The Dordt College Library has an excellent set of web pages called Searchpath which give a good overview of the types of topics addressed on this web page in great detail. If you have not seen Dordt's Searchpath you should read that before proceeding.
This page is here to help my students find literature and decide what sources to rely on when they are writing papers for me. Below I describe two ways of classifying sources. Then I discuss how to search the literature. Finally I list and classify a number of periodical sources that I want my students to be aware of.
[top]A source can be classified according to how close the source is to original information.
In Engineering, a primary source is the first publication of a theory, observation, or other idea. Most articles in professional journals, official government reports, data in data manuals, and published standards are primary sources. (For example, all the IEEE Transactions are primary sources.) A primary source does not have to be in print. It could be a film, a speech, or a web page. In any case, it is the original or first publication of the idea.
The bibliography of a primary source often gives a rather complete history of the background to the idea. This in itself can be valuable.
Secondary sources give wider distribution to the ideas contained in primary sources. These may be explanations aimed at a non-technical audience, interviews, reviews or digests of primary sources, tutorials, etc. Most textbooks, articles in newspapers, and magazines (Circuit Cellar, Newsweek, etc.) are secondary sources when it comes to engineering information. The Proceedings of the IEEE is an example of a secondary source. It is a journal of tutorials and historical reviews on topics of recent interest in the Institute of Electrical and Electronics Engineers (IEEE).
Secondary sources can be good places to look for ideas for a thesis. These sources often have greater breadth of coverage than any individual primary source. The bibliography of a secondary source can lead you to other related primary and secondary sources that may be of value to you.
Tertiary or other sources, are sources such as student papers, and book reviews that do not present original ideas and do not make reference to primary sources. (Notice how the definition is cast in the negative--" . . .do not make reference to primary sources.")
Many high-school papers are themselves tertiary sources. Lately some classes require students to post papers on the web. Yahoo, Alta-Vista and so forth will find them (and probably not tell you the author is a student!). Your paper will be more authoritative if you rely more heavily on primary and secondary sources rather than tertiary sources. You should not neglect any primary sources you run across, and when using a secondary source, look into the primary sources it refers to if you can.The main point is that the further you are from a primary source, the more one can question the veracity of the information.
[top]Literature that has been reviewed by members of a professional society, government organization, or other accepted authority is said to be peer reviewed. Peer review is a quality control process for published literature. A hallmark of peer review is the care usually taken to avoid bias and act neutrally. However, the neutrality of the reviewer usually may be critically appraised. For example, if the peer review is provided by a commercial enterprise or public interest group there may be rather obvious reasons why the organization would publish a certain author's work. Even in highly technical situations, for example an IEEE publication on a new coding method for data compression, the interests of the professional society filter what gets published and that represents a non-neutral act. In spite of any such critical appraisal of the review process, peer review assures that the work published is at least representative of a recognized segment of society.
Publications by commercial companies for profit and for the benefit of practitioners in the field are called trade literature. This literature usually assumes the reader has a certain prerequisite amount of knowledge about the discipline. If the literature were reviewed by an independent body it would become peer reviewed literature, but instead it has been reviewed by an editor or a group who has a profitable interest in the publication of the work. Usually the demands of the professional audience enforce a certain quality in the editing of these periodicals, but the quality does vary between publications. IEEE Spectrum, Circuit Cellar, and Power Engineering are examples of trade publications.
Popular publications are those written for a general audience or for hobbyists. Most newspapers and magazines sold on news stands fit this category. Some examples are The Wall Street Journal, Time Magazine, Scientific American, and Poptronics. The editors of these publications are the primary insurance of quality since the audience is assumed to be non-specialized. Some editors do a better job than others. (Or-- some editors do a rotten job.) In the field of electronics, one must be careful with these magazines. There are some articles with many errors in the details of the schematics, parts numbers, and so forth. Standards are sometimes compromised or skirted in order to simplify the circuitry, often without informing the reader of the short-cuts made. Sometimes explanations of how a circuit works are based on simplified theories--theories not suitable for a college-level paper. Some editors have even gone so far as to publish construction projects that have never been built, simulated, or tested! On the other hand, these magazines can stimulate your imagination and lead you to good project ideas. Besides which, they are fun to read.
Sometimes it is hard to tell if a periodical is a trade or a popular publication. For example Byte magazine started as a popular publication for hobbyists, but evolved into a trade magazine. It is hard to say just when the change occurred. The issues from the mid 1980's through the mid 1990's could be classified either way. You could also make a case that Circuit Cellar is intended for hobbyists, not engineers, and thus is a popular rather than a trade publication. The exact category is not so important. A good assessment of the quality of the editing is more valuable.
Individual publications are those written by one person or a small organization. The distinctive feature is that they have little or no editorial oversight. Anything goes. Individual publications may or may not be addressed to a general audience. In particular, the intended audience of many individually published web pages is intended to be a specific group--maybe not you. For example this web page is intended for students of Professor De Boer. I recommend that if you use such a source, you also provide a parallel and more authoritative source, or justify the authority of the author(s) for your situation.
This web page itself is an example of an individual publication--tertiary no less. (There are no original ideas here, and I'm not drawing on a primary source.) If you are not my student, you are advised that your professor might not agree with the ideas I present here.
However the source is classified, you must assess the reliability of the source. It is possible that an individually published tertiary source (on the web for example--like this page!) is very valuable to you. Maybe the author is quite authoritative on the subject. It is also possible for a peer reviewed primary source to be full of wrong or obsolete ideas. However, these are not the most likely scenarios. In unusual cases it would be well for you to point out the authority of your author or, in the opposite type of situation, to show by means of other sources why an article in a peer reviewed and otherwise trustworthy source is wrong.
The vast majority of quality literature is only available for a fee. Usually, you get what you pay for (or your college pays for). There are some exceptions. For example, government reports are paid for with taxes and are therefore usually available free on the web or for a nominal fee via the mail. Another exception is the case where the information is mostly useless unless one has certain hardware. A specification sheet for an integrated circuit is an example. In spite of the large amount of quality literature available on the web (from government sources, data sheets, etc.) it would be fair to say that most of the information available for free on the World Wide Web is of little value for the research papers I assign. This is because most of it is individually published secondary or tertiary material. Consider the web to be only one of several vehicles for the distribution of information. Do not discount the value of paid-for paperbound journals, books in the library, and microfilm, just because you cannot get to them by clicking your mouse. Keep in mind that your professors are the ones who order the literature in your library--and they do so to support the assignments they give. Click your mouse on the web--where your professors have little idea of what you will find--only after you know what is available locally.
[top]Some periodicals invite certain experts to write papers for them. If the paper is invited that constitutes an endorsement of the authority of the author(s) by the editor of the journal. Sometimes this type of endorsement can go layers deep. The journal's editor might select a guest editor (even more authoritatively, a team of guest editors) to organize a theme section or issue. Then the guest editor(s) usually write some of the papers and invite others to assist them to write more papers. If the journal is peer reviewed, then all the papers are reviewed by experts other than the authors and editors. Usually the journal will note which papers are invited. An example of a journal that publishes mostly invited papers is the Proceedings of the IEEE.
Some periodicals (especially trade and popular periodicals) hire people to write for them on a full or part-time basis. These people have titles such as Contributing Editor or Staff Writer. You can notice these by looking at the periodical's Masthead. The masthead is usually a box of small print on one of the first few pages of the periodical that describes the ownership of the periodical and lists the names of the editors and others who oversee the content of the periodical. If your article has no author shown, then it was probably written by a contributing editor, etc. That is commonly the case for short articles that are amalgamations from a variety of sources such as "New Product Announcements," or "News Digest." Sometimes contributing editors write entire feature articles. Usually they are then listed as the author(s) of the article. If an author of your article is listed in the masthead, then the author is a contributing editor or staff writer, etc. These authors usually publish exclusively and frequently in that one periodical. It is possible to assess the quality of one of these author's works in the context of the entire body of the author's works. For example, Fred Eady is a contributing editor for Circuit Cellar magazine. He usually writes on the topic of embedded microcontrollers. If I suddenly saw an article by him in another periodical, or on a very different topic (say power systems) then I would look twice to assess why this happened. If I have seen good articles by him in the past, then I am more likely to respect his writing in the future.
Other papers in periodicals are offered to the periodical by the author(s) without an invitation. The author simply has an idea to share with the world. These are called submitted papers. Some periodicals accept all submitted papers. More commonly, they accept submitted papers after a recommendation via an editor or a peer review process. Some periodicals never accept submitted papers. Sometimes the author of a submitted paper gets paid a small amount--usually the case for popular and trade publications. Conversely, sometimes the author must pay page charges (usually about $100 per page) to get a paper published--usually the case for a peer reviewed journal.
[top]I presume you have some idea of your topic. For example, suppose you have been assigned to write a research paper about the operation or regulation of an electric utility company. That is a very broad area to start searching. It would be best to review some of the periodicals dedicated to power systems engineering. The EPRI Journal is such an example. In a case like this it is more productive to thumb through about a years worth of the Journal to see what the feature articles are about. Then you can start making some theses statements to see how they work out.
Suppose you try the thesis, "Low reserves of electric power generation during the summers in the late 1990's gave politicians a clear prediction of the rolling blackouts California now suffers." Now there is some focus for your further research. Think of questions that could support or disprove your proposed thesis and see where they lead. For example, were there blackouts caused by insufficient reserves in the 1990's? (yes-- in Chicago) What did the politicians do about that? (new laws and regulations) Are the conditions in California similar to those existing in Chicago? (No--oops, the thesis needs to be changed or refined. . .) Once you have specific questions on specific topics then some of the more specific search engines can help.
Popular web search engines such as Alta-Vista, Google, and Yahoo might get you started (that is--they might), but they do not cover the vast majority of the published literature--only the static and non-password protected contents of the web. If that is where all the good information was, that would be fine, but for the papers I assign, these popular search engines will not be very productive. If you want a broad and general overview, an encyclopedia might be a better starting point. There are a number of them in Dordt's library. Some are on paper, some are in electronic form. Paper has the advantage when it comes to browsing. It simply takes too long to scroll through screens of information on a computer. On the other hand, if you have some good keywords to search for, then the electronic versions (of the encyclopedias, not Yahoo etc.) tend to be better. But woe to the student who can't think up the best keywords. Then it were far better that she or he were cast into a pile of print version Encyclopedia Britanicas for an hour! The keywords themselves come from the literature. It is not likely you will be able to just think them up cold.
Where to look then if not on the web via Google? Look at where the books and journals most likely to be helpful are indexed. In other words, first select your library. I mean that figuratively--select a set of literature--not literally something like "Dordt's Library." (Example: on the topic of the operation or regulation of an electric utility company, you could select related books found in Dordt's library and the Sioux Center Public Library, the EPRI Journal, and Power Engineering Magazine as an initial start for "your library.") Then search through that library's index(es). That will find books, video tapes, and other media on the shelves in a building and available electronically. When you find a book, browse the shelves (or call numbers) nearby because the books are filed in order by topic.
Usually it is best to look for books before looking for periodicals. Books usually cover the subject with greater breadth. They will give you ideas for subject searches and keywords, and the bibliographies in them can lead you to other sources. You are not going to read all of these books--scan them and read relevant portions. Be sure to scan the preface, table of contents and introduction of each book. They were written for people in your situation--people who need to find information and can't spend time reading everything.
To find the periodical literature, select some available periodicals. then browse through them to become familiar with the type of content typically in each periodical. Create a list of periodicals that are relevant to your topic. Then search these with some intensity. Becoming familiar with the literature is one of the reasons I assign papers. It will help you in your career or in graduate school. After graduation, periodicals will become the most frequent type of technical literature you will use.
Look at the table of contents of several issues of a particular periodical. Some are organized by themes. For example almost every issue of the Proceedings of the IEEE concentrates on a theme. The January 2001 issue was on the theme of "Multidimensional Broadband Wireless Technologies and Services." The February 2001 issue was on "Agents in Modeling and Simulation. . ." Or, in some cases there are certain themes or columns that recur periodically. For example every fourth issue of Circuit Cellar has a theme section on "Embedded Applications." If you have the paper copy in front of you, look at the spine. Often the featured themes are right there--you don't even have to take the periodical off the shelf to see the theme. If your periodical is on microfilm, look at the cover. Usually the themes are there. Understanding these patterns can more quickly direct you toward more relevant sources and away from useless ones.
If the periodical is only available to you electronically in a "full-text" database then you can't look at the spine to read the theme of the issue. Also, the cover might have been omitted from the database, and you might not have access to the publisher's masthead, or the original table of contents. Some graphics and equations may have been omitted. (That is, no whole issues are available, just the text from the articles.) Sometimes you can assemble a table of contents for an issue by limiting the search to a particular journal and a range of dates that covers only one issue. (e.g. January 1 through January 31 would pick out the January issue of a monthly). Then leave the "search for" box blank and click the search button. The list that results might be able to substitute for a table of contents.
Hopefully your browsing will turn up at least one or two articles that are related to your topic. Look through those articles for keywords, subjects, and related subjects. Then do subject or keyword searches, but limit the search just to that journal or a few journals. These more focused searches will still probably bring up some (maybe 10?) irrelevant references for every good one, but searching the whole world at once will surely bury the good sources like a needle in a haystack. Is your search now showing the most likely 20 of 384,721 possible matches? Forget it! Do a more restricted search.
An effective literature search must employ several techniques to find sources. Web searches alone (like Alta-Vista and Yahoo) are not enough. Electronic searches through on-line databases alone are not enough. Browsing alone is not enough. You have to read some of the material as you find it too. That way you discover additional keywords and subjects. If you have found a valuable article or book, do some author searches (for books, articles in periodicals, and web sites) to see what else that person has published. Find out the subjects under which those articles are classified and do subject searches to find out what that person's colleagues have published. Each search method generates grist for each other method.
[top]Every index has certain strengths. Below I have listed the indices which will be most productive for the types of papers I assign.
IEEE publishes thirty percent of the world's technical literature in electrical engineering, computer science and electronics and the top-cited journals in electrical engineering, telecommunications, computer theory and methods. IEEE Xplore is the IEEE's full-text access to IEEE transactions, journals, magazines and conference proceedings published since 1988 plus select content back to 1950, and all current IEEE Standards. (copied from IEEE web pages.)
IEEE Xplore is important since it indexes most of the top journals and some of the top magazines in the field of electrical engineering. Recently it added the IET's publications to the database. (The IET is the British counterpart to the IEEE).
FREE TO ALL: Ability to browse tables of contents of all IEEE transactions, journals, magazines, conference proceedings and standards. (All the links in the left-hand frame above the word "Search" work for all users, members or not.)
IEEE STUDENT MEMBERS: You can search and access all IEEE abstract/citation records from all IEEE journals, conference proceedings and standards. You may also access your full-text online personal subscriptions (known as OPeRA) here. You will need an IEEE web account, which is free to student members.
If you are a Dordt College student you may become a student member rather economically. Before completing the membership application it would be best to talk to a local student member of the IEEE Dordt College Student Branch to get advise on how to maximize the benefits you subscribe to at minimum cost (possibly zero cost after rebates).
IEEE Xplore is available at http://ieeexplore.ieee.org.
Our library's primary indexing service for periodical literature is Ebscohost's "Academic Search Premier" database, available from the Dordt library's web page. Ebscohost is an excellent service for our campus as a whole. You should be aware however that most of the prestigious engineering periodicals are not indexed in Ebscohost. (The part of Ebscohost subscribed to by Dordt that is.) On the other hand there are literally hundreds of trade and popular periodicals with engineering content indexed in Ebscohost. It is strong in breadth.
Besides indexing, abstracts and full text are often also available from Ebscohost. Ebscohost's indexing and abstracting service often only goes back to about 1996 (for the periodicals listed below). You can check in Ebscohost to see how far back each periodical is indexed and abstracted. The Ebscohost database will also tell you where you can find the article. If the "Note: This title IS available in the Dordt Library" line shows, then you will find the article on the shelves either in the "Current Journals" area, or in the "Bound Journals" area. Ask a librarian if you cannot find these areas. If the "full- text" icon shows, then you can get the full text from the Ebscohost database.
In Ebscohost's "Academic Search Premier" Database, you can get information about a journal if you know a few words in the title. On the top menu bar, click on "Publications." Then enter a key word and click on the "Match Any Words" button before clicking "browse" to start the search. For example, let's find out if the magazine "Electrical World" is peer reviewed. Enter "Electrical World" in the box, click the "Match Any Words" button, click "Browse," and Electrical World pops up, in this case right on top of the list. Click on the title and you get SOME of the information about the publication. Then click on "Publication Details" and you get all the information. This one is not peer reviewed. Try again with "IEEE Power" as key words in the publication title, then click on "IEEE Transactions on Power Delivery," etc. and you will see that this one is peer reviewed. You can also find out if the publisher has a web site. Sometimes the web sites contain supplemental articles.If you need an older article (not in the databases) for a paper I've assigned, talk to me or a librarian about it and we will look into other ways to find it besides those described here.
[top]Below I have listed periodicals I am aware of that are accessible to my students and that may be relevant to papers I assign. Of course the periodicals mentioned below are not the "vast majority of the literature." Instead, I expect they will be a productive and available subset of that "vast majority." If you have personal subscriptions or access to another library, then you may find still more sources. If nothing much is stated about a particular publication, it is available either via Ebscohost or in the Dordt College Library. Most of these are indexed in the Applied Science and Technology index or in Ebscohost. The titles listed here are listed because Professor De Boer endorses them.
You do not have to restrict your research to these endorsed periodicals. It is just that Professor De Boer thinks you should at least be aware of the publications below to see if any of them fit your area of interest.
Official government reports are peer reviewed literature. An official government report will always include information in the front pages which explains the origin of the report, how to get additional copies of the report, and any contracts that relate to the report. In order to be an official government report (and thus peer reviewed) there must be a claim of government sponsorship. Look for something like, "This report was prepared as an account of work sponsored by an agency of the United States Government." (In contrast, ordinary web pages are not official reports-- therefore not peer reviewed--even though the URL ends with ".gov".)
If you know you are after a government document but do not know how to search for it, you can start with the Dordt college library's web page of Online Reference Links, the Governmental Resources section. Note that FirstGov.com is a good place to start looking for federal government documents. For example, if you are looking for Bush's National Energy Policy, a search for "National Energy Policy" on FirstGov.com brings up several copies of the the document including one hosted by the whitehouse.
If you still can't find what you are looking for on your own, ask for help.
Does one of your sources make some claim against a standard? E.g. "USB allows hot-plugging" whatever that is! Then go get the standard and find out. Go directly to the primary source if you can. Our library has some standards on the shelf. Use Dordt's Webcat to find these. (USB is one standard we have.) Some more are available via ftp or the web. A Google or Yahoo type search might find those or give you a lead to where to find it. If you know who sponsors the standard, check their web site.
Technically, a "draft standard" is not (yet) a peer reviewed document. Professor De Boer might accept it as peer reviewed anyway--talk to him about it and see if he approves.