Keyword or Hierarchical Searching? A Quantitative Comparison of World Wide Web Searching Methods by High School Juniors in North Iowa

In the spring of 2005, twenty 11th grade students from north central Iowa were given 5 topics to search on the Internet. All students were allowed 5 minutes to search for each of 5 topics and used identical equipment to perform the searches. The population consisted of 10 male and 10 female students. As there is a marked lack of ethnic and cultural diversity in the community, all students were Caucasian with English being their first language. All students were of similar socio-economic class since there is little variance in the community. Students were chosen based primarily on a range of grade point averages. Students who were selected to participate in the study had a cumulative grade point average of between 2.5 and 3.5 for the current school year. Students who participated had a working knowledge of the hardware and software being used. The participants were divided into 2 groups of 10 with 5 boys and 5 girls in each. Each group searched for 5 topics using Google. One group searched using Google Search, a text based search engine utilizing keywords to locate related websites, the other group used Google Directory, a hierarchically organized directory of websites catalogued by topic. Since the directory search was new to most of the participants, a short lesson was given to all participants on the basic concept and usage of hierarchically organized directories. Each group was given 5 minutes to search for each of the 5 topics. A search log kept by each participant asked questions relating to the participant\u27s experience with the search of each topic. The search log allowed participants to rate each search according to overall difficulty, difficulty determining keywords or search path, difficulty finding information, and the speed that the information was found. The log also contained items that the researcher addressed by analyzing the search histories of each participant to determine the number of inappropriate sites found, the number of nonfunctional sites found, and the number of sites found unrelated to the search. Only the top 20 sites returned at the end of a search were considered. While the data showed that both types of searching returned no non-functional sites within the top twenty returns. Both also did exceptional jobs of returning relevant sites with Google Search returning a minimum of 19.5% and Google Directory 22.2% of sites that did not pertain directly to the search query. There were larger differences between the two methods when considering participant perceived speed and difficulty. In nearly all instances the participants chose Google Search over Google Directory as faster and easier. Factors that may have influenced these results include lack of prior experience with the directory form of searching, the cognitive form of the topics used, and lack of prior knowledge of the topics. Keyword searching showed a definite advantage in these circumstances as a term from the topic question could be searched even if the searcher did not have knowledge of the topic itself. Where as without prior knowledge or understanding of a topic, a searcher would not know where to start using a topical or hierarchical directory. These factors and others combined to create a situation where participants favored the path of least resistance or at least the path that was more familiar. Keyword searching, with modem relevancy ranking technology was definitely the favorite of these students