were reluctant to embrace open access due to existing university reward systems and misperceptions about quality. He argued that faculty are motivated to publish in high-prestige journals, or journals that bear a reputation of quality, in order to get tenure and promotions. Open access journals are thus at a double disadvantage: they have not yet had the time to earn prestige, and they are often overlooked when faculty flock to established, usually traditionally published journals to disseminate their research. Thus faculty are often under the misperception that they must choose between prestige and open access, despite the facts that prestigious open access journals are growing, open access articles are often cited more (especially after the first year), and many traditional journals allow for open access through self-archiving in repositories.
Of course, before one can promote any scholarly communication concepts, one must understand the attitudes and practices of the faculty at one's particular institution. Using surveys to make these assessments is a common practice, and their findings have been almost universally in line with Suber's declarations. Library literature shows that even at major research universities, faculty often have limited knowledge or misconceptions about open access and institutional repositories and share concerns about how their work can be used, complying with copyright and publisher agreements, and time investment. Abrizah (2009) surveyed faculty at the University of Malaya about their use of open access repositories and found they wanted to make many types of work accessible, including theses and dissertations, post-prints, and conference papers. Though altruistic in motivation, they were concerned with copyrights, plagiarism, and publishers' policies. Several years later, Singeh, Abrizah, and Karim (2013) conducted a survey of faculty from five Malaysian research universities (including the University of Malaya) about their awareness of self-archiving at their institutions, perceptions of the practice, and hesitations. Nearly half had little knowledge of or experience with self-archiving, but were open to the idea when given the opportunity or under a mandate. The main causes of hesitation were fear of plagiarism, concern with how their work might be used, inability to publish their work elsewhere, and time required. Vandegrift and Colvin (2012) conducted an environmental scan as one method of assessing Florida State University faculty's open access knowledge. Their survey revealed limited awareness about open access and its mechanics, and illuminated the need for education about self-archiving and institutional repositories. At the University of Wisconsin-Eau Claire, a smaller liberal arts institution, Kocken and Wical (2013) found similar results, as well as concerns about the quality of open access publications. Lercher (2008) details a survey at Louisiana State University about how the institutional repository should be organized to meet faculty needs. Faculty who felt they had "valuable unpublished work" were more likely to submit to a repository, especially if it were search-able via Google Scholar. This indicates the need for repositories to accommodate work that cannot be published elsewhere, in addition to traditional journal articles. Lercher also notes that some departments had a low response rate, making figures calculated for them less reliable and weakening correlations between responses to different questions. Mischo and Schlemback (2011) surveyed engineering faculty at the University of Illinois at Urbana-Champaign about their open access attitudes and practices. They found high rates of confusion about the nature of open access (especially the workings of gold open access and peer-review for green open access), as well as reluctance to deposit articles and data in repositories due to unfamiliarity and confusion over copyright and journal permissions. However, rates of deposit into disciplinary repositories like arXiv were significantly higher, indicating a willingness to make work publicly available. Harley, Acord, Earl-Novell, Lawrence, and King (2010) report on the Center for Studies in Higher Education at the University of California Berkeley's study of scholarly communication needs and practices in seven disciplines. They determined that faculty are universally concerned with peer-review, prestige, time for publication, theft, and misinterpretation. Opinions on sharing pre-prints and data varied widely by field, though many faculty yearned for new publication models that quickened the timing of peer review; could accommodate various lengths, media, and embedded links; and supported new research methods, such as natural-language processing and visualization.
The University of California's Office of Scholarly Communication and the California Digital Library (2007) administered possibly the largest survey of this kind. They found that although there was high interest in scholarly communication (as evidenced by the large response rate, number of lengthy optional comments, and low abandonment rate), faculty tended to conform to conventional publishing behavior. This was due to concerns with low quality output in open access venues, the belief that everything would eventually appear online in an accessible format, and concern with the effect of non-traditional publishing on tenure and promotion. While many expressed a need for change in the current systems of scholarly communication, they did not want to instigate it themselves and tended to dissociate from the problems. The report noted that senior faculty may be the best targets for innovation, as they are free from tenure concerns and do not feel the same pressures as junior faculty. Housewright, Schonfeld, and Wulfson (2013) report on another large-scale study by Ithaka, in which they surveyed a random sample of faculty at American four-year institutions. They found that there was a trend toward growing acceptance of print-toelectronic transitions for scholarly journals, though most still valued established dissemination models. Though half of the respondents found preprint repositories such as arXiv and the Social Science Research Network very important to their research, less than a third made their work available through personal webpages, blogs, or repositories, and even fewer deposited their data. When choosing publishing venues, faculty were more concerned with factors that affected them, such as the cost to publish and the amount of circulation and prestige. This study did not break down faculty by academic rank, so it is unknown whether faculty share similar views pre-or post-tenure.
Librarians have also used surveys to assess the scholarly communication knowledge of each other. Bresnahan and Johnson (2013) conducted a survey of liaison librarians at the University of Colorado at Boulder, in order to prioritize areas of scholarly communication for training so they could better serve faculty needs. Creaser (2008) reports the findings of two complementary surveys, one to academic librarians and one to researchers, regarding open access in the United Kingdom. These surveys showed that many researchers were unaware of open access policies and the existence of institutional repositories and had a wariness of open access publications from both an author's and a user's perspective, largely due to lack of knowledge. While much has been written about the use of surveys, "failed" surveys are a rarer topic in library literature. When it comes to scholarly communication surveys, articles may refer to low response rates from one of several populations, but literature about completely unsuccessful surveys is lacking. One must turn to more generalized library literature to find information on dealing withor preventing-failed surveys. Webber, Lynch, and Oluku (2013) spend considerable time analyzing the low response rates of postgraduate students to the UK Postgraduate Taught Experience Survey, which reduce the reliability of the data collected each year. They found that students who were unclear of the survey's purpose were less likely to respond, as were students who saw no incentive to completing it. They offer a few ideas for improving future response rates, including financial rewards and better advertisement of the survey's purpose. However, as the paper was focused mainly on the survey process, they do not discuss how the limited data they did collect could be used to improve student experience, which was the survey's point. Melssen (2012) discusses two surveys administered by the University of Pittsburgh Health Sciences Library System regarding e-books. These surveys suffered from a low response rate, inconsistent or unclear jargon that confused respondents, too varied a pool of respondents, and non-response bias. As a result, the data was ungeneralizable. Melssen notes the need to test questions on a representative sample to confirm clarity, and provides a few recommendations for the library to facilitate e-book usage. Jowitt (2008) discussed a study conducted at the Universal College of Learning in New Zealand to collect quantitative and qualitative data about use and perceptions of library podcasts. This data was also ungeneralizable due to the survey's low response rate. The author identifies the fact that respondents were self-selected, and therefore a certain bias was likely (i.e. only respondents interested in podcasts were likely to respond), as the primary reason for non-participation. Rather than give suggestions for improving response rates, the article actually calls for other institutions to perform similar studies so the results can be compared. Adeleke and Habila (2012) surveyed the librarians of the Nigerian Librarian Association forum on their awareness and ownership of weblogs, and just over six percent responded. They conclude that a primary reason for the low response rate was unfamiliarity with blogging or low technological skills. Thus, there was a large non-response bias where those unfamiliar declined to respond. The authors recommend more education about the benefits of blogging, more training for librarians in general, and improvement of library school curricula regarding technology as possible solutions.
These articles about failed surveys vary in their analysis of the reasons for each survey's low response rate, and generally refrain from giving ideas on how to use the limited data collected to meet the surveyors' goals. However, a few do give advice for how to improve participation rates. Nonetheless, the advice is often very specific to the particular surveys discussed in each article, rather than presented as generalized best practices. Thus, in many cases, it may be difficult to apply these suggestions to other surveys or assessment endeavors, including scholarly communication surveys.

METHODOLOGY
The survey at Chapman University was created in September 2013 and based on the "Faculty Activism in Scholarly Communications Opportunity Assessment Instrument" that Lee Van Orsdel (2007) created for the Institute on Scholarly Communication. This instrument later became part of the ACRL Scholarly Communication Toolkit, a helpful collection of resources for a new scholarly communications librarian looking to begin campus conversations and start relevant library programs. The breadth of topics covered were considered foundational to any future activities, especially since open access, data management, and deposit in repositories are all becoming important considerations in the securing of research funding. While there are many other quality survey instruments that could have been used, this one was deemed a good fit because it did not presume the existence of a formal scholarly communication program and wasn't tailored to any specific university or library system.
The instrument itself was meant to be used by liaison librarians in conjunction with their departments, with the idea that librarians would work with their faculty to identify areas of promise they could investigate together. However, because the instrument was designed as a handout for librarians to use while engaged in conversation with faculty, and the scholarly communications librarian had to serve all of the departments on campus rather than just a few, using the instrument as originally intended would have been time-consuming and delayed the development and implementation of the library programs it was hoped would be the final result of the process. Thus, the instrument was converted from a handout for the librarian to fill out into a survey addressed to faculty. While in many cases this just involved simple rephrasing of questions, several questions were omitted because the answers were already known. For example, one question asked whether the library had recently made scholarly communications presentations to the faculty; it was already known from conversation with the other librarians that there had been no formal workshops or presentations at Chapman University, and only a few librarians had discussed any scholarly communication issues with individual faculty. Names of liaison librarians, grant officers, and others across campus were also known and did not need to be included in the survey. Other questions were removed because they were thought to be beyond the scope of the faculty's knowledge or would have required them to research external policies of publishers and societies. These were not considered relevant to the author's goal of assessing faculty attitudes and practices. Other questions were added or split in order to allow more specificity in the answers. These include questions about what faculty were reading and why they were not performing certain actions, under what conditions they would consider doing so, and whether and how they would like to learn more. A few questions were also converted into multiple-choice format so the data could be more easily quantified. Recruitment text and a waiver form were also created to go with the application to the Chapman University Institutional Review Board.
The administration of the survey faced several delays. The original application was sent back for revisions due to concerns over faculty's identifiable data, and provisions were made to better protect names, emails, and personally identifiable responses to open-ended questions. Once the application was approved, the survey was created using KwikSurveys, an online software service that allows respondents to answer the survey anonymously (needing only to identify their department) and to skip questions they would prefer not to answer. When trying to advertise the survey, the initial mailings were rejected due to a human resources error, and an alternate procedure had to be developed in order to advertise the survey on the campus' weekly faculty email. This first advertisement did not go out until two weeks before the end of the fall semester, a hectic time in which faculty were finishing finals and grading before the winter recess.
In an attempt to gather more responses, the survey was reopened for two weeks after the faculty's return to campus in late January. In mid-February, the data was collected and analyzed. Follow-up emails were sent to faculty who had identified themselves, thanking them for their participation and inviting them to further discussion.

RESULTS
As noted previously, the response rate to the survey at Chapman University was very low. Only twenty of Chapman's approximately 392 full-time faculty (or 5.1%) started the survey, and nearly a third of those abandoned it after just a few questions. To operate at a 95% confidence level with a 10% margin of error, the survey would have needed 77 respondents, or 19.6% of full-time faculty. Because the response pool was so small, there is at least a 20-30% margin of error for each set of answers, and all responses are best treated as anecdotal. 1 Any figures noted should not be viewed as definitive, generalizable, or universal. The small response pool also prevents the possibility of performing more detailed analyses, such as examining trends within disciplines or performing crossdisciplinary comparisons. This is unfortunate, as the respondents came from a wide variety of departments, representing all of the major schools on campus except film ( Figure 1). 1 Twenty-three respondents out of a sample size of 392 would be needed to have a confidence level of 95% and a margin of error of 20%. Fifteen respondents would be needed for a margin of error of 25%. Every question after the first had fewer than fifteen responses, meaning the margins of error for these are closer to 25-30%.

Findings of the Survey
In the first portion of the survey, every respondent reported that they published in peer-reviewed journals. Nearly four-fifths reported that they also perform peer review for a journal.
When asked whether online journals were considered acceptable for tenure or promotion, just over half believed they were: one respondent believed that they were unequivocally acceptable, and the rest reported that they were acceptable, but not as much as traditional print journals. A few faculty indicated that online journals were not considered acceptable at all, and nearly a quarter were unsure. When asked about the acceptability of open access journals, the same group was unsure, but all of the remaining faculty believed they were acceptable to some degree. The respondents that found online journals less acceptable than print felt the same about open access journals, while the rest thought open access was acceptable flat-out. This points to an interesting disconnect: a small group of respondents deemed online journals unacceptable for tenure or promotion, but thought open access journals were fine ( Figure 2). Furthermore, no faculty reported that their department had an explicit policy against either online or open access publications, but that any biases against them were implicit. This is in contrast to a later question where nearly half of faculty reported that their departments had explicit standards relating to citations and impact factors.
Regardless of their feelings on the acceptability of online and open access publishing, the minority of faculty had published in either model. Just under half reported having ever published in an online journal, and in a separate question, just over a third reported they had published in an open access one. However, five-sixths would consider publishing in either type if they had more information.
Faculty were asked to detail any grants they had ever received, then asked whether any of the funding agencies allowed for the payment of author fees to publish in an open access journal. Half said that at least one of their funding agencies did allow this, while one said that none of theirs did. A third of respondents did not know.
The section on data presented a wide range of responses. Faculty were first asked to detail what kinds of data they or their departments were generating and storing. Then, if they were not storing or otherwise preserving that data, they were asked why not. Though this question received the least responses of any in the survey, several concerns were mentioned by multiple parties: where to store data, the time and cost involved, and commercial or other  JL SC reuse. One respondent was also concerned about how to manage, format, or package data. One saw no need to store or preserve data (possibly because data is not a huge part of their discipline), and one was concerned with their department's handling of assessment data.
Faculty were almost evenly split on whether a disciplinary repository existed for their field, with responses of "yes," "no," and "I don't know" being nearly equally represented.
Relatively few respondents (about a fifth) had deposited work in a disciplinary repository. However, nearly all respondents said they would consider it if they would not lose any of their rights. Dissemination, preservation, and communication with colleagues were other motivators, with half to three-quarters of respondents expressing interest in those issues. Nearly half would deposit in order to comply with funding terms (or perhaps, had funding terms with which they needed to comply). Just under a third reported they needed to know that such a repository existed, while another third wanted some say over how their work could be used. One respondent felt they were already well-served by arXiv and that ResearchGate worked well for others (Figure 3).
Faculty seemed to be fairly well in tune with the open access activities of the societies they belonged to or whose publications they followed. When asked whether any of those societies published open access journals, only two did not know. (Nearly half said yes, and about two-fifths said no.) When asked if any of the societies ever made their backfiles open access, the same percentage did not know. (Here, about a third said yes, and just over half said no.) Finally, faculty were asked about their relationship to the library and librarians. Both questions in this section yielded a diverse range of responses. When asked about venues for scholarly communication education that the library could provide, almost three-quarters expressed interest in a research guide. Just over half were interested in workshops or lectures, talks, or forums, while just under half wanted one-on-one consultation and just over  a third wanted email conversations with the scholarly communications librarian. Nearly half were interested in an occasional newsletter about new developments in scholarly communication. One respondent wrote that they wanted visits to departmental meetings, a useful option that hadn't been included in the original list of answers ( Figure 4).
Likewise, when asked about which aspects of scholarly communication they were most interested in learning about, the answers varied widely (which may be partially due to the open nature of the responses). Nearly half of respondents wrote about data management, preservation, and storage. The rest of the responses were split equally among areas faculty felt would help their careers: publication venues, author rights, new developments, open access requirements in the European Union, and how it could help them do a better job.

A Low Response Rate Doesn't Have to Mean Failure
A low response rate to a survey doesn't have to spell doom for the savvy librarian. One may not be able to determine concrete facts and figures, but a few thorough and thought-out responses may be just enough to gauge attitudes, general knowledge, or interests; to determine possibilities and opportunities; and to inform future plans of action. It is quite possible, even likely, that those few answers may reflect larger trends. After all, faculty do not work in isolation: they collaborate, they engage in conversation with their colleagues both inside and outside of the university, they all have specific requirements they must meet in order to gain tenure, and more frequently they are engaging in interdisciplinary research. If a few researchers are calling for a particular program, service, or educational opportunity, it is unlikely their needs exist in a void; there are probably other faculty who could benefit from that program or service as well.
A low response rate can indicate that there is a large degree of non-response bias at play: busy faculty may not have answered the survey due to their workloads, and faculty with limited knowledge of scholarly communications issues may have declined to participate. In the case of the former, the survey was administered during one of the busiest portions of the semester, and it may not be too surprising that these faculty did not prioritize the survey.
In the case of the latter, faculty may not have responded  A low response rate does not have to spell doom for the librarian looking to learn. The data generated, even if only from a few sources, can provide an idea of general trends and needs, and also illuminate gaps in knowledge to be filled or interests to be cultivated.

Lessons Learned from the Survey Itself
The data collected from the survey, though limited, is useful for deducing general trends of scholarly communication attitudes and practices among the faculty of Chapman University. For example, the responses reveal confusion over the meaning of open access. Some faculty are possibly confusing it with online publication, as evidenced by the discrepancy in responses when asked (separately) if one or the other were acceptable for tenure. The fact that open access journals are found more acceptable than online journals is surprising, because open access journals are, by nature, published online.
Open access publications also often suffer from perceptions of lower quality, as evidenced by open-ended responses to later questions. Several faculty claim they are "low quality," "weaker," "easy to publish in," and "probably not scholarly," or that they have "less prestige" or "less quality control" than traditional print journals. Education about open access can lead to education about other scholarly communication issues as well and help faculty advance their research and manage their publication prospects. Given the low number of faculty who know much about open access, it is not surprising that many faculty may be unsure of whether their grants allow for the payment of article processing charges. It is impossible to know from the limited survey data the reasons for this lack of knowledge; it could have been because the faculty had not thought about the issue, had not read their award letters that closely, or were not planning to publish in an open access journal and thus did not care. Nevertheless, this presents another opportunity for educational outreach, so that researchers can understand the possibilities afforded to them by their grants, allocate money for author fees if they decide to pursue a (gold) open access route, and plan their applications, research, and publication prospects in advance. This would allow them to consider publishers previously unknown to them and also better equip them to comply with increasinglyprevalent funder mandates.
The survey results also point to a need for education regarding data management, storage, preservation, and sharing. As mentioned previously, the question about how faculty are currently storing or preserving had the least responses, half of which indicated that faculty were not taking charge of this area of their research. The fact that the responses to this question were significantly lower than the others may illuminate additional non-response bias; faculty may not have answered because they are not storing or preserving their data. 2 Also of note is that nearly half of the responses to the final question, which asked what areas of scholarly communication faculty would like to learn more about, had to do with data. Overall, these responses reveal that some faculty simply do not know how to approach data issues and are unsure of how to manage data or where to deposit it, if they are worried about it at all. However, there will be greater need to pay heed to data issues as more mandates come into play, especially at the federal level--but any feeling of worry could be transformed into one of preparedness through outreach and education. The survey responses also showed concern over how shared data could be used by others. In this regard, it would be useful to provide education about Creative Commons licenses or other ways of indicating allowable reuse.
Perhaps the most important lesson learned from the survey results is two-fold: faculty want to learn about scholarly communication in a variety of ways, and they want to know how it can directly benefit their careers. This is important to remember when planning library programs or services directed at faculty. A workshop titled "How to Make Your Data Comply with Your Funders' Demands" puts the direct and immediate benefit to the faculty--compliance with the body paying for their research--front and center, as opposed to something like "Data Management Workshop." The same is true with a talk labeled "Get Your Work Noticed by a Worldwide Audience" over "Learn about Open Access." If librarians can rope faculty in with a promise of knowledge that will directly benefit their careers, they will also gain the opportunity to educate their audiences about the larger, philosophical issues in the background. 3 2 The number of faculty who completed any portion of the survey and came from data-heavy disciplines was higher than the number of respondents to this question, so it cannot be assumed that the response rate was so low merely because data was not an important consideration for this group of faculty.
3 It may be worth noting that the title of this survey did not stress a direct and immediate benefit to the faculty. The advertising materials did mention that completing the survey would help the library plan programs and services to help faculty with their research needs, but the title was fairly dry and jargon-y: "Faculty survey on scholarly communications attitudes and practices." (This was partially due to the limited amount of space allotted to the article title.) If a similar survey effort were undertaken in a few years, it would be interesting to see if a title that advertised a more direct and immediate benefit to the faculty would elicit better participation.

Lessons Learned from the Process of Conducting the Survey
Regardless of the survey's title, there are other lessons to take from its administration that may benefit future endeavors. A low response rate may indicate lack of investment from the respondents, but it could also indicate poor planning on the part of the administrator. In order to get the best possible response, it is important to have a good sense of appropriate timing, as well as realistic expectations for how long the process will take.
It is important to target a time that will not be overwhelming to the anticipated respondents. In this case, the survey was unable to be disseminated until two weeks before the winter break, at a time when the faculty were busy with finals and then grading.
Communications on non-teaching activities were low, and most of the activity on campus was focused on wrapping up work before the holidays. Any lack of interest on the part of the faculty was likely reinforced by the need to focus on finishing other duties on a schedule. When the survey was extended to cover the first two weeks of the next semester, the response rate went up by twenty percent. This was an improvement, but it was still a very busy time for faculty as they began new instruction. If the survey had been disseminated in early October, as originally planned, faculty may still have been preparing for midterms and conducting research, but they would not have had a ticking clock looming over their heads and may have been able to more easily cobble together a few minutes for the survey. It may also be useful for survey purveyors to consult their institutional research offices to determine if there is any sort of master survey calendar. If such a resource exists, it can help the survey planners determine an appropriate time for dissemination that won't be in conflict with other surveys and avoid divided attentions and lower response rates for both. This is why it is also useful to overestimate the time needed for preparation. If one builds extra time into the process, it helps to prevent disruptions in the anticipated workflow from delaying later steps. One should begin planning a survey or similar activity several months before one plans to administer it, especially if there are other units that need to provide input or assistance. If one ends up finishing preparations too early, it is a simple matter to just hold on to the survey materials until the time for proper dissemination is at hand. 4 One should also determine a marketing plan early on, including how many times potential participants will receive invitations to participate. While one does not want to overwhelm or annoy potential participants with constant advertisements, a few well-timed messages can be useful reminders in an environment where many things beg for attention. Each time an email reminder went out for this survey, there was a jump in responses for several days.
It can be speculated that the lack of an immediate and direct reward may have contributed to faculty apathy toward answering (and completing) the survey. Because the rewards offered were indirect and delayed (helping the library plan services that will help them, someday), a better response rate may have been achieved by offering some sort of immediate and tangible reward, such as gift cards. Of course, the ability of any researcher to offer such rewards is dependent on oft-constrained budgets. In an upcoming and separate research project, the author will be coadministering faculty focus groups in which participants are offered a small gift card. It will be interesting, and possibly enlightening, to compare the differences in participation rates between these two projects.
Finally, it is important to think critically about all the possible ways one might want to analyze data when planning a survey. In this case, questions were rephrased from the original instrument to address faculty, but little consideration was given to the faculty's diversity. If the survey could be redone, it would have been useful to ask what stage each respondent was at in his or her career. Even a simple checkbox for pre-or post-tenure status would have allowed for analysis of whether latercareer scholars were more or less amenable to open access publishing, or whether earlier-career faculty were hesitant about innovative scholarly communication venues.
Having this information would have opened a wealth of other opportunities for the author to analyze and perhaps could have helped in targeting services to specific groups.

NEXT STEPS AND CONCLUSION
In the summer of 2014, Chapman University will launch an institutional repository under the direction of the scholarly communications librarian. In advertising the repository, whether it be through presentations to the faculty senate, library workshops, or visits to individual researchers or departments, the scholarly communications librarian will have the opportunity to introduce and reinforce concepts of open access, data management, copyright, and author rights. The initial survey, though it "failed," has provided a wealth of information about how to tailor these discussions so they will resonate with faculty's needs, desires, and preexisting knowledge. For example, rather than espousing on the ideological benefits of open access or data management, there will be a greater emphasis on usage data, citation counts, content examples, and case studies from existing repositories. Thus, the faculty will be exposed to the direct benefits that depositing in the repository can have for their work. The language of the presentations will also be adjusted to reflect the terminology that faculty tended to use in their open responses, and efforts will be made to steer clear of, or at least succinctly explain, terms they showed hesitance toward. One presentation was held for a selected group of administrators and faculty in the spring, but after a soft open over the summer, several more are anticipated for the fall semester.
A mini-presentation was also given at a faculty workshop about National Science Foundation funding, where issues of data management and sharing were discussed. This workshop was co-sponsored by the campus' Office of Undergraduate Research and Office of Sponsored Research Administration, who have already become allies in supporting the repository due to the benefits it will provide Chapman's faculty in terms of funder compliance. They will be key allies in marketing the repository as a solution for data storage and dissemination. It is likely that there will be additional opportunities for collaboration in the future, possibly with a wider variety of offices around campus as the repository gains momentum. As these offices push faculty toward the repository (and thus contact with the scholarly communications librarian), it will provide additional opportunities for further individualized education.
Teaching faculty about the repository may open avenues for deeper discussion of scholarly communication issues as well. Interest in the repository may lead to the need Journal of Librarianship and Scholarly Communication | jlsc-pub.org JL SC for and possibly higher attendance of more general presentations and workshops about aspects of scholarly communication, such as the overall concept of open access; data management, preservation, sharing, and storage; copyright issues; author rights; and alternative publication venues. For example, if faculty see that their research is being heavily downloaded after deposit in the repository, they may gain an interest in open access publication for future work, either to procure instantaneous worldwide dissemination or to allow for quicker upload into the repository.
Due to the surprising amount of faculty who reported interest in a scholarly communications newsletter (especially given the amount of email they already receive), the repository will be investigated as a publication platform. However, it is still to be determined whether such a thing would need administrative approval, despite the scholarly communications librarian's role as the repository director. If allowable, this newsletter could contain information about the repository itself, as well as scholarly communication developments of interest to faculty, such as new mandates, article processing charge waivers, or new publication venues. Faculty could choose to follow either the series or the author in order to receive updates. Another option would be to create the newsletter within the scholarly communications research guide, but there it would be less likely to be noticed, harder to advertise, and harder to track statistics for.
As for the research guide itself, it already had much of the information faculty wanted, but clearly needs better advertising in order to serve all faculty effectively. This will require investigation into possible advertisement venues; the repository newsletter (if allowed) or the webpages of similarly-focused units (such as Sponsored Research or Undergraduate Education) may be options.
Finally, future methods of determining faculty's knowledge of scholarly communications issues and their satisfaction with upcoming services and programs will need to be investigated. As mentioned previously, part of the reason for this survey's low response rate may have been unfamiliarity with scholarly communication issues. While this sort of pre-assessment is useful for determining areas of focus for library outreach, it will be equally important to perform post-assessments of library programs in order to gauge their efficacy. A repetition of this survey after a year or two of outreach and programming could reveal changes in faculty's scholarly communication attitudes and practices in several ways. If participation in the survey was substantially higher the next time it is administered, it could indicate that more faculty are aware of scholarly communication issues and view them as important enough to merit a response. The data could also be compared to the little collected in this instance to determine how well the library has been doing: if fewer people express confusion or at a loss of what to do regarding data, for example, it could help demonstrate that outreach efforts have been effective. Other methods of assessment may be beneficial and provide more immediate information, too. Feedback forms after specific programs, individual conversations with faculty, and more abundant anecdotal data as the scholarly communications librarian forges deeper relationships with faculty and staff across campus may yield useful insights or inspire other methods of assessment.
Though this survey "failed" in terms of administration and faculty response, it still imparted valuable lessons about faculty attitudes and practices regarding scholarly communications, as well as the process of survey administration and assessment. These lessons will inform future services, programs, and practices at Chapman University, helping to ensure long-term success for the library, the faculty, and the university as a whole.