This section asks LAIP participants to rate the effectiveness of the cluster concept in terms of organization, administration, and support. Availability (up-time) and other issues of shared access are assessed, along with preferences for system interface.

Cooperative management of clusters, system support from cluster hosts, and sharing centralized databases and software are ranked highly satisfactory. SUNYNet support and the geographic organization of clusters are evaluated slightly lower. Comments describe the cluster environment as very effective in developing cooperation and mutual support among LAIP members; however, a couple of comments mention that existing consortia arrangements should have been considered.

The availability (up-time) of the local campus network, cluster host network, and SUNYNet are rated above satisfactory. Problems due to security issues are rated as low. The impact of scheduled downtime is rated as moderate.

Responses on questions relating to the OPAC interface show that LAIP participants want both an Internet browser and a text-based interface. Ease of use and sophisticated searching capabilities are described as most important with the OPAC interface.

For staff functions, text-based and graphical interfaces are rated slightly more important than an Internet web browser. Comments focus on speed of interface and response time: routine check-in, linking, and circulation transactions cannot take any longer with a graphical interface than with a text-based interface. The multi-tasking of Windows is needed, but it has to work quickly.

MultiLIS / LAIP Software

This section asks libraries to do two things: (1) rate the MultiLIS software by broad function, and (2) rate the usefulness of having more of a common, shared catalog than multiple catalogs.

LAIP libraries rate the MultiLIS software module by module, and overall place them into two groups, with relatively tight agreement among libraries about those ratings. The higher-rated modules all average better than satisfactory, are rated by 95-100% of libraries, and include the core functions of Circulation, Cataloging, Authority Control, OPAC, and Policies Management.

The lower-rated group of MultiLIS software modules average at the bottom end of satisfactory to poor. Standard Reports, Serials Control, Inventory Control, and Report Generator (the report writer) all average in the lowest satisfactory rating. The Backup Circulation module is rated below the others, in the poor category.

Although MultiLIS documentation was not offered for rating, comments in the Summary Section on LAIP weaknesses did not overlook it: "weak documentation from vendor"; "accepting such poor documentation ... passed off as 'user manuals'"; and "seems to be missing pieces, poor indexing."

Document Types are a customizable set of non-MARC, authority-controlled identifiers in MultiLIS. The overwhelming majority of LAIP libraries rate the search and retrieval capability of Document Types as extremely useful, and extremely important to retain in future library systems. The Committee recommends that the Document Type functionality be required in future library systems.

Sharing authority records is rated as highly helpful by the majority of the 38 libraries that rated it. An equally high rating is given to the coordination of the authority work within clusters. A slightly lower opinion overall -- but still above the middle -- is given to how well the authority clean-up is actually being carried out within clusters. Focusing on individual clusters: the libraries in the Plattsburgh cluster unanimously give the highest possible rating to the implementation of their ongoing clean-up; the Albany cluster generally rates its cleanup between extremely and moderately well done; the other three clusters generally rate their ongoing clean-up as moderately well done.

This positive opinion of sharing authority records and their maintenance (beyond the automated processes incorporated into the MultiLIS software and the authority updating contract with BNA) is consistent with most of the other very positive ratings accorded to sharing the management, system support, and databases that the cluster concept requires.

When considering sharing bibliographic records, libraries emphasize the need to retain their own unique records rather than sharing common records with cluster holdings attached. Due to ambiguous wording in survey questions 10-12, the Committee cannot conclude how useful libraries find the ability to view the holdings of other libraries in their cluster. From the many comments, librarians regard the capability of seeing other libraries' holdings as less important than having a system that will describe and retrieve as precisely as possible what their own collections contain.

All of the MARC fields listed in the survey are rated by libraries as at least moderately important for characterizing their unique collections and for the retrieval needs of their users. The two highly important fields are locally-added subject headings (60x, 61x, 65x), and local notes (normally added as 590's when cataloging in OCLC but converted to 500's by MultiLIS). Locally-added personal headings (1xx, 7xx), local subject headings (69x), and locally-added series (4xx) are moderately important for reflecting unique needs. Other fields mentioned in the comments were the contents note (505), links to WWW sites (856), and periodicals "Indexed by" notes (510).

Process / Migration / Implementation

The purpose of this section is to assess the original process for acquiring and implementing MultiLIS software.

The process for acquiring the automated system is viewed as satisfactory by most libraries. The initial contract negotiation process is rated slightly more satisfactory than ongoing negotiations, but the few comments on this topic recognize that difficulties are more attributable to the vendor and to State processes, rather than to inattention by LAIP staff.

The phased implementation of clusters, campuses, and modules is rated as highly effective, and the coordination of accompanying training needs is rated as very satisfactory, although some negative comments were expressed about each. A similar approach might be employed for migrating to a new system with greater consideration given to the desire and perceived need to implement several modules simultaneously, so that libraries can maintain workflow.

Initial clean-up of databases by both SUNY/OCLC and BNA is rated highly satisfactory by most libraries.

Of the 40 libraries participating in LAIP, approximately 10 libraries responded to question 18 concerning migration from a previous online system. Since libraries that did not migrate were requested to respond N/A, this indicates that about 25% of LAIP libraries migrated from some other system. While one library responded "Not at all important" to all of the items in question 18, the remaining libraries rate data integrity in all categories as highly or extremely important, with bibliographic data of highest concern to the largest number of respondents, followed by item records and authority records. User records, circulation status, and serial item records are rated as extremely important by 2 of 3 libraries responding.

During the implementation phase, communication between the LAIP Office and individual libraries is characterized as extremely useful. Assistance with the MultiLIS profile, assistance with other vendor profiles, and the onsite visit by LAIP staff are also rated as highly to extremely useful, with mean scores varying by only hundredths of a point. The local test database, used at least once a week by most libraries during implementation (question 24.a.), is rated as highly useful by most respondents. The Committee recommends continuation of this service. Direct assistance from MultiLIS is rated as only moderately useful. The data are supported by several positive comments citing the assistance from LAIP staff as "invaluable," "very helpful," and "fantastic."

Support / Communication

The LAIP Office excels in its support and communication to its constituents. LAIP participants are extremely appreciative and complimentary of the quality of support received from the LAIP staff. Although some survey comments express concern that support and communication may have been weak at the beginning of the project, these services have improved over time and are now rated very useful.

Training classes are rated highly useful. LAIP training classes are rated as more useful than the VAX/VMS training provided by the Training Center or the e-mail class on Report Production. Survey comments confirm the quality of LAIP training classes. The Committee recommends that LAIP continue to offer training classes for users new to MultiLIS, as well as training classes for new modules and/or systems as they are implemented.

Early documentation was only moderately useful, with the LAIP Letter more useful than the LAIP Operations Manual and much more useful than the MultiLIS Implementation Manual. Several libraries commented that the LAIP Operations Manual was published after their implementations, thus reducing its potential usefulness.

Overall communication between the SLSC and the LAIP community is rated highly satisfactory by all libraries. Many comments support this rating. Other comments encourage frequent and wide distribution of information beyond the library directors, especially via the SLSCk newsletter and the listservs, so that all necessary parties can be informed.

Although overall communication is ranked highly satisfactory by all libraries, no single mechanism of communication is ranked quite as high. The SLSCk newsletter and communications to library directors are rated highly useful, with all libraries rating them as at least moderately useful. Cluster listservs are also rated highly useful. The MULTILIS listserv and minutes from the LASG and the SLSC Advisory Board are all rated as moderately to highly useful; however, several comments cite the lateness and thus the lack of usefulness of the LASG and SLSC Advisory Board minutes. Interestingly, among the various mechanisms for communication, the LIB_STAFF account ranks highest (as highly useful) and the web site the lowest; however, the resources of each are virtually identical.

The LAIP services available via the LIB_STAFF account and on the gopher/web are rated moderately to highly useful, with Member Contributions rated slightly higher than the sample database and LAIP documentation. Searching the Target database is rated less useful, but still is ranked moderately useful. These services were used extensively during implementation, but receive much less use currently. The Committee recommends that similar services be set up for any new or future implementations.

Member libraries are extremely satisfied with the LAIP Office as both the first line of support for campuses and the primary point of contact with the vendor. 35 of 40 libraries responded to these questions with the highest possible rating. Since support is being transferred from LAIP to the vendor just as this survey is completed, the Committee recommends that the issue of who provides first-line support be revisited in the future.

Cost / Governance

This section asks participants to assess the value of the LAIP project in terms of campus and library costs, equitable pricing model, and contributions from various advisory boards, user groups, and cluster representatives.

The LASG Financial Plan: The fairness of the cost model (per record and flat fee charges) is rated just above satisfactory. LAIP expenditures are perceived to return a high value-for-money, although there are some questions from campuses about how LAIP expenditures might compare to overall costs for other automated systems. Almost all libraries rate the set-aside fund as extremely important to fund hardware replacement, software development, and special projects. While expenditures from the set-aside fund are rated as highly satisfactory, there is some confusion about how the set-aside has been spent to date. The Committee recommends that an annual accounting report be issued to members.

On the question of value, MultiLIS is rated highly satisfactory when including all the related campus expenditures for library, campus and people resources. Value-for-money of MultiLIS to each of several library constituencies (campus, library staff, off-campus) is rated higher than satisfactory, with comments indicating that the value is not as apparent or high for the off-campus communities.

LAIP libraries are undecided about the decision to exclude non-SUNY libraries from LAIP. Written comments indicate that excluding non-SUNY libraries was a good idea initially, but sentiment is expressed that non-SUNY institutions should be considered for inclusion in the future if that will yield advantages such as controlling costs and encouraging sharing.

The contributions of LASG and the SLSC Advisory Board are rated highly satisfactory, closely followed by NYSMUG. Communication by Cluster Coordinators, NYSMUG Enhancement Representatives, and Cluster Authority Control Coordinators is rated close to excellent. Communication by SLSC Advisory Board Representatives is rated highly effective. Communication by cluster host staff and LASG Representatives is rated only moderately effective. Ratings for communication from the cluster host range from extremely effective for the Albany cluster down to moderately effective for the Buffalo cluster.

Of concern is the difference in ratings for LASG, with its contributions rated highly satisfactory, but communication from its representatives rated only moderately effective. Also, comments in the Support and Communications section note the lateness of minutes. The Committee recommends that LASG and its representatives increase their efforts to communicate their important work to the SUNY community.

Summary

Pages of testimony from the 40 LAIP campuses proclaim the success of LAIP in terms of cost effectiveness and cooperative resource sharing. The overall evaluation of LAIP received close to an excellent rating on the survey. 38 of the 40 libraries say they will stay on board with the next generation LAIP project. Listed among the strengths of this program are: SUNY-wide shared resources of talent and expertise; pooling of computer resources; economy of scale and cost effectiveness; and the expertise and responsiveness resulting from one central office.

In spite of the unfailing support provided by LAIP staff members, their vulnerability to SUNY budget woes has been devastating to the project. Along with the loss of LAIP staff, poor vendor responsiveness and the MultiLIS software are cited as weaknesses of the program.

Consistent with the high rating given to cluster governance elsewhere in the survey, respondents show extremely high agreement on keeping the same number of clusters (five) in the future. Very few libraries indicate there should be less than five clusters, and even fewer advocate eliminating the cluster concept altogether.

The question of future funding received not only the highest level of agreement on the survey (with a standard deviation of .27), but also received the highest mean ranking (4.93 out of a possible 5). LAIP libraries are unanimous about the need for state appropriations for the next generation library system.

Conclusion

In reviewing both the statistical data and written comments from LAIP members, it is clear that the Library Automation Implementation Program works because of the people who manage it: librarians and staff at 40 libraries across the state allied with the centralized support staff in the LAIP Office. Also contributing to the success of LAIP is the cost effectiveness which results from shared expertise and pooled resources. Most of the problems cited by members reflect obstacles beyond the control of individual libraries--instability at System Administration, SUNY budget woes, and an unresponsive vendor.