A complete literature search can and should incorporate several resources. Resources for a literature search include Internet search engines, databases, and library catalogs.

Internet Search Engines

• Include Google, many others. 
• Cover all subjects
• Are free
• Retrieve many results
• Provide no quality control
• Include some full-text literature

Databases

• Include TRID, many others
• Provide a level of quality control to their content
• Cover information a specific area
• Are sometimes subscription-based. The library has full-text access to hundreds of databases for students and faculty, including many on transportation information.

Library Catalogs

• The Transportation Library's catalog searches books and articles on topics related to transportation.
• NUsearch, the main Northwestern University library catalog, searches all resources at Northwestern.
• WorldCat searches the collections of more than 10,000 libraries around the world.

After topics are defined and well understood, and the resources to search are chosen, it is time to select search terms. To choose search terms, look again at the search topic that was developed in Step 1.

• List words that describe the topic first. \
• Think of synonyms, plurals, and different word endings (e.g., climate and climatic) for these words.
• Consider technical, local, and international terminology as well as acronyms and abbreviations that are related to these words.
• Take spelling variations into account (e.g., behavior and behaviour).
• Check on the availability of a thesaurus, subject headings, and index terms from the search resource and look up related terms. 

The resulting list will be the initial search term list. The list of search terms may very likely expand or change as the search progresses.

Defining the relationships between search terms and combining them are critical steps in the search process. This is called developing the search strategy. How terms are linked in a search strategy can significantly affect the outcome of a search. In order to create an effective and efficient search, it is worthwhile to put some time into developing a search strategy instead of simply adding any potential words into a single search field. 

Search Tips

Controlled Vocabulary vs Keyword Searching

• Controlled vocabulary may also be referred to as subject headings, descriptors, thesaurus, or index term. Advantages to controlled vocabulary include:
  • A list of subject terms may help a user find an appropriate search term for their topic.
  • It can provide a searcher with suggested terms for narrower, broader, or related topics. Using index terms helps the searcher avoid the need to think of every possible synonym or alternate spelling of their search terms.
• Keyword searching: any search terms chosen by the searcher
  • Unlike controlled vocabulary searches, this will include jargon and new terms that haven't yet been added to a controlled vocabulary.

Exact Phrase Searching

• Most databases and search engines, including Google, support exact phrase searching using quotation marks to retrieve results in the exact order that phrases are typed.
  • Example: "pedestrian bridge" will only retrieve items that contain the phrase pedestrian bridge. Items such as pedestrian-friendly bridge would be excluded from results.

Truncation

• Truncation broadens the search by including all word endings. The common symbol for truncation is an asterisk (*) but this may differ in some databases and search engines.
  • Example: wood*  will retrieve wood, woods, wooden, etc.

Boolean Operators: And, Or, Not  

• Retrieve specific search results by using the “and” operator. “And” generally retrieves fewer results than “or.” 
  • Example: Pedestrian AND Bridge will only retrieve items that contain both the words pedestrian and bridge.
• Broaden the search by using the "or" operator. This functionality is useful in a search where you are aware of similar terms for a concept, such as active transport and sustainable transport. 
  • Example: Pedestrian OR Bridge will retrieve items that contain either of the words pedestrian or bridge.
• Exclude certain items using the "not" operator. 
  • Example: Pedestrian NOT bridge will return items that contain only the word pedestrian, excluding those that also have the word bridge in them.

A literature search is usually not complete after the first set of results has been retrieved. These first results should be reviewed in order to determine if more searching is necessary, and whether the search strategy needs modification. Review the initial results of the search by skimming titles, abstracts, and keywords or subject areas. Then organize the citations into three categories:

• Definitely related to your topic.
• Possibly related to your topic.
• Not related to your topic. 

The search results that are in the “definitely related” category can be the base for further searches. Use these relevant results to identify keywords, index terms, or subject headings that have been assigned to those items.

• Run the search again, using the most relevant keywords, index terms and subjects headings.
• Note any recurring authors that appear in the definitely related category and conduct an author name to identify other relevant research by top authors they may have published on the topic.
• Use the list of works cited by a particularly relevant item as a resource for other relevant works on a similar topic.

Too many results in the “not related” category? Not enough results in the definitely related category? If the initial results are not what you expected or if no relevant results were found, refine the search strategies. Questions to ask yourself include the following

• Have an overwhelming number of results been retrieved? If so, consider simplifying the search to include fewer terms.
  • If the topic is very new or very narrow, the possibility is there may not have been much published. If this is the case, consider:
    • broadening the focus of the topic. 
    • Looking more closely at the “possibly related” category of results. You may find articles that are tangentially related to your topic.
    • Consulting sources other than online resources. 

Looking Beyond Online Resources  

Online databases and catalogs contain a wealth of information. However, not everything is available online, and a thorough literature search should at least consider the following: 

• Relevant information may be found as a component of a larger document and may not be indexed separately (e.g., a table within an article or a chapter within a book). Do not immediately discount more general material in your search results. 
• Some documents may exist only in print format. Depending on the breadth of your search, time to locate and review documents should be factored into the search schedule. 

When you start finding useful resources, collect them. For each useful item, record full bibliographic information: title, author, year of publication, journal title, and volume number (if applicable). The bibliographic details are called a “citation” or “reference,” and provide details needed to assess whether a document is worthy of review, and to help locate it. You may also wish to keep notes about the content and relevance of resources and other details, such as what database was used to locate them or libraries where they might be housed. Keeping good records helps you locate relevant resources at a later date.

Bibliographic Management Tools allow users to save, organize, and export citations with a personal database of references.

See: Zotero Guide

Excerpted from Casey & Landgraf, "Literature Reviews: How to Put It All Together," in Literature Reviews and Literature Reviews for Transportation Research Projects.

As detailed by Cooper’s taxonomy (1), literature reviews may be comprehensive, representative, or concentrated on pivotal works. The research problem statement and the detailed scope of the research project should clearly indicate what is sought from the literature review and promote a common understanding on the part of the agency and investigator before the work begins. Does the agency requesting the research desire a broad review of nearly all applicable literature on the topic to give background and historical perspective? Or is the interest narrower, perhaps focused on a particular time frame or specific subproblem of a larger issue? 

While all literature reviews support research, their specific functions and relation to that research vary. Several methods of classifying literature reviews have been proposed. These classifications inform the research and writing of a literature review.

Under Cooper's taxonomy, literature reviews can be classified based on the following:

• Focus: While all literature reviews support research, their specific functions and relation to that research vary. Several methods of classifying literature reviews have been proposed. These classifications inform the research and writing of a literature review
• Goals:  Goals include synthesis, criticism, and identification of central issues. Nearly all reviews synthesize past literature, which encompasses generalizing from multiple specific instances, proposing explanations that can resolve conflicts between contradictions found in the literature, and closing gaps between theories or disciplines by creating a linguistic framework that can be shared. 
• Perspective: Literature reviews can either present evidence neutrally or advocate for a specific position. Advocating for a specific position is not necessarily an indication of bias; it is possible for an author to fairly review and present conflicting evidence but still reach a conclusion about the correct interpretation and present it. 
• Coverage: Reviews may be comprehensive (presenting all works relevant to the topic); comprehensive with selected citations (basing conclusions on all works relevant to the topic, but only presenting a selection of the most important works in the review); representative (presenting samples of the relevant material); or concentrated on central or pivotal works. 
• Organization: Effective literature reviews can be organized chronologically, conceptually or methodologically.
• Audience: The audience for a literature review—whether specialized researchers, general researchers, practitioners, policymakers, or the general public—will affect the writing style and language used. 

____________________________________________

1. . Cooper, H. Organizing Knowledge Syntheses: A Taxonomy of Literature Reviews. Knowledge in Society, Vol. 1, 1988, pp. 104–126. 

Excerpted from Casey & Landgraf, "Literature Reviews: How to Put It All Together," in Literature Reviews and Literature Reviews for Transportation Research Projects.

1. Determine the purpose of the literature review.  All literature reviews perform some basic functions: informing the research by clarifying what is and is not known about a topic, providing context by summarizing the current state of research on the topic, and establishing a researcher’s authority by demonstrating his or her understanding of related existing research. 
   • What is the goal of the review? Who is the audience? What is the focus of the review: research outcomes, research methods, theories, or practices and applications? What is the perspective of the review: a literature review may present information neutrally, or make a case for a specific position.
2. Determine the scope of the literature review. The scope includes three major facets:
   • ​Defining the specific topic that the literature review will cover and topics that will not be covered.
   • Determining how comprehensive the review will be. It may be appropriate to seek all relevant works, a representative sample or only the significant works on a topic.
   • Defining the time period the review will cover. Literature reviews that seek to synthesize current knowledge often focus on recent research, while reviews that seek to demonstrate how a field has developed over time will naturally incorporate more historical research.
3. Review the research. While it is not generally necessary to read every piece of marginal literature in depth, thorough note taking that includes bibliographical information is critical to the research process. A University of Colorado–Denver tutorial (1) presents two approaches to note-taking:
   • The “summarize-as-you-go” method, in which the researcher writes complete sentences with citations that can be pasted into the literature review nearly verbatim. These notes should summarize a study’s context, methods, findings, conclusions, and implications.
   • The “note-basic-details” method, in which the researcher captures more basic information about a study’s context, methodology, findings, implications, and suggestions for future research, without trying to generate nearly publication-ready prose. Prevalent themes in individual studies should also be noted so they can be compared and organized when all studies have been reviewed. 
4. Evaluate the Research.  Levy and Ellis (2) outline a six-step framework for processing the information gathered:
   • Know the material. This step includes understanding the information in each cited work and the methodology used to reach its conclusions instead of simply identifying works that are relevant without describing their conclusions.
   • Comprehend the material. This step involves demonstrating how the information in a source is significant and relevant to the subject of the literature review rather than simply repeating the information within the cited source.
   • Apply the material. In this step, the review author identifies the major concepts of each work cited that relate to the study and organizes the information appropriately so it can support the story told by the literature review.
   • Analyze the material. Analysis involves demonstrating why the information pulled from sources and presented in the literature review is important. The review author should make the value of the information explicit rather than simply presenting it and leaving the reader to draw conclusions.
   • Synthesize the material. A literature review is a narrative, not a collection of facts, and synthesis is what turns it from the latter into the former. The narrative should effectively generalize the material while noting any gaps in knowledge and areas of dispute.
   • Evaluate the material. The review author must distinguish between facts, theories and opinions in the works cited instead of simply presenting all material as if each source has equal supporting evidence and validity. 
5. Organize the material and write the literature review. Remember that it is a narrative, not simply a listing of resources or an annotated bibliography. Organizing the content in a logical, thematic manner that supports the literature review’s overall goals is the most critical part of this step. Poor organization is one of the most prominently cited shortfalls in literature reviews. According to Washington et al. (3), the literature review should be organized by topic, with connections between papers made as appropriate. Within each topic, cited works should be given prominence according to their importance and relevance rather than being presented equally. There are several valid topic organizations, including:
   • Chronological, which is useful to show how knowledge in a field grows and changes over time. Descriptive, which presents what several authors write about a specific topic, followed by analysis for that topic. This method highlights topical themes that make up the entirety of the subject.
   • Descriptive–analytical, which is a variation of the descriptive organization. In this method, the analysis presents the similarities and differences among the sources for each topic rather than presenting them at the end.
   • Big-to-small-to-big, which begins with the largest and most wide-ranging studies before progressing to smaller ones and then branches out to larger studies. This organizational method highlights how the results of broader studies differ from smaller ones and is particularly useful for empirically oriented reviews.
   • Methodological, which groups studies by the methodologies they use. A brief analysis after each methodology shows what it does and does not cover, while a master analysis at the end compares and summarizes the findings.
   • “Big camps,” which is useful when there are distinct interpretations of a set of data. It can either present various topics and how the different camps’ interpretations are similar and different for each, or present each camp and its interpretations of all relevant themes as a single unit.

According to Cooper (4) and the University of Colorado–Denver tutorial (1), literature reviews may also blend these methods as appropriate. One common organizational method that many sources discourage is presenting literature author by author—that is, presenting the full content of one paper, followed by the full content of the next and so on 

_______________________________________________________

1. Writing a Literature Review, University of Colorado Denver School of Public Affairs, undated.  

2. Levy, Y., and T. Ellis. Towards a Framework of Literature Review Process in Support of Information Systems Research. Proceedings of the 2006 Informing Science and IT Education Joint Conference, 2006.

3. Washington, S., J. Leonard, D. Manning, C. Roberts, B. Williams, A. Bacchus, A. Devanhalli, J. Ogle, and D. Melcher. Scientific Approaches to Transportation Research. NCHRP Report 20-45, Vols. 1 and 2, Transportation Research Board of the National Academies, Washington, D.C., 2001

4. Cooper, H. Organizing Knowledge Syntheses: A Taxonomy of Literature Reviews. Knowledge in Society, Vol. 1, 1988, pp. 104–126.