WebSurveyor Analysis

Survey: CDL-hosted Database Test Accounts

Author:
Responses Received: 276

"Other" responses:

research staff

postdoctoral fellow

research administration

Postdoc

Administrative Staff-CDC

Field Assistant for UC Coop Extension

Reference Librarian

Librarian

post-doc

Research staff

pgr

Librarian with CBSR/selector for English

Librarian at CBSR/selector for English

Librarian

LIBRARIAN

Librarian

research assoc

"Other" responses:

CDL

UC Coop Extension Office, Sonoma County

affiliated library

Additional Comments

what's the difference in ABI editions? why more full-text and page images in Bell and Howell?

silverplatter can link to CDL Periodical Titles; can Bell and Howell?

What bothered me is the lack of display of the database name. I made my search on trust that I was in ABI. Lookinf at my results, I assumed I was in ABI.

I liked the Gale product MUCH better than the ISI product.

The best.

Yuk.

Not as streamlined and easy to use as OCLC. Inferior access to full text.

Horrible, because of my comment at Question 8.

ABI article images are the THE advantage for the proquest version. We must upgrade!

INSPEC before 1968 was not available. Any database UC would buy should have papers before 1968 as well.

Very bad.

Axiom is not good with author search ( my najor searc mode). Ovid is slightly better.

this version of BIOSIS seems far superior to the Ovid version on first use

The interface is counterintuitive. tools such as these will lead to more frustration and mroe deamnds for librarians, not less. we cannot work such interfaces without training--that is what librarians are for. finally, even Google does a better job of finding things.

This is the best looking interface for INSPEC of the three choices. It does not require me to learn INSPEC syntax. however, if none of htese tools are going to interface with the melvyl catalog, what the hell is the point?

give us an interface that does not require a web browser. We'll sort out the rest. melvyl is much better than this in terms of what I can do in a small amount of my time. Note: it might be different if the help buttons worked, but they didnt. melvyl has help on all subjects and has EXAMPLES.

I love the innovations the library has been making over the past four years, and I can tell you that researchers, both professors and students, love them too. However, webspirs is not one of your best databases.

Thank you for the opportunity to contribute to making the library better. We do appreciate your hard work.

Thanks for taking the time to make the databases more efficient and user friendly. We appreciate your efforts.

ABI Inform is a great database; very important; B&H seems to be a good vendor

One search I did ended up with just one result, which automatically took me to the full article in html, along with the abstract. I couldn't figure out how to print just the citation and the abstract, and found it difficult to tell where the abstract left off and the article began. Hope someone will pass on to Gale that this database and all others should be tested with naive users. Just wondering, too, if this is the only option for MAGS, or if there are others-- e.g., a similar but different database like ProQuest (I haven't used ProQuest, so am not sure of coverage & quality, but I know other libraries do.) It would be nice to be able to make some comparisons. Esther Grassian, UCLA College Library,estherg@library.ucla.edu

In general I find CSA interface to be poorly designed, not intuitive, not user-friendly.

Again, I hope someone passes along to Gale the need for naive user testing and inclusion of other advanced features. Also, I'm used to CSA and appreciate the flexibility offered by some of its advanced features, and will comment on those shortly. Esther Grassian UCLA College Library estherg@library.ucla.edu

this one was much better than silverplatter.

the other vendor was MUCH better

My impression was that the PsychInfo version unsed at UCI Melvyl is less inclusive than the Ovid version.

I tried several vendors/interfaces, and Ovid is the best by far.

The learning curve will be heavy, especially for teaching at the reference desk.

The first link to holdings I tried - to BMJ - was not correct. It appears that UCB has no holdings after 1980 accession#: 11273241 Not sufficient for heavy users -- health sciences professionals and grad students. simple features maybe fine for UG's - but advanced features so clumsy.....

I don't see any links to full text articles in here. I assume that will be added later? It is essential that these vendor databases connect with Endnote. Nobody prints out references and types them back in anymore.

Terribly inefficient. It should list the same exact results (or close to) for the same search in telnet and does not. I would have missed half of what I was looking for.

Horrible responses - not comparable to same exact search on telnet The same author yielded over 100 responses on telnet and a mere 15 on this system. Simply awful.

The only one of 3 psych info searches that worked as well as telnet

I dnn't know if I like it or dislike it; but after you get the results from your initial search terms a list is produced of secondary ways to focus and you have to choose. Patrons may be frightened at first faced with so much but if they use the database frequently they WILL grow accustomed to it.

I think Ovid is the second best vendor behind Silverplatter.

This is the worst verndor in my opinion. Silverplatter is the best, followed by Ovid.

Our on-line library is a real source of pride and accomplishment and I hope it stays that way. The chemical sciences are unique because most compounds can only be described using drawn structures rather than typed words. Thus, Beilstein Commander (for reactions) AND SciFinder Scholar (for chemical substances) are absolutely essential. SciFinder Scholar has exceptional bibliographic coverage for chemistry, but it does not allow full text searching. Thus, chemists will prefer to augment Beilstein and SciFinder Scholar with a full text search engine (for Current Contents) witht the best full text coverage.

PubMed is better than the telnet version of Medline.

I was a little confused as to why a good portion of the pdfs were not scanned images. If they don't have the rights to the images and they already offer the option to reformat the text for the browser, why bother with creating a pdf?

same as expanded academic...why bother with a pdf, if it truly isn't a scanned image. Some of the scanned images could have been scanned at a higher resolution. Some barely legible on the monitor, and not much better printed out.

Arrow that moves back and forth while system works on your request is annoying. Help screen hard to read.

Please work on refining the userfriendliess.

When I searched MLA (SP), I got 50 hits for George Woodyard's name; but OVID only had 39 (I don't know why this happened) Rhonda--MLA Test with Ovid platform 4-30-01

This is an okay platform, but OCLC needs to quit selling its other products on the top of its individual database screens. When/if we dont subscribe to these products, we have just increased the expectations of our users, without being able to help them.

Extra step required for basic limits (had to go to another screen) Limits not very obvious (text indicator yes?...how about full-text!) When initially entering SilverPlatter, annoying that you have to select the the individual files to search. Pre 1985? Didn't like it as much as Bell & Howell.

Subject headings should really be part of "extended citation" ESPECIALLY since there is no searchable thesaurus.

as with all silverplatter products, it's annoying to have to select the time/disc you want to search inadvance. An extra step is required to get to the limits Search builder is a poor substitute for an advanced search option

Like it better than silverplatter & know I'll like it better than OCLC. Less flexible download,print, and e-mail options than silverplatter, however.

like it better than silverplatter, but not as much as CSA Why offer an option...limit to full-text...when there is no full-text in the database? What's the S inf GeoRefS stand for?

very nice! assume they link to holdings as well?

Using the 'browser print' option with Netscape 4.7 and a Macintosh causes a problem: the document re-displays and the Mac's print function is engaged. After the user hits 'print' the Mac dialog box for printing appears again. The user must hit 'cancel' to get out of the loop. I was not able to replicate the problem using Internet Explorer for Mac nor using a Windows computer with either browswer. This bug should be fixed before full patron access is implemented.

Functional SFX enabled is a plus. Didn't like it as much as Axiom.

Although better than the other three, this INSPEC still was not as good as the current INSPEC we've been using at MELVYL, which already has some problems. I could not find a way to e-mail my search results, either.

-Browser printing is problematic with Mac computers and Netscape 4.7. After choosing this option, the Mac printing function is invoked automatically and the system continues to prompt me to print until I hit 'cancel' on the Mac print dialog box. - If I mark a record in the full display, then hit 'back' in my browser, the record is not added to the marked list. If I mark the record in the full display and hit the Infotrac button "Back to citations", the record is saved. This is confusing.

Overall, like better than LANL, even though I didn't see any links to full-text or SDI, but not as much as Axiom.

The word "basic" on the icon for basic searches might throw off those not comfortable with computer. They may interpret it as the BASIC programming language, thinking that rather than helping them, it is a button reserved for programmer types.

Nice, easily searchable database!

Ovid is clearly inferior to OCLC

does the ovid version of mla contain the thesaurus and MLA Directory of Periodicals?

Silverplatter is NOT an acceptable replacement for MELVYL ABI.

ABI on Proquest is the only viable option to MELVYL ABI. It is perfect, but is much better than SilverPlatter.

inferior to present ABI access. PLEASE DO NOT CHANGE

current system is easier to use--one search for all years...do not see any advantage in changing

Not a comfortable feeling. Probably the low bidder. Too generic.

This is a clearly inefficient search interface. I would not recommend.

The icons and text associated with the icons are to small. From the perspective of a typical users, Ovid's ERIC interface would be difficult for most users.

Some of the links to the CDL were not helpful, i.e. they were misleading (i.e. the dates for the articles online were not listed)

I also dislike the fact that there users are presented with an additional selection page(s). e.g., when you select an Advanced Search you are given an additional step of having to "Continue to Advanced Search". Also, the user's ability to modify an existing search is limited.

Unable to test. PsycINFO was not listed as an available database. However, I do like the OCLC interface.

The icons and the text associated with the icons themselves were to small. The search interface would be difficult for most users.

I don't recommend this interface.

Infotrac has a clear interface that is not cluttered with features. It is easy for patrons to use and allows for minimal instruction for this use. However, it is missing useful features that are available in CSA. Please see CSA comments for PAIS from UCR.

These comments also apply to the other Gale testdatabases I looked at since they seem to have the same basic layout/interface

The CSA databases can be difficult to use for beginners. However, the information for citations is more complete than Infotrac and the CSA databases have the ability to link to a library's holdings. Once the learning curve is mastered I consider CSA PAIS a better database than Infrotrac PAIS.

I liked ovid better than silver platter for medline, the data base I use most often.

I liked Ovid better than Silver platter

I like pub med for medline searches and it matches with ref manager (which I use) very well.

It's important that we have access to these databases from our home computers as well as from on-campus IP addresses.

I did not try out the enhanced service but did look for the ability to set up an alerting service which Axiom has. This feature is a high priority item for my users, as is the ability to download into bib software (EndNote etc). From an ILL staff viewpoint, I liked that Copyright Clearance center information is fully listed in the records.

Very user friendly. I liked LANL and Axiom about equally but for different reasons.

fast response

Field search syntax complex for novice users Need to explain use of "combine checked" button bar (Provide ? for "and" and "or"

love it

It took a long time to find my way through the screen clutter at first; didn't automatically understand how to return to the collective display once I had gone to a long record display.

I hate this interface -- find it very NON-intuitive, very clunky to navigate, and definitely confusing to the beginning user. I have had numerous complaints from students and faculty about the interface -- it seems to put up obstacles to finding one's citations, rather than guiding one to them. Also the thesaurus/index is next to useless, and a real time-waster as you sift through screen after screen of topics arranged alphabetically, rather than being grouped by concept.

There were times in my searching when it looked like clicking a link would take me to my citations; instead, it took me to another link that then took me to my citations. It gets tiresome, especially if you're doing multiple searches.

There seemed to be excessive steps in Basic Search to get to a listing of citations. The little icons across the top were annoying until I figured out their functions -- then I found them great! Under "Record Format" in "Citation Display" feature -- what in the world is "Ovid"?

Would not let me in. (I am working from a machine on the UCI campus.)

Bell and Howell ABI much easier to use!

Regarding this survey: I'd like to see a "What feature(s) did you like the least?" question after #9. #10 is not sufficient. The survey could also have provided the option of emailing the respondent a copy of the completed survey. (That would have been helpful in collecting data locally.) The small scrolling text boxes make it difficult to print out a copy. Regarding the Database: Overall screen design is too busy, with too many different places to search to find the function you're looking for. Interface should be simplified. The frames may also cause problems... For instance, why should your "Help" button disappear when you scroll through the results list? If you limit to full text articles, why do you get a line that says "At least 10 articles are full text"? Aren't they all?

I preferrd Bell & Howell for ABI/Inform. However, I have several suggestions for CDL to negotiate for improvement: 1. Add more sort options, such as by journal/source, reverse chronological order 2. Showing number of articles should be default. 3. In naturla langaguage search option page, add otpions for a. limit to peer reviewed; b. show number of articles 4. Add "Subject List" link in the Guided Search page. Thanks.

why are we changing from melvile

why are we changing from melvyl?

LANL is much better but Melvil sytem is the best. Not because I am used to it but because it has more options to modify search

This system is far superior to OVID but still does not measure up to Melvil. I would suggest selling our Melvil system to one of these vendors and letting them manage it or we should continue managing melvil ourseves.

wish we were sticking with the Melvyl system, but if we have to change could adapt to the lanl system. Melvyl seems to have more options for power searching, but probably just familiary with the lanl system enough to know how to fully utilize its capabilities.

I would prefer to stick with the Melvyl system. Easier power searching

This is probably the better of the two ABI versions, which is kind of a sad comment. The displays of search results are good, but searching by subject is confusing. The help screens are very good but how many people take the time to read help screens any more? While I would vote for this version over the SilverPlatter version, I think it's not a good thing to have a hodgepodge of interfaces for the old CDL databases. There would be more commonality of interfaces if SilverPlatter were chosen.

I preferred ISI for CC. I can't answer #5 since there is no email option.

In a previous institution for the past 10 years I used SilverPlatter Medline and PubMed (since it became available). While I have not used SilverPlatter for the past few months, of the three I think PubMed would be the easiest for undergraduates. The ONLY reason I would see purchasing OVID or SilverPlatter MEDLINE would be if CDL is going to go with one of these search engines for a majority of databases--not just 1 or 2.

I liked the presentation and arrangement; it was easy to read and choices were easy to find.

Misleading red dot and back arrow in record view looks like a link, but isn't. On search screen, pull-down menu says "Click here for database information". There's a button on the right that says "Database Information". Is there a difference? How is the user to choose? Add to that Help and About SilverPlatter, and you've got too many branches to explore to find what you need. On one record, clicking on "View complete record" brought up an empty window.

Felt very limited in search results. Didn't get the number of hits and subject searching wasn't as helpful. Didn't like the little icons above Author, Title, etc.

Get duplicate entries in search results which takes time to sort through.

Database has the feel of being in transition from design for pros (mediated searching) to use by amateurs (unmediated).

Running the same searches in this version of ABi and in the ProQuest and CDL versions, SilverPlatter consistently came up with smaller results, and fewer full-text articles.

the default display is difficult to read, particularly the information in brackets it is not obvious that you can change display format until you scroll down to the bottom of the page I couldn't figure out if you could modify an existing search

If we subscribe to this database must include "Locate Document" feature.

Couldn't get to.

This database is a mess. Buttons are poorly placed and have tiny identifying print. Have to click on "Select more than one database" to be able to search all files of PsychInfo at once. Use of mapping function (and it is the default) is confusing. "Click to begin search" buttons down left side small and dim - at first I thought it was wallpaper. In "Citation Manager" sort keys not clear. To print you have to select "print preview" (whether you think you need/want on or not), then print from browser - but it doesn't tell you that. Look - if a reference lib- rarian has "figure it out" it's not a database for most under- grads.

This is a test message from John Kupersmith at the California Digital Library. If it seems to be cut off and does not end with an e-mail sig block, please notify me at john.kupersmith@ucop.edu ... One of our campus librarians has expressed concern that this survey form may truncate respondents' comments if they are longer than 4 rows of 40 characters. He bases this on the tag

. However, checking an HTML guide suggests that this tag simply defines the size of the input box displayed to the user, not the amount of text the user can enter. I'm writing this long comment to test this idea. --jk
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 John Kupersmith 
 Service Design Analyst, California Digital Library 
 University of California 
 1111 Franklin St., Oakland, CA 94607-5200
 john.kupersmith@ucop.edu tel: 510-987-0554
 http://www.cdlib.org fax: 510-763-2471
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~</td>
 </tr>
 <tr>
 <td>	a) Include option for comments to be mailed to oneself or elsewhere. (Information might also be desired by someone other than the CDL or survey respondent, such as local campus library.)
	b) Do we �need� this database when we already have Lexis-Nexis? What exactly is covered in each resource for the individual newspapers? How much is covered for each title (not just years, but features and editions)? What would we miss if we kept L-N and not NNI? Are indexing features better in NNI?
	c) FYI, one unexpected occurrence concerned email results. I emailed items to myself on 3 separate days but did not notice why I didn�t receive my first 2 results right away. Somehow the results do not list the date feature that will allow it to be sorted by my Eudora mail program. What happened was that w/ �no date�, the email results got placed at the very beginning of all my messages in my IN box. I was expecting them in the most recently received area. (The date of the email is retained in the message itself.)
	d) On one help screen it states that a �...keyword search lets you match words in the articles themselves, not just controlled index terms...� [What? You can search this file full text? What full text?] On the Index help page for keywords one finds that kw=words in title, author and Gale Group-assisted index terms.
	e) Given that there are so few abstracts, searching in the abstract field seems for the most part to be a waste of time.
 
Submitted by Judy Lee on May 5, 2001. judy.lee@ucr.edu</td>
 </tr>
 <tr>
 <td>This seems easier than previous CDL journal searches</td>
 </tr>
 <tr>
 <td>LANL is NOT the preferred vendor for BIOSIS database. the interface and features are not user-friendly. Too many anomalies and spawning off to BIOSIS to get to taxonomy is unacceptable.
Taxa should be integrated in the vendor product.
How do they get any work done at LANL using this product?
I can't believe this vendor product is a "contender".</td>
 </tr>
 <tr>
 <td>Too bad Silver Platter didn't make it to the "finals"; the SP product was comparable to Ovid.
Both SP and Ovid BIOSIs products are superior to LANL's product.</td>
 </tr>
 <tr>
 <td>I liked the various colors which are used to delineate between searches in the search history. The search icons once I found them at the top of the screen. They aren't the typical icons one is accustomed to seeing. They are on the cutesy side, but they'll do. </td>
 </tr>
 <tr>
 <td>I hope that CDL will stay with SilverPlatter for the MLA Bibliography! </td>
 </tr>
 <tr>
 <td>hands-down favorite of the three MLA options.</td>
 </tr>
 <tr>
 <td>It was quite difficult to gain access to this database. I tried several times on different days before I was successful. ("Page currently unavailable") Is this normal operating conditions for this vendor? Otherwise, this is my second choice among the 3 MLA vendors on trial. </td>
 </tr>
 <tr>
 <td>I had difficulty getting access to this database ("page currently unabailable"). Tried on several occasions over two days before getting in.</td>
 </tr>
 <tr>
 <td>I can't believe CDL is dropping biosis and current contents.
Is this the beginning of the end for UC's reign as the top public
university system in the country?</td>
 </tr>
 <tr>
 <td>Again, thanks for this additional opportunity to comment! Hope you can
find my comments re PsycInfo and add them to this survey. They were
pretty positive, but provided more detail. Basically I said that at first I 
went for the interface I most familiar with--CSA. But when I went back, 
I liked OCLC a *lot* better. 
Esther Grassian
estherg@library.ucla.edu</td>
 </tr>
 <tr>
 <td>These comments apply to PsycInfo via Ovid as well. Thanks again for
asking. 
Esther Grassian
estherg@library.ucla.edu</td>
 </tr>
 <tr>
 <td>I would enjoy a link from the search results abstract form to availability at UCB libraries.</td>
 </tr>
 <tr>
 <td>This database does not contain much for my research interests.</td>
 </tr>
 <tr>
 <td>thanks for asking us for our opinion! dawn@ucsd.edu</td>
 </tr>
 <tr>
 <td>Of the three vesions of CC offered, this is the best, but still not as good as what we currently use!!</td>
 </tr>
 <tr>
 <td>Emailing -- you can be viewing in a format with graphics or images
and then email and you naturally expect you'll receive that version
but I rec'd just the text -- even though I was looking at graphics.
I might need to test further or read online help but an an
initial observation this didn't match my expectations</td>
 </tr>
 <tr>
 <td>Much better than OVID or ISI interfaces. Those two have
pathetic advanced search screens.</td>
 </tr>
 <tr>
 <td>We don't really need it since we have NewsBank and Lexis-Nexis</td>
 </tr>
 <tr>
 <td>Help is complete but dense. Would lose a freshman with no instruction. Links in help good.
With initial search screen, the limit with words too far down. </td>
 </tr>
 <tr>
 <td>#1 - I chose to search MLA and Current Contents at the same time. I like this feature, to be able to search more than one database at a time.

#2 - Only "somewhat easy to use" because it's complex.  It's not difficult!

#3 - When I know my exact author, I'd rather not be presented with the other possible choices.

#4 - Searching by subject doesn't seem to be an option!  If it's the "sh" command, it's very confusing.

#5 - E-mailing my search is a piece of cake, good directions. I like the feature of saving my search history.  The format of the citations in e-mail, however, leaves a lot to be desired. From the user's perspective, especially, a regular citation format.

#6 - There's so much to read through in the help features. One would have to have patience, but I'm thinking that once learned, the advanced feaures will be quite nice. It will be difficult in some places to teach new users.</td>
 </tr>
 <tr>
 <td>OK, overall, I like this one alot. I say that here because below are some fairly negative 
comments intended to give details of my assessment , and to point out areas needing 
improvement. I am particularly concerned about the last thing I typed into this comments field:
how punctuation effecs search results. I am very worried about this problem.
Thanks, Dawn Anderson, UCI andersod@lib.uci.edu

PS I would have really liked it if this form could have included the ability to copy myself when sending this in. so I could have a copy for my records.

Hopefully the first screen reached by user will not be the one I reached 
testing MLA OCLC.  I would want the user to arrive at the MLA database itself, 
rather than having to orient to a new page which has no mention of MLA anywhere,
then the user has to figure out how to choose MLA.

BASIC Search:

Once I searched, it was nice to have the link to full text (ECO title) (although why is it called "remote full text"?).  
But, when I got to the full text, and then wanted to go back to results, I clicked on "Back"
in my browser, but the page was "sticky", and would not return to results.  I had
to  click on the "Return" icon in the upper right.  I see this as an issue that will commonly 
annoy users.

Also, I am concerned about how libary holdings will appear in relation to 
the full text links.  If holdings info is not available next to those full text links, the user won't have 
the complete information, and may assume the article cited is not owned.  Users may blow by citations without full text links, even though
we own the titles or have full text access through other providers (say JSTOR).  Holdings info is evident in full display in close proximity to full text link which is good.  
This is a concern I have about all of the databases.

I like that there are context sensitive instrucations on the screen to assist users.

I like that authors are clickable from detailed records to find other works by them.

I don't know if this feature didn't work because it's only in demo mode, but the "search the catalog at University
of California, Irvine" feature didn't work for me.  Error: "The requested URL /WebZ/tp://antpac.lib.uci.edu/search/i 
was not found on this server."

I wish the descriptors were clickable from individual citations to enable quick searches using those descriptors.

When I marked two records in the results list, and then  clicked on "E-Mail", the default setting at the enxt screen was to send all records on the
screen I just came from (NOT just the marked records).  I would change the default to e-mail marked records.  USers will overlook
the default and receive totally different records than the ones they worked
hard to select and mark.

When I clicked to e-mail while looking at one detailed record, I was confused about the difference
between two of the three options: one option is to send "Record 4 - Dismemberment of Things Pas..."
and another option is to send "Holdings Record 4 - Dismemberment of Things Pas.."  What the heck is 
the difference between sending the record and sending the holdings record?  Is this for when (I hope) we have holdings linking?
If so, the labels need to be changed to make it clear to the user what the options are, for example, "Record citation only", and "record citation with holdings info" or something like that.

I liked the search limiting feature a lot.  The "+" and "-" symbols may trip up less expert users, but you can click on teh category names, so it's not a problem.
When I clicked on "limit by language" and also when I clicked on "Publication type", I was instead offered the years to limit by.  Assume this is just a demo glitch.

Did not like info under "finding related items" in Help.  I thought this would explain about descriptors, but instead it points users to other databases.
Clicking on "Expanding Search" in Help didn't work when I tried it.

When I tried to use the "Expand search" option by clicking the button at the top of the results list, it only offered to find other articles by the same author.
This is too limted to be very useful.  Hoped to be able to use descriptors from records in results list here.

When I clicked on "Print" from my results list, having marked two records, the database simply displayed ALL records from the search in BRIEF format, with no obvious option to print only marked records, or to change formats (for example to detailed).
If I used the navigation bar to go to "Marked records", and THEN clicked on Print, I received a screen with just those records for printing.  However, again there was no option to get the detailed display for printing.
When I chose "Detailed Display" for my results, and then clicked on print, it only displayed one record at a time in detailed display format.   Can I only then print one detailed format record at a time?  This is tedious, and would leave alot of white space per printed page (not efficient).

------------------------
USING Advanced Search

Like the menu overall.
A couple of issues:
One problem is that the "browse" feature, which is powerful and extremely useful for descriptor searching (less so for others), is 
effectively hidden under that tiny blue button with the stacked hash lines (at right of each field).  The thesaurus searching ability is therefore hard to find.
Another issue is that the nicely conspicuous and useful limit features aren't as useful as they could be because
1) user cannot select more than one publication type ( I like to select all but dissertations for example)
2) user cannot choose a specific language that is not English

This is true in expert mode as well, unless one learns the command language and hand keys it.
 
******
Subject searching

BIG PROBLEM:
Puncuation in names
The database requires EXACT matching including punctuation.  When searching for an author as subject who has a hyphen in their name, 
if you do not include the hyphen, the search is botched!!!  And even if you do include the hyphen the search can be botched because of inconsistencies in the use of the puncutation in the records.
Example:
Subject: Valle-Inclan [with hyphen]:  590 hits
Subject: Valle Inclan [wihout hyphen]: 1 hit

Example: 
Subject phrase:  Lyotard, Jean-Francois [hyphen]: 25 hits
Subject phrase: Lyotard, Jean Francois [no hyphen]:  27 hits
The records in these two searches do not overlap (at least by looking at ten records in each set).

This is a BIG problem.
This is the opposite manifestation of the same major problem in our current SilverPlatter product (in 
which if you DO include the hyphen the search is botched).
Being able to browse indexes (if you think to use it and can find it) is not enough to overcome this.

</td>
 </tr>
 <tr>
 <td>is the downloadable to endnote?</td>
 </tr>
 <tr>
 <td>I really liked the interface. Very easy to use, quick, simple, and user friendly.
However, is it still possible to use it from home? </td>
 </tr>
 <tr>
 <td>The print looks somewhat larger than the current database. Not quite up to visually handicapped, but improving.
Dislike the initial page where you have to try to decide what subjects to search. You have to guess where your database might be. Then the next screen which allows cross database searching, hardly useful for PAIS and PsychInfo.
"View record" is small print, but at least in the window. What about View FULL Record?
Help: Advanced example uses "Truncate" But help uses "Wildcard" . Inconsistent
Good internal links for help, author, descriptor. (At least citation includes descriptors, unlike Gale).
No links to our holdings. But links to free titles, ESPECIALLY GOVERNMENT PUBLICATIONS.
Can truncate more than one word unlike OCLC.</td>
 </tr>
 <tr>
 <td>Wow, I have a feeling that this is a pretty powerful search interface, but I was totally overwhelmed when I started using it. There are so many buttons and options it really makes it difficult to use the interface.</td>
 </tr>
 <tr>
 <td>Being able to move easily to a request function is very important to me. 
So is being able to mail a whole list.</td>
 </tr>
 <tr>
 <td>I found OVID much more difficult to use than OCLC, but I like some of its features very much. There was a greater chance of confusion throughout (than in OCLC) but much more to explore.</td>
 </tr>
 <tr>
 <td>links to full-text appeared inconsistent - ie, didn't always work even if
other articles in same journal did!</td>
 </tr>
 <tr>
 <td>"modify last search" brought me to a blank search screen - not what I would have expected.
links to holdings and to full text of IEEE conference papers are both absolutely critical!!!</td>
 </tr>
 <tr>
 <td>There is a lot more functionality and navigation than in the
current A&I database protocols.

IT IS CRUCIAL TO HAVE CDL REQUEST LINKED TO THE
NEW A&I DATABASE FORMATS. I UNDERSTAND THERE ARE
A LOT OF FACTORS INVOLVED IN THIS, SOME STILL PENDING FIRNAL
DECISIONS. HOWEVER, ACCESS TO REQUEST THROUGH THE
CURRENT A & I DATABASE STRUCTURE MUST NOT GO AWAY
BEFORE THE NEW ONE IS HOOKED UP.</td>
 </tr>
 <tr>
 <td>i think it would be a real mistake to select this product. </td>
 </tr>
</table>
<hr>

<hr>
Generated: 6/5/01 1:13:40 PM

</BODY>
</HTML>