Student Support
A Guide to Using the Internet for Research
The Internet is a massive network of computers which can
offer users access to vast quantities of information and easy global
communication. There are resources available on the Internet for almost every
field of interest. This guide serves to explain what Internet resources are
useful to research and gives some pointers on how to find information relevant
to your research project.
It is convenient to classify the Internet resources into
e-mail based communication and the World
Wide Web (WWW) and this guide has been structured as such.
An additional
guide for students writing their theses has been prepared by Tom
Moultrie. It gives additional information on referencing and some pointers
on academic standards on the web.
The Internet
Research FAQ is another guide to using the Internet for research.
E-mail programs allow one to send messages to
individuals or to groups of people. Discussion groups and mailing lists have
been created where one can send mail to, and receive mail from, groups of people
with similar interests.
Mailing lists and discussion lists
Mailing lists and discussion lists (groups) are devoted
to particular subjects or fields of interest. There are a huge number different
lists available on the Internet and new lists are being created almost
continuously. A list related to your field of interest can be an excellent tool
for research. Lists allow you keep up-to-date with the most recent
announcements, developments and thoughts in a particular subject.
These lists are run by programs called listservers which
store e-mail addresses of people who are interested in a particular subject. If
a single e-mail message is sent to a particular listserver's list, then that
message will be duplicated and sent to all members of the list. Listservers are
thus an efficient means of communicating, with and between, a large group of
people with a common interest (rather than sending each one a separate message).
The difference between mailing lists and discussion
lists is that mailing lists are structured as announcement-only services whereas
a discussion list is a forum where all subscribers can participate by asking and
answering questions. With mailing lists there is usually only one person or
organisation that can post information. Mailings lists are thus mainly used for
distributing abstracts of new papers (published and/or working papers), for
calls for papers, conference announcements, job advertisements, etc. Discussions
groups on the other hand offer open discussion and debate amongst all
subscribers. They can provide a forum for asking questions that will hopefully
be answered by experts in the field. Although there are important differences,
some people use the terms mailing lists and discussion lists interchangeably.
Discussion lists can be moderated or unmoderated. All
mail sent to a moderated list must be approved by the moderator (the person in
charge of the list) before the message will actually be distributed to all the
list members. In unmoderated lists, any subscriber can post a message to the
list without it first being approved. In some cases unmoderated lists contain
many off-topic posts and suffer from a high "noise-to-signal ratio".
On the other hand they can contain a wide range of freer thought and discussion.
Which is the case depends on each individual list, its culture and the
personalities of the subscribers. Join some lists and find out...
Finding mailing and discussion lists
Sometimes it is difficult to find out what lists there
are in your field of interest. Lists are often advertised or mentioned in Web
pages relating to your subject or in Usenet newsgroups (discussed below). One of
the best ways of finding lists on a particular subject is to make use of some of
the many Lists of Lists that are available on the Internet. Most of the better
Lists of Lists are searchable and you can thus easily search for your subject in
the list descriptions. However, currently no one List of Lists offers a complete
listing of all discussion groups and thus it is advisable to use several List of
Lists to search for your interests.
Links to some Lists of Lists are given in the Virtual
Investments Library. In addition you can also find the "Wouters guide to
finding mailing lists" there.
Subscribing to lists
Once you have found suitable lists, you join them by
"subscribing". This is accomplished by sending a simple mail message
to the listserver (NOT TO THE LIST). There are 3 common listserver
programs: listserv, majordomo and listproc. An example of how to subscribe to a
list is given below:
- Name of the list: BUS422W_INVESTMENT
- Address of the list server: A majordomo@southernlife.com
To subscribe to this list you send a message to
majordomo@southernlife.com. Leave the subject blank and enter: SUBSCRIBE
BUS422W_INVESTMENT in the body. (Note: if you have any signatures or comments
that are automatically added to the bottom of your messages you must turn them
off when subscribing to lists. Alternatively you can use the END command after
your subscribe line to prevent the listserver from reading the rest of the
message and your signature).
The above process is similar for both listserv and
listproc lists. For listservs, however, you must add your first name and last
name after the list name on the subscribe line.
- e.g. SUBSCRIBE RISKNET Joe Bloggs
Soon after subscribing you should receive confirmation
of subscription from the listserver. This message will explain how to
unsubscribe from the list, how to send messages to the list, how to access the
archives of previous postings and provide other general list information.
To send mail to a list to which you have subscribed you
use the list name in the address. In the above example this would be,
bus422w_investment@southernlife.com.
Many lists keep archives of all messages that have been
posted to the list. These are either available by mail from the listserver or
are located at a Web site. Details on how to retrieve archives are normally
given in the welcoming (information) message when you join the list. When
archives are kept on the listserver (as opposed to the Web) they are normally
accessed by using GET commands sent to the listserver. You can usually get
instructions on how to retrieve files from the listserver (and on all other
listserver commands) by sending a message containing HELP in the body to the
listserver.
Finding e-mail addresses of individuals
Currently the easiest way to find out someone's e-mail
address is to ask them! Searchable directories of e-mail addresses are available
but at present these directories are far from comprehensive.
Four11 Directory
Services is the largest "white page directory" of e-mail
addresses. It contains mainly e-mail addresses of people in the US but it may
help you find someone. There is also a South
African directory of e-mail addresses but at the time of writing it
contained very few addresses. In most of these directories you can search by
country, organisation and occupation as well as by name.
In addition, if this fails, you may find a directory of
e-mail addresses at the Web page of the person's employer. Universities and
research institutions generally have detailed directories of employee's e-mail
addresses.
Usenet groups
With e-mail or a listserver, a mail message gets sent to
a specific person or group of people. Usenet, also works by means of e-mail
messages, however, it operates like a bulletin board rather than a postal
service. Messages get sent to a central bulletin board where they are seen by
anybody who looks at the board (not only the person to whom it was sent or the
people on a specific list). Answers to messages are also posted to the board. In
a newsgroup you have to actively enter the forum to read the messages
postedwhereas with a discussion list all the messages get sent directly to your
e-mail address. If you don't read Usenet news for a few days you can miss out on
messages entirely as old messages expire and are removed from the "bulletin
board" after a few days to make space for new messages.
Each newsgroup deals with a specific topic. Topics are
grouped together into major categories. Below are some of the major categories:
- alt: controversial or fringe topics
- bionet: research biology newsgroups
- bit: bitnet (a network of educational sites)
newsgroups
- biz: business related newsgroups
- comp: computer and related subjects newsgroups
- misc: well supported discussions which do not fit
elsewhere
- news: news about Usenet itself
- rec: hobbies, games and recreation newsgroups
- sci: science other than research biology
- soc: social (often ethnic) newsgroups
In addition there are country and even organisation
specific categories of newsgroups, such as:
- za: South African newsgroups
- uct: UCT newsgroups
There are now search engines that allow you to search
through all the Usenet groups and messages for topics of interest. DejaNews
currently offers the largest searchable Usenet archive. You can search by
subject, keywords, author, etc.
Reference.COM is
a useful search engine that provides access to the archives of over 16,000
newsgroups and a large number of mailing lists. Furthermore, Reference.COM
allows you to create "stored queries" that will automatically rescan
the archives at a user-specified interval and will e-mail you the results. This
stored query is an example of an "intelligent agent" that will seek
out news and information that interests you and then give you details on how to
access this information.
Reference.COM has an advanced search function where you
can interactively refine your search using message header fields such as author,
organisation, subject and newsgroup name. There are also ready-made query
templates which pre-select the right newsgroups and lists for a particular
topic. Just complete the rest of the form and submit your query.
The World Wide Web (WWW) is the most versatile tool
available on the Internet. It consists of hypertext pages that can contain text
and graphics (and even sound and video). These pages can be linked to other
pages by means of hypertext links.
Uniform resource locators (URLs)
The "address" of resources on the Internet are
specified by a URL (Uniform Resource Locator). URLs identify both the type of
resource and specific location of it on the Internet. An example of a URL is
given below:
- http://www.southernlife.com/actuary/uct/research.htm
The first part of the URL (before the colon) specifies
the access method. The part of the URL after the colon is interpreted as the
address. In this example the access method or protocol is "http". This
identifies the Hypertext Transfer Protocol (HTTP) which is the communication
protocol used for the World Wide Web. The part after the double slash in the URL
is the address of the computer (www.southernlife.com). The last part of the URL
(after the first single slash) is the name of the file at the site. In this
example the file referred to is this research guide.
Note that there are other access methods or protocols
(e.g. ftp and gopher) but most Web browsers support all the protocols and so you
should be able to access almost all resources from within your Web browser.
Finding sites on the WWW
The Web is so big that it is often difficult to find
resources in your field of interest. Good starting points are the search engines
and the various Web Directories. Virtual Libraries can be also be very useful
for finding resources of interest.
If you are looking for the Web site of a company or an
educational institution you can often be quite successful with educated guesses.
Most organisations will use a URL that contains their name in a logical way so
as to make them easy to find. When guessing URLs try http://www.company.com
or http://www.company.co.za for South African companies. Common endings
for academic sites are .edu for American universities and .ac.za for South
African educational sites. Another common ending for URLs is .org which is
normally used for official bodies or non-profit organisations.
-
Yahoo!
Yahoo! is one of
the most popular and well-known directories. You can browse through entire
sections or you can search the whole directory for your desired subject.
When you search Yahoo! it provides a list of links
matching your search criteria. It also lists the names of several other
search engines at the bottom of the page. Clicking on a link to one of these
other search engines will conduct the current search on the specified search
engine. Currently there is no one search engine that is best. They each
search different databases using different methods and so using a range of
search engines is often the best way to find good links.
Note that Yahoo! searches are case-insensitive -
i.e. Yahoo! doesn't differentiate between upper and lower case. Yahoo!
supports both the and and or Boolean operators. You never actually type the
Boolean operators in your search string - you just separate the words with
spaces. By default Yahoo! conducts an and search for all the words in your
search string. But you can specify on the search options page that Yahoo!
rather conducts an or search.
Also on the search options page you can specify
whether Yahoo! searches for your string as a substring (this is the default)
or as complete (and distinct) words. If Yahoo! is searching for substrings,
then the words must appear consecutively in the order specified in your
search string. The final setting on the search options page is the number of
matching documents to be displayed on each of the pages generated. The
default is 25 but you can specify 10, 25, 50 or 100 matches per page.
-
Alta Vista
Alta Vista is
a very powerful search engine that has indexed a massive number of Web
sites. It supports some quite complex searching techniques including Boolean
searches. This search engine's database is so large that you must be very
specific when using it otherwise you will find far too many links! Alta
Vista has a simple search and an advanced search facility. The advanced
search is actually better and is often easier to use. Alta Vista can search
both the WWW and Usenet newsgroups.
At the Alta Vista Web site there are several pages
explaining how to use the search engine. Some of the most important aspects
of the help are summarised below but this guide does not attempt to give
complete details on using Alta Vista.
When searching for words or phrases Alta Vista
ignores all punctuation. Thus U.S.A. and U S A are identical
(both 3 word phrases to Alta Vista). These two phrases are not the same as USA
(1 word). If you are searching for a specific phrase (i.e. the words must be
found together) then "enclose the phrase in double quotes".
If you specify a word in a search string using lower
case letters then Alta Vista conducts a case-insensitive search for that
word. If, however, you use an upper case letter anywhere in a word then you
force Alta Vista to find exact case matches on that entire word. Thus, internet
will find Internet, INTERNET, internet, or even iNtErNeT.
But if you use Internet, Alta Vista will only match Internet
and none of the other capitalisation variations. It is recommended that you
use all lower case letters unless you specifically know the case of the word
you want to find.
You can use Alta Vista to search for the occurrence
of any of a family several related words by using * notation. For example, invest*
will find invest, investing, investment and investments.
However, it will also find words like investigate. To use the *
notation Alta Vista requires that the * appears only at the end of a word
and that there are at least three characters before the *. Furthermore, Alta
Vista will only match strings of up to five additional lower case letters
(no capitals and no digits).
When Alta Vista displays the list of matching
documents the search engine puts the links that it thinks are most useful at
the top of the list. "Usefulness" is based on weighted scores for
pages. The exact scoring technique is quite complicated and not published
completely, but, for example, documents with titles containing the words of
the search string are scored highly. Also documents containing many words
from the search string (if an or search is used) are weighted higher than
those which contain only a few of the search string words. And documents
that contain the matching words multiple times are also scored highly.
Due to this weighted scoring, even if Alta Vista
returns a few thousand matching documents you should still find what you are
looking for in the first 30 or so links. If not, then you should refine your
search string and try again.
Alta Vista also supports various constraints in the
searching process. You can use the title constraint to find only
matching words in the title of the document. For example, title:bond
will only find pages that contain the word bond in the document
title. Other useful constraints are link (finds pages that include a
link to a certain site) and url (finds pages whose URL include the
matched words). Note that if you wish to search for multiple words in a
title you must either use quotes (e.g. title:"bond rates")
or proceed each word with title:.
Simple Query: In this search Boolean
operators like and,, or and not are ignored in search
strings. Alta Vista will search for Web pages that contain any of the words
in the search string. Those pages that contain all or most of the words in
the search string are displayed near the top of the list of matching pages.
Below are some example search strings for the simple search.
bond rates
This will find all documents containing the words bond or rates
or both. Pages containing both words will appear towards the top of the
list. Note that in this search the words do not necessarily appear next to
or even close each other in the document.
bond +rates
This will force Alta Vista to return only matches that contain at least the
word rates and here again pages with both words will be at the top of
the listed. This search will not find any matches which contain bond
but do not contain rates (while the first search would have found
pages with bond only). [Note there is a space between the word bond
and the +sign but there is no space between the + sign and the word rates.
bond +rates -gilt
This is similar to the second search but it will exclude all documents that
contain the word gilt.
"bond rates"
This will find only those documents that contain the exact phrase bond
rates. There can be no words separating the search phrase on the
matching documents. This search would probably be most useful in finding
sites which provide bond rates.
Advanced Query: In the advanced searches you must
use at least one of the operators and, or, not and near
to combine words and phrases in search strings. Parentheses are used to
control nesting of operators. The operators can be specified in upper or
lower case. The results of the search can be optionally ordered by
specifying (in the Results Ranking Criteria field) a set of words or phrases
which then receive higher weighting. Some examples of advanced searches are
given below.
bond or rates
This finds documents contain at least one of these words (documents with
both words are also returned and these are near the top of the list).
bond and rates
In this search both words must be present in the matching documents.
(bond or gilt) and rates
This will find documents containing bond rates or gilt rates
(or both). Note that parentheses are used to control the precedence of the
Boolean operators.
bond and not gilt
This will find all documents containing bond that do not contain gilt.
Note that to use the not operator you must proceed it with the and
operator.
bond near rates
This will find documents where the words bond and rates appear
within ten words of each other. The near operator is particularly
useful when searching for someone's name. A name may be in the form of Joe
Soap, Joe A. Soap, Joe Andrew Soap or Soap, Joe and the near operator
will find any of these orderings.
Virtual Libraries
Virtual Libraries can also be very useful resources for
conducting research. On the Internet you will find both general reference
libraries and subject-specific Virtual Libraries. The general libraries contain
a broad range of subjects and information. There are links to some good general
libraries in the Virtual Investments Library. Subject-specific libraries
obviously have much more detail on the topic than general libraries. A
subject-specific Virtual Library relevant to your research topic should be an
excellent resource. However, currently there are many subjects that do not have
any dedicated online libraries.
The Virtual Investments Library is an example of a
subject-specific library. In this case the library is in fact specific to the
needs of students studying an investment and asset management course at UCT.
This Virtual Investments Library is structured as a list of links to investment
related sites on the Web. It links mainly to on-line magazines and newspapers as
well as financial information sites. Some of the other Virtual Libraries contain
complete written works or at least abstracts of published material.
Electronic Journals and Magazines
Many printed magazines, journals and newspapers (and TV
stations) are now publishing on-line versions. The on-line equivalents are not
generally the complete publications. Some sites do publish complete on-line
versions but require you to subscribe to the service. There are a few sites
(e.g. Financial Mail) that do offer complete
on-line versions that are available at no charge.
Due to the ease of publishing on the Internet there are
now many electronic journals and magazines that have no printed equivalents.
These are called e-journals and e-zines. Some e-journals have already
established a name for themselves as acclaimed academic journals. Many
e-journals and e-zines are distributed in mailing list form and can usually be
found in the same way as you would find a mailing or discussion list (process
outlined above). Alternatively you can use one of the lists of electronic
journals and magazines. There are links to some of these lists in the Virtual
Investments Library.
Referencing Material from the Internet
Referencing information obtained from the Internet is
slightly different to referencing printed material and currently no official
referencing standards exist. One of the problems with Internet material is that
there is no guarantee that a Web source you used will still be available in the
future at the same location or even anywhere. Past Usenet news messages and
mailing list messages may be stored in an archive but not all lists or groups
are archived (yet).
It has been suggested that references include the date
that the information (e.g. Web page) was accessed. Most referencing styles
proposed for Internet resources require the following information (some of which
may not always be available): the author, the date published or written, the
URL, the (most recent?) date accessed by the researcher, the name of the
information supplier (on-line publisher) and the medium. In the absence of
standard referencing styles, it is best that you record at least all this
information for all Internet sources that you use. Below are some examples of
how Internet references can be presented:
Individual works:
Author/editor. (Year). Title (edition), [Type of medium]. Producer
(optional). Available protocol (e.g. HTTP): Site/Path/File [Access date].
Example:
Pritzker, T. J. (No date). An early Fragment from Central Nepal,
[Online]. Available HTTP: http://www.ingress.com/~astanart/pritzker/pritzker.html
[1995, June 8].
Write "No date" when the electronic
publication date is not available.
Parts of works:
Author/editor. (Year). Title. In Source (edition), [Type of
medium]. Producer (optional). Available: Supplier/Database identifier or number
[Access date].
Discussion list messages:
Author. (Year, Month Day). Subject of message. Discussion List, [Type of
medium]. Available E-mail: DISCUSSION_LIST@e-mail_address [Access date].
Example:
RRECOME. (1995, April 1). Top ten rules of film criticism. Discussion on All
Forms of Cinema, [Online]. Available E-mail: CINEMA-L@america.edu [1995,
April 1].
Author's login name, in uppercase, is given as the first
element. If the author's full name is known (i.e. specified in the message) then
this can be used. The general consensus is, however, not to give the
author's full e-mail address.
For more details on Internet referencing see the
following sites:
Tom Moultrie
has prepared an additional guide for students writing their
theses. It gives
additional information on referencing and some pointers on academic standards on
the web.
The Internet
Research FAQ is another guide to using the Internet for research.
This research guide was prepared by Steven Levin - Last
updated 16/07/96
Any suggestions or comments about this guide are gladly
accepted. Did you find it useful? Do you know of any better Internet research
guides? What should be added?
| 
