JSTOR results

A previous post, JSTOR search, discussed my attempts to gather information about good web design using JSTOR, a large multidisciplinary database. After three searches I came up with a satisfactory list, in which about 1/3 of the top 25 results contained information that would be significantly useful to a novice web designer.

The highest-ranked useful document was result #4:


In this study, Webster and Ahuja investigated user disorientation on websites and concluded that there are a couple of concrete ways that web designers can minimize disorientation (and thus frustration, giving up, and other problems) on a multi-page website:
1. Design each page so that the navigation bar is always in view no matter where on the page the user is (for example, if you have a page where the user may scroll down, have the navigation bar move down as the user scrolls so that it stays visible)
2. Have your navigation system use a 'tree' design rather than a 'bar' design: that is, present a small number of initial choices to the user, each of which has its own set of follow-up choices, rather than presenting a large number of more specific choices all at once.

Other promising results that this search produced were:

Critical Design Ethnography: Designing for Change
Sasha A. Barab, Michael K. Thomas, Tyler Dodge, Kurt Squire, Markeda Newell 
Anthropology & Education Quarterly, Vol. 35, No. 2 (Jun., 2004), pp. 254-268

Understanding the Impact of Web Personalization on User Information Processing and Decision Outcomes
Kar Yan Tam, Shuk Ying Ho 

MIS Quarterly, Vol. 30, No. 4 (Dec., 2006), pp. 865-890

Factors Affecting Web Site Visit Duration: A Cross-Domain Analysis 

Peter J. Danaher, Guy W. Mullarkey, Skander Essegaier

Journal of Marketing Research, Vol. 43, No. 2 (May, 2006), pp. 182-194

The JSTOR search was fairly simple to conduct because I am in the early stages of gathering information on web design. So, like a student who just needs 'a bibliography of ten sources, any ten,' almost anything I find on the topic will be relevant and useful to me. However, I still learned important lessons from this search. The most significant thing I learned was the usefulness of the fact that JSTOR classifies its included journals by discipline and allows you to use this fact to limit your search. My biggest problem with the search for 'web design' was that JSTOR is a large multi-subject database and that 'web' and 'design' are common words that are meaningful in multiple contexts. My initial search was undermined by this fact and returned a lot of biology articles about spiders. The ability to limit which parts of the database were searched (which as far as I know is unique to JSTOR among similar multidisciplinary databases) enabled me to search JSTOR as though it were a subject-specific database enough though in fact it is not. This limiter was the key in going from mostly irrelevant to mainly relevant results. Once I had narrowed down the search to discipline-specific journal collections it was a fairly simple matter to  refine the search further in order to collect some relevant documents.

JSTOR search

My first search for web design resources used the JSTOR database. Since I have not gathered any information on the topic yet and because JSTOR is a database that covers a wide range of academic disciplines I started with a broach natural language search for "web design" to get an idea of what was out there.
This is what my results looked like:


As you can see, not only are there an unwieldy number of results, but two of the three top hits are about...spiders. I probably should have anticipated this homonym problem.

Accordingly, I went back to the Advanced Search screen and re-ran the same natural language search, this time using the "narrow by discipline" function. This is a pre-search limiter that allows you to limit your JSTOR search to a subset of JSTOR journals, either by selecting individual journal titles or by selecting disciplines, academic field categories into which JSTOR has sorted all its journals.

My new search then showed in JSTOR as:
(web design) AND disc:(anthropology-discipline OR business-discipline OR historyofscience-discipline OR libraryscience-discipline OR markadvert-discipline OR psychology-discipline OR sociology-discipline)

This search produced about 11,000 results and, as the screenshot below demonstrates, the top hits are much more relevant than in the previous search.






11,000 results is still obviously too many. I now have to start making choices about exactly what information I want. I am all for adding limiters rather than changing the language of the search when you can, because the behavior of limiters is (or at least should be) completely predictable whereas in most cases you can never be sure how changing the language of your search will affect the results. So I again modified my search to add the limiters of English language, publication date of 2004-2014 (inclusive), and document types of article, book, or miscellaneous (excluding pamphlets and reviews). This new search is expressed in JSTOR as:
(web design) AND ((cty:(journal) AND ty:(fla OR edi OR nws OR mis)) OR cty:(book)) AND (year:[2004 TO 2014]) AND la:(eng) AND disc:(anthropology-discipline OR business-discipline OR historyofscience-discipline OR libraryscience-discipline OR markadvert-discipline OR psychology-discipline OR sociology-discipline)

It cut the results list to 5,888. I read the first 25 results and found that about 1/3-1/2 were relevant. This was satisfactory so results will follow in the next post.