This report is intended to be a short summary of the most interesting results from the online survey. It isn't a complete report by any means. For more information, you can view the response statistics.
Our methods were simple. We devised the short survey and publicized it through multiple channels for maximum exposure. We received 591 responses. Results were calculated as a percentage of the total responses for a question. Open-ended responses were collected, normalized (i.e., categorized under a controlled keyword), and counted. Cross-correlation was not performed, but may be at a future date.
Caveat: Nearly half (48.8%) the respondents were librarians, library staff, or information professionals. These respondents are generally more versed in information retrieval methods, which may be reflected in the data.
1. A majority of respondents indicated they were most interested in online journals and reference materials when they went online to look for information.
More than 80% of respondents selected these two types of digital resources (online journals 88.8%, reference materials 80.1%). None of the other responses for resources listed matched these high percentages (e.g., data 56.2%, images 53.2%).
2. Additionally, respondents indicated that these were the digital resources they were unable to find online.
This seems to point to an apparent trend, however, without further data analysis we don't know if these responses reflect differences in respondent roles. For instance, librarians might make up the majority of the 88.8% of respondents who indicated they were interested in online journals, while students might make up the majority of the 20.7% of respondents who indicated they were unable to find online journals. Cross-correlation may indicate that appropriate interfaces and publicity would be needed to reach certain audiences.
3. Top problems that respondents noted when looking for information online included not retrieving the resource itself, not finding older materials, having rights problems, not knowing of a comprehensive service for finding resources, and the ever-present problems with searching.
Respondents are clearly interested in being able to access the complete digital resource (67.1%), and not having to view the citation and take an additional step, whether that is performing another search or finding the resource in the physical library.
When respondents mentioned not being able to find older materials (32.2%), we considered this a plea for having more materials online. As one respondent put it: “It's great to have so many online journal articles! More please! I'd love to see more pre-1900 texts available in this format.” It seems we have a duty to perform, as the demand is clearly there.
Rights issues (8.9%) are sticky ones. Respondents want copyrighted items, which are often unavailable for digitization. A respondent states that he wants “free access to all the quality materials.”
One respondent says: “There are so many sources that I'm overwhelmed. I don't know what's out there.” A service that could offer one-stop, comprehensive access to all digital resources (8.2%) is, naturally, something OAIster hopes to address. Another respondent noted: "Can find most of what I need – but then again there may be resources out there I don't yet know I need." A comprehensive service could also benefit users interested in serendipitous searching.
Searching problems (13.7%) are nothing new. Particular problems that respondents noted were that they:
- don't have enough information to describe something
- don't know what words to use
- don't know how to narrow searches
- don't find relevant materials
- don't know where to start looking
One respondent puts it rather bleakly: “It's not that I haven't found them, it's the usual difficulty of finding the correct key word/phrase to bring it up. I think this is just an unsolvable problem.” That's probably an entirely separate discussion, but it's worth noting that we need to address these issues through a combination of designing appropriate interfaces, education and assistance.
4. Surprisingly, a number of respondents (5%) indicated that they were generally successful in finding resources online.
Typical responses included “I've been able to find what I need” and “So far have found all I wanted.” These comments could reflect the abilities of expert searchers, but could also point to the fact that searchers might not be aware of what they are missing, aren't looking for in-depth or comprehensive results, or, of course, are generally satisfied with the resources they retrieve. This would be a fascinating area for further research.
5. Several potential features of a service like OAIster were of interest to respondents, including looking for resources in a certain subject area, using a service that is continually updated, and searching the full text of the resource.
Respondents wanted to be able to narrow or limit their searching scope to a particular subject area (82.6%). The high percentage of responses to this question points to a pressing need for some type of subject categorization so that users can perform their searches within certain subject areas. We hope to research this issue and its solutions towards the end of this year, and hopefully into the next.
Respondents also indicated that using a service that featured a wide range of subject areas was not one of their top features (53%). One interpretation of this result is that it indicates that while we wish to offer an all-encompassing service for all types of audiences and all types of uses, we also need to make specific subject area access available and easy to use.
The desire for continuous updating of the service (78.4%) implies that respondents don't want to use a service that does not add resources on a regular basis or that is stagnant in the types of services it offers. Naturally, this is a challenge, but one that needs significant attention for the continued popularity of a service.
The desire to search full text (78.2%) is also nothing new. Providing the option to do so can be problematic as it can result in way too much information for the user. Our desire is to provide users with more tailored results by having them search within metadata that has been created to reflect the content of the resource. One respondent put this succinctly: “Tagged information so I don't get fire hosed.” If time permits, we hope to research the potential for offering even better search results through metadata normalization.
We particularly liked one comment we received: “You will never beat Google. No way.”
Probably true. However, it's not our intention to beat Google, but to provide an adjunct method for accessing information online. Our hope is that by providing a comprehensive service that caters to user needs – e.g., finding resources by subject, finding resources by format, retrieving the full resource – and addresses multiple searching problems, we can provide access to more, and more varied, useful and informative digital resources that are currently difficult to find.