A software company has designed a new version of its search engine, and wants to investigate consumer response. A randomly selected group of users of the old search engine is selected, and given training with the new version of the search engine.
a. The users are given a set of search tasks to complete. The total time required to locate the required information with the old search engine and the new version is recorded for each user (the order of use of the search engines is randomized). Is there evidence, at the 5% level of significance, that the new search engine is faster?
b. Researchers realized that the data from part a would be affected by the knowledge gained during the first search. They design similar tasks, and then repeat the test. Under the new conditions, is there evidence, at the 5% level of significance, that the new search engine is faster?
c. Users are also asked to rate their impressions of the old and new search engines. Twenty one of the users preferred the new version, eight preferred the old version, and five found them about the same. Is there evidence, at the 5% level of significance, that users prefer the new search engine?