How Unique Are Our Users? 451 450 How Unique Are Our Users? Part 2: Comparing Responses Regarding the Information-Seeking Habits of Education Faculty Sarah Robbins and Karen Rupp-Serrano Sarah Robbins is Director of Public Relations & Strategic Initiatives and Karen Rupp-Serrano is Direc- tor of Collection Development & Scholarly Communication, both at University of Oklahoma Libraries; e-mail: srobbins@ou.edu, krs@ou.edu. © 2013 Sarah Robbins and Karen Rupp-Serrano, Attribution- NonCommercial (http://creativecommons.org/licenses/by-nc/3.0/) CC BY-NC This follow-up study examines whether or not findings of single institution studies are applicable to other institutions by performing an institution-to- institution comparison of the results obtained from an information-seeking behavior survey sent to education faculty at twenty research institutions. The results from this study corroborated what was found in the previous study conducted on the information-seeking behavior of engineering faculty in 2009. It indicates that general information about information- seeking behavior of faculty holds true across institutions, while information related to specific library services or facilities should be validated locally. s budgets tighten across high- er education, libraries are under greater scrutiny for their spending and are being held increasingly accountable for their contributions to the missions of their home institutions. As a result, librarians must astutely demonstrate their contri- butions to the teaching, research, and service of the university; one way to do this is through user surveys determining the value students and faculty place on library resources and services. The results of many of these studies are published in the professional literature; however, these published studies, often written by prac- titioners, usually focus on the situation at a single institution with little attention paid to the transferability of the findings to broader audiences. This means readers must determine for themselves how, if at all, the results apply to their situations. To promote a culture of evidence-based practice within librarianship, library practitioner-researchers need to develop studies and report results in a way that promotes educated consumption by their peers. It is costly to undertake user stud- ies, so it is important to know what types of information about library users hold true from institution to institution and what types should be locally validated before taking action. For the current study, identical user surveys related to the information-seek- ing behavior of education faculty were conducted at twenty research institutions. The researchers then examined the results by institution and compared the findings at other institutions included in the study crl-390 How Unique Are Our Users? 451 to ascertain to what extent librarians can apply the findings of single institution research to their own situations. While we often think our situation and users are unique, are these beliefs grounded in reality? Background Information In the previous article discussing the uniqueness of users, the literature review focuses on practitioner-researchers in libraries, the nature of research in library and information studies, and user behav- ior studies and user satisfaction surveys at large. While a recap of the literature review in these areas is not necessary for the understanding of the current study, several articles have been published since the writing of the previous article that warrant mentioning and indicate that the topic is still relevant. In a series on evidence-based librarian- ship 101, Virginia Wilson focuses an entire article on the importance of determining if and to what extent research findings are applicable to one’s own situation. She de- lineates how practitioners can determine if presented evidence is applicable to their situation by considering the user groups, timeliness, cost, workplace politics for application, and the severity of the situ- ation in need of a solution.1 She reminds practitioner-researchers that: You will more likely find evidence that resembles your situation, but that needs to be replicated and validated at the local level…. It is worthwhile keeping in mind that if you go the route of validating the evidence you have found by repli- cating it at your level, the greater LIS community will benefit if you write up your efforts and find a way to disseminate the information.2 This emphasizes not only the impor- tance of using research in practice but also the importance of sharing research results with a broad audience so others in the field benefit from the findings. Ray Lyons recounts the plethora of issues surrounding the use of improper research methods and statistical analysis within the field.3 He specifically discusses convenience sampling, selection bias, and the problem of nondata. This article heightens awareness of common pitfalls found in the LIS literature and educates readers about the dangers of these issues when present in studies. An understand- ing of these limitations is important for practitioner-researchers who often use their home institutions, a sample of con- venience, as the basis for their studies. Echoing sentiments shared by Ly- ons, Greifendeder and Seadle discuss problems with the data gathering for a study about mystery shoppers as a way to evaluate reference services in an editorial for a recent issue of Library Hi Tech.4 The authors indicate that, while the researcher’s findings may indeed be helpful to the institution, the limitations of the study should be communicated. They conclude: The people who do the real disser- vice to the institutions they study are those who misrepresent the validity of their data and attempt to draw unsupportable conclusions based on those data. That can be avoided without trouble or expense. Avoiding useless results requires openness and transparency about how the data were gathered and a reasonable judgment that does not exaggerate what these data can pos- sibly mean.5 This reiterates the importance of researchers clearly and accurately com- municating their research methodologies and being honest about the limitations of the study. A clear understanding of a study’s limitations as well as the method- ology assists practitioners in determining if the findings might be relevant to their particular situation. Lili Luo examines the role that the LIS research methods course plays in pre- 452 College & Research Libraries September 2013 How Unique Are Our Users? 453 paring practitioners to conduct research once they are in the workplace.6 Luo finds that the “majority of LIS practitioners did involve research at work” though it was primarily using existing research to solve problems or improve services rather than conducting original research. She also found that practitioners considered the research methods course helpful for evaluating research studies and applying them, improving services to patrons, and designing and implementing studies for gathering data to facilitate decision mak- ing or solve problems. These findings are encouraging for those who bemoan the state of LIS research and suggest that requiring LIS students to complete a re- search methods course as a part of their master’s degree program might help to improve the quality of research studies published by practitioner-researchers. Methodology To test whether the findings of user stud- ies conducted by researchers at a single library can be applied to other, similar libraries, the researchers conducted a user study of education faculty members at multiple institutions and compared institution-to-institution results. This methodology replicates a study con- ducted on engineering faculty in the fall of 2009 and is meant to test whether the findings of the initial study hold true in a different discipline. A separate paper presents the aggregated results of the information-seeking behaviors of educa- tion faculty based on the survey data. The researchers surveyed education faculty members at twenty research in- stitutions from across the United States. The 15-item survey consisted of demo- graphic, open- and close-ended questions (see Appendix). The survey gathered both qualitative and quantitative data and was designed to take less than ten minutes to complete; all questions were optional. Responses were anonymous and confidential. The survey was derived from the surveys used by Robbins et al., Jankowska, and Cannon.7 In October 2010, an e-mail invitation to participate in an online survey was sent to approximately 2,878 education faculty members at twenty public research insti- tutions. The institutions were selected as a purposive sample and represented differ- ent regions of the United States with top- rated education programs as determined by U.S. News & World Report8 and with libraries that are members of the Asso- ciation of Research Libraries. The sample was chosen to ensure that the faculty were from institutions of similar size and reputation, which could influence faculty information-seeking behaviors. Student assistants gathered e-mail addresses of all faculty listed on the institutions’ websites for their education department or college. This typically included both tenured and nontenured faculty as well as researchers and faculty emeritus; the survey was sent to the entire population as denoted on the institutional websites. Faculty members were given three weeks to respond; a reminder e-mail was sent after two weeks. Results & Discussion Of the 2,878 e-mail invitations sent, 538 education faculty members responded, for an overall response rate of 18.69 percent. By institution, the response rate ranged from 9.88 percent to 30.85 percent. If fewer than twenty responses were received from an institution, that institution’s data were removed from the data set. This left the data from twelve institutions to be included in the analysis. A chi-square test of association was conducted on the data gathered by eleven of the survey questions to determine the statistical significance, if any, of the rela- tionship between the respondents’ institu- tion and the answers given in response to the survey. The results of this analysis are presented in tables 1–9. Few of the p values indicate a statis- tically significant (p≤0.05) association between a given response and the respon- dent’s institution. This suggests that, for most types of information gathered, the results found at one institution would How Unique Are Our Users? 453 mirror the results found at another in- stitution when education faculty were asked the same questions. The p values presented in the tables do not indicate the level of importance associated with any of the given services, merely that the answers given were or were not statisti- cally significant in their association to a particular institution. Departmental Duties In the survey, faculty members were asked to provide answers to several demographic questions such as area of specialty within education, their length of time in the field, their institutional rank, and what was included in their departmental duties. Much of this was for the aggregated study on the information-seeking behavior of education fac- ulty and was not analyzed for the purposes of the current study because it was assumed to be (1) institution-specific and (2) readily accessible to librarians at an institution without having to conduct a formal survey. However, the research- ers did perform a chi- square analysis of asso- ciation on the responses received to the question, “Which of the following are included in your de- partmental duties?” (see table 1). Overall, the re- sults did not show a sys- tematic difference among the twelve institutions, with the exception of su- pervision of doctoral re- search (p=0.028), graduate instruction (p=0.050), and field and/or laboratory research (p=0.050). While the association between institution and responses to this question were not statistically significant for the most part, librarians at the institution are best po- sitioned to know the job requirements of their institutions’ faculty and should interpret survey results with this knowl- edge in mind. Information Use The chi-square analysis of the responses received to the question, “How many of the following have you completed within the past 5 years?” indicated that there was no statistically significant association be- tween the responses and the respondents’ institution (see table 2). The researchers asked faculty how frequently they sought information to complete a series of tasks common to Table 1 Departmental Duties Correlated to Respondents’ Institution Pearson χ2 Value df p Supervision of Doctoral Research 21.527 11 0.028 Graduate Instruction 23.618 11 0.050 Field and/or Laboratory Research 19.688 11 0.050 Undergraduate Instruction 18.009 11 0.081 Grant Preparation 14.707 11 0.196 Commercial/Proprietary Research 14.655 11 0.199 Administrative Duties 12.999 11 0.293 Table 2 Frequency of Completed Research Projects Correlated to Respondents’ Institution Pearson χ2 Value df p Nonrefereed Journal Articles/ Book Chapters 59.568 44 0.059 Conference Proceedings 55.233 44 0.119 Books 48.486 44 0.297 Grant Applications 48.342 44 0.302 Patents/Commercial Projects 44.986 44 0.430 Refereed Journal Articles/ Book Chapters 44.276 44 0.460 454 College & Research Libraries September 2013 How Unique Are Our Users? 455 education faculty members. Five of the six tasks did not show a systematic dif- ference among the institutions. The single statistically significant associated task was preparing a new research proposal/ grant application (p=0.008) (see table 3). To determine the possible implications for practitioners, the researchers com- bined responses to look at the range of faculty responses by institution indicating that they sought information to complete a given task at least monthly. When look- ing at the data in this way, the ranges that seem to indicate a meaningful difference are writing/researching for publication and preparing for a conference presenta- tion. At Arizona State University, only 48.6 percent of the faculty indicated they sought information to write/research for publication at least monthly, whereas 89.7 percent of the faculty at the University of Missouri indicated seeking information for this purpose at least monthly. By con- trast, only 27.5 percent of the respondents at the University of Missouri indicated they sought information to prepare for a conference presentation at least monthly, while 56.9 percent of the responding faculty at Texas A&M indicated seek- ing information for this purpose at the same frequency. Since each of these tasks requires similar types of information, it is unlikely that these differences would have a significant impact on collection development activities or the disburse- ment of funds. Table 3 Frequency of Information Seeking to Complete Tasks Correlated to Respondents’ Institution Pearson χ2 Value df p At Least Monthly (%) Prepare a New Research Pproposal/Grant Application 83.677 55 0.008 16.1–39.2 Professional Development/Remain Current in the Field 69.832 55 0.086 74.4–87.5 Prepare for Student Lectures 61.76 55 0.247 66.7–96.1 Write/Research for Publication 61.275 55 0.261 48.6–89.7 Determine Protocols for Field Research 57.663 55 0.377 20.8–36.4 Prepare for a Conference Presentation 50.745 55 0.638 27.5–56.9 Table 4 Ranked Importance of Sources for aiding Research Correlated to Respondents’ Institution Pearson χ2 Value df p Very Important + Important (%) E-mail Discussion with Students 57.466 33 0.005 43.3–86.4 Face-to-face Discussion with Students 49.197 33 0.035 34.8–71.9 Attendance at Conferences 45.445 33 0.073 65.2–96.4 Face-to-face Discussion with a Colleague 36.875 33 0.294 82.8–95.7 E-mail Discussion with a Colleague 36.247 33 0.320 69.5–91.3 Internet Resources 33.745 33 0.431 80.6–100 Books 33.626 33 0.437 85.7–100 Scholarly Journals (in print/online) 33.109 33 0.462 91.4–100 Textbooks 29.235 33 0.655 35.7–66.0 How Unique Are Our Users? 455 Finding Information While it is typical for practitioner-re- searchers to inquire about the productiv- ity of practitioners and the tasks that led them to seek information, knowing where they go to find information once they have an expressed need has more immediate implications for librarians. To determine this, the education faculty members were asked to rank the importance of sources for helping them with their research, for staying current in their field, and for identifying older resources that might be relevant to their needs. Of the nine resources listed as poten- tially helping faculty with their research, two were found to be statistically sig- nificantly associated to the respondent’s institution. Those that were significantly associated include e-mail discussion with students (p=0.005) and face-to-face discus- sion with students (p=0.035) (see table 4). The chi-square analysis of the respons- es to “How do you keep abreast of current developments in your field?” provided a similar finding. With the exception of personal communication (p=0.037), the overall results did not show a systematic difference among the twelve institutions (see table 5). Only one of the responses to how faculty members became aware of less recent journal articles was sta- tistically significantly associated to the respondents’ institutions: using citations at the end of book chapters (p=0.033) (see table 6). Table 5 Sources for Remaining Current in the Field Correlated with Respondents’ Institution Pearson χ2 Value df p Range of Responses (%) Personal Communication 20.698 11 0.037 50.9–73.9 Electronic Discussion List 17.3 11 0.099 15.6–45.8 Follow References or Leads from an Article/Item of Interest 12.974 11 0.295 72.5–93.8 Current Awareness Service 12.686 11 0.314 0–8.6 Attendance at Conferences 12.076 11 0.358 75–96.6 Scanning Current Issues of Journals 10.366 11 0.498 85.7–100 Scanning Recent issues of Abstracting/Indexing Tools 7.698 11 0.740 20–43.5 RSS Feeds 6.038 11 0.871 4.2–14.5 Table 6 Tools Used to Discover less Recent Journal articles Correlated with Respondents’ Institution Pearson χ2 Value df p Range of Responses (%) Citations at the End of Book Chapters 21.045 11 0.033 40.6–81.3 Retrospective Searching of Indexing/ Abstracting Tools 12.408 11 0.334 35.5–62.1 Browsing through Older Volumes 12.201 11 0.349 11.4–35.5 Personal Communication 10.394 11 0.495 31.3–58.3 Citations at the End of Journal Articles 9.191 11 0.604 79.2–96.9 456 College & Research Libraries September 2013 How Unique Are Our Users? 457 The researchers were also interested in the factors influencing a faculty mem- ber’s choice to use a resource or service and asked three questions related to this. Faculty were asked which factor most influenced their use of a current information source and were given six choices as well as the option to provide their own answer. These choices included: least time to track down the information, convenience, currency, authoritativeness (reliable, complete information), famil- iarity, and reliably available/no wait or hassle. The faculty members’ responses to this question were not found to be statisti- cally significantly tied to the respondents’ institution (df=66, p=0.436). Respondents were asked to indicate which factors might limit their use of the library’s electronic services and resources and were allowed to select all that ap- plied. Of the seven choices, only two were found to be statistically significantly associated to the respondents’ institu- tions—hard to find on library website (p=0.022) and unavailability of needed electronic resources/services (p=0.048) (see table 7). It stands to reason that these factors would be influenced by a faculty member ’s home institution, given that subscriptions and library websites are specific to the institution whereas access restrictions and lack of time are likely more ubiquitous. Faculty were asked how the library’s electronic resources and services improve finding needed information and were al- lowed to select all that applied (see table 8). Only one choice—provides access to full text publications (p=0.038)—was found to be statistically significantly as- sociated to the respondents’ institution. While the p value indicates the response variation is statistically significantly tied Table 7 Factors limiting Use of library’s electronic Services/Resources Correlated with Institution Pearson χ2 Value df p Range of Responses (%) Hard to Find on Library Website 22.322 11 0.022 25.8–46.4 Unavailability of Needed Electronic Resources/Services 19.841 11 0.048 40.0–74.2 Access Restrictions 8.8888 11 0.632 14.6–24.5 Lack of Time 7.312 11 0.773 22.4–52.4 Unaware of Available Eelectronic Resources/ Services 5.821 11 0.885 6.5–40.0 Lack of Instruction 4.414 11 0.956 0–10.2 Table 8 Benefits of Electronic Resources/Services Correlated with Institution Pearson χ2 Value df p Range of Responses (%) Provides Access to Full-text Publications 20.605 11 0.038 54.2–92.2 Helps me Assist Students in their Research Efforts More Effectively 14.288 11 0.217 41.7–74.2 Speeds up Research Process 6.756 11 0.818 79.6–100 Allows Working from my Office/Home 6.148 11 0.863 87.5–96.9 Does Not Improve Finding Information 5.247 11 0.919 0–3.9 How Unique Are Our Users? 457 to the respondents’ institution, the range of responses indicates that a majority of the respondents at each institution felt that providing access to the full text of a publication was a benefit of electronic resources. For a practitioner, this would indicate that full text of publications is beneficial to faculty, not just speedier ac- cess to resources or out-of-library access to the tools. Using the Library As practitioners, it is important to know the relative value faculty place on various library services. Libraries may be adept at providing a certain service that is of little value to faculty, and knowing this can guide librarians in allocating finite resources and staff time better. The educa- tion faculty were asked a general question about the importance of library research in their field, with the answer choices including very important, important, neutral, unimportant, or not applicable. The chi-square analysis of the responses indicated that they were not statistically significantly tied to the respondents’ in- stitution (df =11, p=0.647). The education faculty were asked to rate ten library services as very impor- tant, important, neutral, unimportant, or not applicable to their needs (see table 9). With regard to the importance education faculty members place on various library services, there seems to be a distinction between physical space, the services themselves and providing access to resources. Regardless of insti- tution surveyed, researchers are likely to get similar results when asking about the importance faculty place on libraries providing access to electronic journals, physical or electronic books, and data- bases. However, the value faculty place on services such as document delivery (p≤0.001) and interlibrary loan (p=0.020) is statistically significantly tied to their home institution. To better understand this difference, figures 1, 2, and 3 show the percentage of faculty indicating the importance they place on document delivery, library sub- scriptions to scholarly journals in print, and space to study and conduct research by institution. Responses of important and very important were combined, as were responses of unimportant and not applicable, because the researchers felt these responses would have similar im- plications for practitioners. It is interest- ing to note that the importance faculty place on document delivery was shown Table 9 Ranked Importance of library Services in Meeting Information Needs Correlated to Respondents’ Institution Pearson χ2 Value df p Very Important + Important (%) Document Delivery 82.887 44 0.000 39.0–91.3 Interlibrary Loan 65.368 44 0.020 53.7–95.7 Assistance from Library Personnel 59.836 44 0.056 40.7–79.3 E-access to Archives of Scholarly Journals 42.237 33 0.130 91.6–100 E-access to Current Scholarly Journals 41.468 33 0.148 94.6–100 Print Subscriptions to Scholarly Journals 51.237 44 0.211 35.5–78.0 Library Databases (e.g. ERIC) 36.373 33 0.314 80.9–100 Physical Book Collection 45.585 44 0.406 59.4–83.8 E-book Collection 42.575 44 0.533 57.1–86.6 Space to Study/Conduct Research 42.013 44 0.557 25.8–42.8 458 College & Research Libraries September 2013 How Unique Are Our Users? 459 to be statistically significantly associated to institution (p≤0.001). Yet, looking at the responses by institution when the percentage rating the service “very im- portant” and “important” are combined, it appears to matter very little which insti- tution is surveyed—the response would indicate that document delivery is an important library service. Pitt faculty had the smallest percentage of respondents rating document delivery as important, with only 39.0 percent rating it as either very important or important; only 19.0 percent of their faculty indicated it was unimportant or not applicable, so many remained neutral on the topic. Figure 2 depicts the percentage of fac- ulty indicating the importance they place on libraries providing access to print jour- nals by institution. The importance faculty FIgURe 1 Importance Placed on libraries Providing Document Delivery by Institution 0% 20% 40% 60% 80% 100% Texas A&M Texas Pitt Ohio State Missouri Kansas Iowa Illinois Georgia Florida State Connecticut Arizona State Unimportant/NA Neutral Very Important + Important FIgURe 2 Importance Placed on libraries Providing access to Print Journals by Institution 0% 20% 40% 60% 80% 100% Texas A&M Texas Pitt Ohio State Missouri Kansas Iowa Illinois Georgia Florida State Connecticut Arizona State Unimportant/NA Neutral Very Important + Important How Unique Are Our Users? 459 place on libraries providing access to print journals was not shown to be statisti- cally significantly associated to institution (p=0.221). This suggests that, regardless of the institution surveyed, researchers could expect to receive similar results. However, a look at figure 2 tells a somewhat different story. If, for example, the study was con- ducted solely at the University of Kansas, one might get the impression that faculty minimally value print access to journals, with only 35.5 percent of the respondents indicating it was very important or im- portant. By contrast, if the study had been conducted solely at Pitt, it would appear that education faculty still place a high value on print journals, with 78.0 percent rating it as very important or important. The difference in these findings would have entirely different implications for the allocation of collection development funds. At institutions where print is still highly valued, the cancellation of print subscrip- tions in favor of electronic would require greater preparation among library con- stituents; however, at institutions where print is deemed less important, librarians might easily be able to shift spending away from print subscriptions if they have not already done so. Figure 3 paints a similar picture; the importance faculty place on libraries providing a space to study and conduct research was not shown to be statisti- cally significantly associated to institution (p=0.557). Again, while the chi-square analysis indicates that, regardless of the institution surveyed, researchers could expect to receive similar results, the re- sults themselves suggest otherwise. Of the twelve institutions included in the study, three had a greater percentage of faculty indicating space to study or conduct re- search was important or very important than indicating it was unimportant or not applicable. The University of Missouri had the most faculty members indicat- ing library space was important to them, with 42.8 percent of the faculty rating it as either important or very important; however, 21.4 percent of Missouri’s fac- ulty rated it as either unimportant or not applicable. At the University of Iowa, 30.4 percent of faculty members rated library space to study or conduct research as important, while 60.8 percent rated it as unimportant or not applicable. This sug- gests that the extent of importance varies by institution, but the fact that no institu- tion had a majority of faculty members FIgURe 3 Importance Placed on libraries Providing Space to Study/Conduct Research by Institution 0% 20% 40% 60% 80% 100% Texas A&M Texas Pitt Ohio State Missouri Kansas Iowa Illinois Georgia Florida State Connecticut Arizona State Unimportant/NA Neutral Very Important + Important 460 College & Research Libraries September 2013 How Unique Are Our Users? 461 least monthly. They found that the ranges were not so wide that a practitioner would draw differing conclusions on the popu- lation based on the results at any single institution. For this particular question, the types of information sources a faculty member would need to perform most of these tasks are quite similar; so if a ma- jority of faculty indicate they are seeking information for any one of these tasks, practitioners are likely to feel validated in their ongoing collection development activities. Finding Information In both studies, faculty were asked to rank the importance of a list of sources for aiding their research. For engineering faculty, the ranked importance of three sources was statistically significantly as- sociated with the respondents’ institution, while that was the case for only two of the sources for education faculty (see table 11). However, for this question, the two showing a statistically significant associa- tion for the education faculty were also statistically significant for the engineering faculty, and both had to do with faculty- student communication. This theme resonates in responses to the next question faculty answered in regard to the sources they use to remain indicating space for study and research was highly valued that the data found at one institution may indeed be “good enough” for other institutions. A Tale of Two Studies As indicated in the methodology, this study replicates a study conducted on engineering faculty in the fall of 2009 and is meant to test whether the findings of the initial study hold true in a different disci- pline. By comparing the data between the studies, it is possible to begin to determine what types of questions asked in single- institution user studies can be accepted as descriptive of an equivalent population at other, similar institutions. Information Use In both studies, the researchers asked how frequently the faculty members sought information to complete a series of tasks. For the engineering faculty, the frequency with which faculty sought information to complete three of the six listed tasks was statistically significantly associated to institution; for the education faculty, only one of the tasks was statistically signifi- cantly associated to institution (see table 10). In both studies, the researchers then looked at the range of the percentage of faculty reporting doing these activities at Table 10 Frequency of Information Seeking to Complete Tasks Correlated to Respondents’ Institution Compared by Discipline Education Faculty Engineering Faculty Pearson χ2 Value df p Pearson χ2 Value df p Prepare a New Research Proposal/Grant Application 83.677 55 0.008 82.022 75 0.271 Professional Development/Remain Current in the Field 69.832 55 0.086 102.010 75 0.021 Prepare for Student Lectures 61.76 55 0.247 63.855 75 0.817 Write/Research for Publication 61.275 55 0.261 103.353 75 0.017 Determine Protocols 57.663 55 0.377 68.415 75 0.691 Prepare for a Conference Presentation 50.745 55 0.638 98.936 75 0.034 Research Patents Not Available 81.065 75 0.296 How Unique Are Our Users? 461 current in their field. For both engineering and education faculty, personal commu- nication was the only response statisti- cally significantly associated to institu- tion (education faculty: df=11, p=0.037; engineering faculty: df=15, p=0.001). From a practitioner standpoint, these findings have little implication for day-to-day li- brary work unless the library makes an ef- fort to actively foster relationships among departmental faculty and students. Next, faculty in each study were asked which tools they use to discover less re- cent journal articles. For the engineering faculty, no tools were statistically signifi- cantly associated with the institution, and only one tool, citations at the end of book chapters, was statistically significantly associated for education faculty (df=11, p=0.033). From a practitioner standpoint, a library is going to continue to order books, in some format, regardless of whether or not faculty are using the cita- tions at the end of book chapters to locate older materials. Since this is not a primary purpose for ordering books, it seems that a library’s collection development policy would not be changed based on this dis- covery, regardless of whether only 40.6 percent of faculty indicated using this method of discovery in a study. Knowing that faculty do or do not use citations at the end of book chapters to become aware of less current journal articles might inform instruction practices; but, again, if a librarian covers citation tracking in an instruction session, it would likely be a part of a larger lesson on citation track- ing in general, regardless of whether the citations come from a book or journal. Using the Library In all likelihood, practitioners conduct- ing similar studies at their institutions are most interested in how their library is meeting the users’ needs and how services might be improved. To this end, faculty in both studies were asked to rank the importance of library services in meeting their information needs. In each study, there was a marked difference in responses for collections, space, and ser- vices. For both studies, document deliv- ery and interlibrary loan were statistically significantly associated to institution, while the ranked importance of resources and library facilities were not statistically significantly associated to institutions (see table 12). This seems to imply that, for the most part, one could trust the data from a study conducted at an institution similar to one’s own. Table 11 Ranked Importance of Sources for aiding Research Correlated to Respondents’ Institution Compared by Discipline Education Faculty Engineering Faculty Pearson χ2 Value df p Pearson χ2 Value df p E-mail Discussion with Students 57.466 33 0.005 62.173 45 0.046 Face-to-face Discussion with Students 49.197 33 0.035 64.360 45 0.031 Attendance at Conferences 45.445 33 0.073 62.318 45 0.044 Face-to-face Discussion with a Colleague 36.875 33 0.294 50.597 45 0.260 E-mail discussion with a Colleague 36.247 33 0.320 47.498 45 0.371 Internet Resources 33.745 33 0.431 61.466 45 0.052 Books 33.626 33 0.437 61.466 45 0.052 Scholarly Journals (in print/online) 33.109 33 0.462 41.088 45 0.638 Textbooks 29.235 33 0.655 42.685 45 0.570 462 College & Research Libraries September 2013 How Unique Are Our Users? 463 However, the researchers for each study highlighted responses by institu- tion for several of the listed services to aid in understanding the dangers in relying solely on another institution’s data for de- cision making. In the study of engineering faculty, the researchers note: If, for example, the study was con- ducted solely at UCLA, one might get the impression that more faculty find library assistance unimportant or not applicable than find it impor- tant (though 42.8% of the UCLA re- spondents remained neutral on the topic). This finding might call into question the necessity of reference departments at engineering libraries across the country. By contrast, if the study had been conducted at the University of Oklahoma, the results would suggest that assistance from library personnel was considered important to the majority of engi- neering faculty (63.4%) and have entirely different implications for practitioners.9 This difference has strong implications for practice and indicates that librarians should be cautious consumers of research studies and validate the findings lo- cally when feasible, especially when great stakes are concerned. Conclusion After the previous study of engineering faculty, the researchers questioned the ap- plicability of the findings to practitioners. Regardless of whether data gathered at one institution was comparable to similar populations at similar institutions, how could the data be used to inform practice and assist with decision making? This study used a modified survey instrument in hopes of providing more useful infor- mation for practitioners while maintain- ing many of the original questions for the sake of comparison. The findings of the current study of education faculty confirmed the find- ings of the previous study of engineering faculty. For the most part, responses to questions related to general productiv- ity, duties, and preferred sources are not Table 12 Ranked Importance of library Services in Meeting Information Needs Correlated to Respondents’ Institution by Discipline Education Faculty Engineering Faculty Pearson χ2 Value df p Pearson χ2 Value df p Document Delivery 82.887 44 0.000 108.881 60 0.000 Interlibrary Loan 65.368 44 0.020 127.392 60 0.000 Assistance from Library Personnel 59.836 44 0.056 72.878 60 0.123 E-access to Archives of Scholarly Journals 42.237 33 0.130 57.777 60 0.557 E-access to current Scholarly Journals 41.468 33 0.148 57.171 60 0.580 Print Subscriptions to Scholarly Journals 51.237 44 0.211 73.601 60 0.112 Library Databases 36.373 33 0.314 72.005 60 0.138 Physical Book Collection 45.585 44 0.406 54.691 60 0.669 E-book Collection 42.575 44 0.533 49.970 60 0.819 Space to Study/Conduct Research 42.013 44 0.557 70.777 60 0.161 Access to Laboratory Protocols Not Available 81.735 60 0.033 How Unique Are Our Users? 463 institution-specific. Librarians could rely on the data from a single-institution study to develop a better understanding of their own faculty’s information-seeking behav- iors in a given discipline, provided the in- stitutions are similar in size and research focus. Responses to questions about why faculty seek information and how they go about finding it were also fairly consistent among the various institutions. However, as common sense would seem to dictate, librarians should use caution when relying on others’ data for highly institution-specific areas such as services provided or strengths and weak- nesses in resource availability. The areas where there were the most statistically significant associations between responses and institutions related to areas with fairly clear local ties such as accessing items via an institutional website and services such as interlibrary loan or reference assistance. Essentially, librarians should proceed with caution and seek to validate findings lo- cally when the stakes are high. This study is the second of many needed to fully understand how to make survey results from information-seeking behavior studies relevant to practitioners. The researchers intend to further fine- tune the survey instrument to collect information that is meaningful to prac- titioners and then conduct future multi- institutional studies that examine faculty and/or students in other disciplines such as the humanities, fine arts, and/or life sciences. Given the inherent limitations of a purposive sample, it might also be ben- eficial to repeat the current study within the fields of education and engineering with a random sample of faculty at large research institutions. Notes 1. Virginia Wilson, “Applicability: What Is It? How Do You Find It?” Evidence Based Library and Information Practice 5, no. 2 (2010), available online at http://ejournals.library.ualberta.ca/index. php/EBLIP/article/view/8091 [accessed 4 August 2011]. 2. Ibid., 112. 3. Ray Lyons, “Statistical Correctness,” Library & Information Science Research 33 (2011): 92–95. 4. Elke Greifeneder and Michael S. Seadle, “Research for Practice: Avoiding Useless Results,” Library Hi Tech 28, no. 1 (2010): 5–7. 5. Ibid., 7. 6. Lili Luo, “Fusing Research into Practice: The Role of Research Methods Education,” Library & Information Science Research 33 (2011): 191–201. 7. Sarah Robbins, Debra Engel, and Christina Kulp, “How Unique Are Our Users? Comparing Responses Regarding the Information-Seeking Habits of Engineering Faculty,” College & Research Libraries 72, no. 6 (2011): 515–32; Maria A. Jankowska, “Identifying University Professors’ Informa- tion Needs in the Challenging Environment of Information and Communication Technologies,” Journal of Academic Librarianship 30, no. 1 (2004): 51–66; Anita Cannon, “Faculty Survey on Library Research Instruction,” RQ 33, no. 4 (1994): 524–41. 8. Accessed online at http://grad-schools.usnews.rankingsandreviews.com/best-graduate- schools/top-education-schools during July 2010. 9. Robbins, Engel, and Kulp, “How Unique Are Our Users?” 526.