Friday, December 13, 2013

How to evaluate these databases by acquisition?

It is important to have an initial assessment of the product to see it it meets the basic criteria. In essence these are the factors that will allow a title onto the desiderata list and those that will exclude it. This generally comes down on three things: a. cost, b. content, c. capability. The following questions will serve as a guide in choosing a database:
    1. Has the dataset been requested by the large number of users?
    2. Will it be heavily used by a large number body of readers from all sectors within the institution?
    3. Will it provide access to information which is not readily available in other formats?
    4. Will it provide a set of tools or features which is not readily and enhance the use of material?
    5. Will it offer the possibility of a potential cost or space saving through the cancellation offprint subscription?
    6. Does it have, or is it likely to have, any competitors?
    7. Is it essential to maintaining the research profile or competitiveness of a department/faculty/company?
    8. It is based in a subject area that has been relatively underrepresented by previous dataset purchase?
    9. Does it complement existing digital services?
    10. Has it received promise of funding?
    11. Is it a one-off purchase (ex. it does not add to recurrent cost?)
    12. Is a consortium deal possible or it regarded as good value for money?

After the title has been selected it is important to request a trial subscription or evaluation copy. The evaluation should draw the on a pool of experts and these would ideally represent:
·        Subject specialist
·        technical experts
·        readers or readers services
·        interface experts
·        Administrative personnel (to consider such things as the implications of the budget, the proposed license, etc.)

The following list of questions provides a fairly comprehensive guide of the types of issues an evaluation might cover. This would also be of use once a dataset has been purchased, at the renewal stage of the life cycle. The collation of work by other institutions drawing on Oxford University’ Evaluating a ( dataset and Harvard’s Stewards’ checklist of questions (http://

Evaluation Checklist:

Initial Details
·        Full title of the dataset
·        Publisher
·        Details of corresponding print publication (if applicable)
·        Pricing details
__ payment model
__standalone cot
__network cost
__ recurrent or one-off payment
__ additional fees to third parties or under national taxes
__(if available) average inflation for product over past few years
·        Type of product (e.g.) CD-ROM plus platform, internet access, tape, etc.)
·        Machine specifications used for evaluation  include details such as internet connection if applicable)
·        Date and time of evaluations
·        Brief description content

Setting up or accessing the dataset

Local Products
·        Is the dataset easy to install
·        Is it easy to network?
·        Was all the software needed supplied with the installation CD-ROM?
·        Were the instructions clear?
·        Does it appear to conflict with other functions of the computer/server?
·        Can the database be uninstalled easily?

         Remote access
·        Did you manage to access the site easily?
·        Was the URL intuitive?
·        Can you bookmark the site or section of it?
·        Does the service use any of the following; frames, cookies, Java or JavaScript?
·        What browsers will it run under, and what version is needed?

Both local and remote

·        Does the dataset require any plug-in?
·        Does it create an otherwise excessive burden on the client
·        Accessing the opening screen: on average, is this instantaneous, or does it take few seconds
·        Accessing the search screen: on average, is this instantaneous or does it take few seconds?
·        Have you ever experienced a message saying “too many connections/Users?”
·        Is yes, approximately how often does this occur?
·   Is there a clear and easily accessible list of the complete contents of the dataset? If this is a growing resource. Or one with updates, how accurate is the list?
·  Are there errors noticeable in the text, broken links, or anything else’s which would indicate a lapse in quality control?
·  Would you say that the dataset contains everything that is in the print version? What is the relationship between the electronic product and the print holdings of the institutions?
·  Have you seen any other publications to compare to this in content?
What are they and how they compare?
·        How often is the product updated?

     Interface and Searching
·        Is there a main menu or some kind of main list that outlines all the functions clearly?
·        Is the interface generally easy to use?
·        Is the navigation intuitive
·        Does it conform to accepted standards for disabled access?
·        Was it obvious how to browse or start a search?
·        Can you:
__ save searches?
__combine search words?
__modify searches?
__narrow/widen searches?
__use proximity searching?
__use Boolean operators?
__truncate search terms and/or use wildcards?
__limit by year/publication/type/language/other?
__check the indices and pick entries to search for?
__use any command line searches? Is it made clear how to structure these, or how to use special codes
·        Can you search the dataset via third-party software (ex. reference manager tool?)
·        Do you get a hint if your search achieved no results?
·        Can you see history of searches?
·        Can you save the search set for future use?
·        Can users set up or configure personalized accounts?
·        Can the dataset be cross-searched with others?
·        Sending a simple search and receiving the results (e.g. looking for a term in a single dataset) on average, is this instantaneous, or does it take few seconds?
·         Do results display automatically?
·        If not, is it intuitive how to access them?
·        Is there a choice of formats for display?
·        Can you mark useful results?
·        How many records can you mark/save/export at once?
·        Are the steps for saving/exporting/printing easy to understand?
·        Can you save to a floppy disk/hard disk/either?
·        Can you e-mail results?
·        In what format are records saved (ASCII, HTML, other)?
·        Can the results be exported to third-party software?

·        Is it easy to exit/logoff?
·        Are you prompted to save or lose your searches/results?
·        Is there a timeout system for idle session?

·        Is it possible for the administrator to change some settings of the database to suit the users’ needs?
·        What authentication system is employed (ex. Username/Password, simultaneous access, domain name, etc)?
·        Is there a choice of authentication systems on offer?
·        Can you add information to the dataset, such as local holdings? Is it easy to interface the dataset with other electronic products?

        Documentation and support

·        Is any documentation provided?
·        Is it sufficient, covering both installation and use?
·        Are there costumer support and technical support facilities?
·        Are these available by telephone/fax/email/the web?
·        If there is a website does it include information about system status?(ex. Is the system fully operational? Are connections being refused?, etc.)
·        Is there any training available? Is this free of charge?
·        If there is an online help system how useful is it?
·        How easy is it to access the online help mid-session?
·        How closely does it mirror the paper documentation?

General points

·        Is this product available on a different platform or through a different aggregator? If so, has this been evaluated?
·        Is this publication useful? In other words, could you predict a high demand for it (using existing requests for the product, use/importance of the print publication if applicable, ILL request and so on)?
·        Would you say this dataset presented good value for money?
·        Do you think it will be of use to people in the institution other than the proposer?
·        Did you detect any errors or problems either with the actual data or with the search software?
·        On average would you envisage using the dataset:
__lees than once a week?
__1-3 times per week?
__once a day?
__more than once a day?

References: What are proprietary database. Retrieved on December 05, 2013,   

Lee, S.D. (2002). Building an electronic resource collection: a practical guide. Great Britain: MPG Book Ltd.

No comments:

Post a Comment