| Criteria |
Ovid |
ProQuest |
| 1. Software Design: |
|
|
N/A |
N/A |
|
|
Good.
4 seconds for all formats
25 seconds to retrieve full-text+ graphics BUT the gain in speed is
lost because the images are thumbnails. User must tell system to re-display
with full size figures and tables. |
3 seconds for all other formats including full-text
+graphics
35 seconds to retrieve full image first time 10 seconds lost to Netscape
opening Adobe.
25 seconds second time (Acrobat was already up.) |
|
|
Nice layout. Logical progression through screens. |
Good novice end-user layout. Really nice search
assistant. |
|
|
Yes. |
Yes. |
|
|
No for Web-based.
Yes for Java-based.
This is a case where a remote user using Netscape on a Macintosh
would be unable to access Java version until they installed Microsoft
Explorer.
|
No. |
|
|
Yes. |
Yes. |
|
|
No. Just Basic vs. Advanced search screens. |
Not really. Options for Basic (default),
Advanced, and Search Assistant |
|
|
Yes. |
No. |
|
|
Don't know |
Don't know |
| 2. Searching Capabilities: |
|
|
| Abstract (ab) |
| Outline Headings (oh) |
| Accession Number (an) |
| Pagination (pg) |
| Authors (au) |
| Publication Date (pd) |
| Caption Text (ct) |
| Publication Type (pt) |
| Document Type (dt) |
| Publication Year (yr) |
| Full Text (tx) |
| References (rf) |
| ISSN (is) |
| Text Word (tw) |
| Institution (in) |
| Title (ti) |
| Issue/Part (ip) |
| Update Code (up) |
| Journal Name (jn) |
| Volume (vo) |
| NLM Journal Code (jc) |
|
| Abstract |
| All Indexing |
| Article Column Head |
| Article Title |
| Article Type |
| Author |
| Author Affiliation |
| Caption |
| Classification Code |
| Classification Name |
| Company Name |
| Date (Alpha) |
| Date (Numeric) |
| Degree |
| Dissertation Advisor |
| DUNS® |
| Footnote |
| Geographical Name |
| Headnote |
| ISBN |
| ISSN |
| Product Name |
| School Code |
| SIC Codes |
| Source |
| Source Database ID |
| Source Type |
| Subject Terms or Code |
| Ticker Symbol |
|
|
|
Basic Screen:
- Keyword/Author search
- "Limit to" choices -
- Latest Update
- Original Articles
might not be the best for novice end-users to have to decide between.
I have been unable to find OVID's definition of "original article"
Advanced is the default right now.
|
Basic Screen:
"Search by word" searches the following indexes:
- Abstract, Article Title, Author, Company, Geographical Name,
Personal Name, Product, Publication Name (Source), and Subject.
- Users have the option to "Search the full text of articles"
if checked off ProQuest searches title, abstract, and complete
text of each article.
- Basic is the default with the "search full text option" turned
off.
- Advanced has four search fields with drop menus with the boolean
connectors. Article title, author, company, geographical personal
name, Pub title, subject or text are the searchable field options
in advanced mode.
- Although all the fields in item 2.1 are searchable, the user
must know the code and apply it on the search line.
|
|
|
$ (unlimited)
: (unlimited)
$1 (limited truncator)
# (one character at end)
? (one embedded chrct.)
|
? (unlimited)
* (one character, either embedded or at end.)
|
|
|
Uncluttered. Clean. Simple icons. Good background
color. |
Uncluttered. Clean. Simple icons. Good use of color.
Buttons are in the same place on every screen. That's an plus. |
|
|
Yes. |
Yes. With an option of deleting it from the search
history list. |
|
|
Available on screen as checkboxes. |
No. |
|
|
Either by display arrangements, (i.e. sort) or by
"combine" This feature that allows you to take search #1 and #5 and
pick your boolean.
Additional after searching the user can select Limit and select
an earlier search to apply a limit to, not just the last search
executed.
The user can limit by name of journal, publication type, date,
Latest Update Abstracts
Original Articles, & Review Articles. |
User can click on "Show only articles available
online in full text." to narrow after searching. I like this because
users can see how many articles they lose by limiting in the fashion.
|
|
|
The browser back key is probably the way that the
users will choose to undo their prior action. |
The browser back key is probably the way that the
users will choose to undo their prior action. |
|
|
No. No subject headings displayed to have hypertexted!
The complete list of subject headings is available through the Help.
They are not hot-linked, so users can either cut and paste them to
the command line or they can remember them and type them in back at
the search screen. |
There are subject headings displayed but they are
not hot-linked.
The help mentions subject headings and states that it is a controlled
vocabulary list but there is no access to the list online.
|
|
|
The back button on the browser. And the Combine
feature would allow you to go back and combine older searches. |
The back button on the browser. There is no way
to combine older searches. Users can rerun them but not combine them. |
|
|
Yes. All search fields have browsable indexes. |
No. |
| 3. Help/Guidance: |
|
|
Yes. Some jargon like "Original Articles." |
Yes. No jargon. Typical ProQuest user-friendly screen. |
|
|
Yes. But not database specific! Ex: When searching
in Biomedical Collection IV, I asked for help with limiting. The limit
screen displayed in the help was not the limit screen I was working
with, so I could find no explanation of what "Original Articles" means
to OVID. |
Yes. Very nice. |
|
|
"The query that Ovid received, ?"> , did not produce the desired
results."
Is a very nice error message as error messages go. At least it doesn't
assign blame and the syntax error screen has a "Return to search"
button.
"Internal Server Error" was more daunting and did not have instructions
on how to return.
|
Good explanatory error message:
- "Your search failed because it is incomplete or incorrect.
- The problem is immediately before or after the ^ character as
shown here:
- Instead of "?" you should have one of the following:
- a search field
- the name of a month or season
- a word
- a left parenthesis
- a quotated word or phrase"
An end-user should be able to correct their search statement using
this error message. |
|
|
No. The button on the screen takes you back to the
main search page, but it does keep the search history. |
No. The back button doesn't penalize the user though. |
|
|
Yes. |
Yes. |
|
|
Yes. |
Yes. |
| 4. Display/Delivery: |
|
|
No. The citations returned in "Title Display" format
is very jumbled. Even with color-coding to separate the fields visually,
the fields flow together. Especially after using databases where each
field is given its own line. |
Sort of.
The initial screen is busy. Especially with the chart on the right-hand
side that lets the user pick the display format. After this point
though it get cleaner and easier to read.
Has a link for "clean copy for printing" that removes all extraneous
links from the page. Nice. |
|
|
| TOC |
| Citation |
| Abstract |
| Complete Record |
| (an, au, ti, so, ab, issn) |
| Full-Text + graphics |
The default results will return as a citation that is color-coded.
This color-coding won't carry over to the printed copy so I suspect
the result display layout won't be that popular but there is sufficient
citation info on the initial results screen that the patron could
elect to just print that screen.
|
When user gets initial list of results, and they
click on the title, they will be taken to the full-text. They also
can select their other display options in the columns to the right.
Those options are:
- Citation
- Abstract
- Full Text
- Text+ graphics
- Page image
There is sufficient citation info on the initial results screen
that the patron could elect to just print that screen.
No table of contents. |
|
|
| Abstract (ab) |
|
LocalMessage (lm)
|
|
Accession Number (an)
|
|
NLM Journal Code (jc)
|
|
Authors (au)
|
|
Publication Type (pt)
|
|
CODEN (cd)
|
| Source (so) |
| Document Type (dt) |
| Title (ti) |
| ISSN (is) |
| Update Code (up) |
| Institution (in) |
|
Default is full text. |
|
|
- User can specify layout format - OVID or BRS/Tagged. Neither
term means anything to lay end-users.
- The download format can be DOS/PC- Mac- or Unix- ASCII.
- Users can mark individual records or take first 200 in set and
manage and download them. Nice feature.
|
- WYSIWYG.
- The Save feature of the web browser is how exporting is done.
- This is done record by record too, since you can't mark a set
to manipulate. After the user selects email citation or email
full text, the article are saved in a set in the Pending Order
Screen.
- The use of the word ORDER is misleading since this usually means
money to people. Maybe Export Queue would be a better term. This
is as close as ProQuest gets to marking.
|
|
|
Yes, users can check off records to display and
can sort (or "manage") by Author, Title Journal or Date in ascending
or descending order. This is a nice feature. |
No! As far as I was able to tell from the help,
there is no way to browse a list of citations and mark the ones you
want and then manipulate that set. |
|
|
No. Must "manage" citations/records in order to
sort. |
Yes. |
|
|
Excellent on screen. Image are "clickable"
|
Image Display Options are:
- Fax Quality (provides the fastest download)
- Print Quality
- Print Quality with Halftones (provides the best image quality
for articles with halftone graphics)
|
|
|
Any printer will work if article is full-text. Must
have laser to print graphics/images. |
Any printer will work if article is full-text. Must
have laser to print graphics/images. |
|
|
Yes but email will not include images or graphics. |
Yes but cannot use email to deliver Page Image or
Text+Graphics article formats |
|
|
Excellent on screen. Printouts will be as good as
the printer. |
Excellent on screen. Printouts will be as good as
the printer. |
|
|
Nicely done. Default is inline thumbnail, with options
to remove all graphics or to show them full sized embedded in article.
Additionally the article is not scanned so users will be able to copy
and paste from the text and know where the graphics will fall in the
text. |
- Well done. Clean images embedded in the article.
- Right next to graphic is the option to enlarge thumbnail to
50% or %100 of actual image.
- This is different than the select and then reload of OVID.
|
|
|
- No. User must return to title list before looking at next record,
regardless of display format. This is clunky.
- Down the pages: only displays 10 citations at a time.
- There is no fast way to move from record 1 and 47 for example.
A jump to feature would be nice.
|
- A small table box in the corner allows the user to go forward/back
one article, tells you which article you are on and has the "clean
copy for printing" link.
- Down the pages: displays 10 citations as a default but user
can set it to display 20 or 25 per page as well. After results
are retrieved there is the option to "View all the results" And
that will take all results in pool and put them on one web page.
VERY nice feature. Even includes a warning that it will affect
performance.
- The web standard "Select a page for more results" is at the
bottom so the user can quickly jump around the results.
|
|
|
No. Not really necessary since the whole database
is full-text. |
No. |
| Comparative Search:
(arterio? and (geriatr? or elder? or gerontolo?))
|
0 hits |
30 hits |
| aged AND alcohol? |
15 seconds to run |
25 seconds to run |
| 5. UNF Specifics |
|
|
87 |
113 approx. |
|
|
25 |
19 |