The effect of presearch experience on the success of naive (end‐user) searches

Naive (end‐user) searchers were assigned to two different experimental conditions and given conceptual instruction in online searching. The subjects, fourteen graduate students in mass communication, were searching on topics framed in a mass‐communication context. Seven subjects were assigned to conduct a search in printed indexes followed by a search in DIALOG electronic files. The other seven subjects were assigned to conduct a search in DIALOG electronic files followed by a search in printed indexes. The precision of the electronic searches was measured, and the total relevant items found in the combined search was computed. The first hypothesis, that the manual/electronic searches would be more precise than the electronic/manual searches, was not supported by standard measures of statistical significance, but the direction of effect was supportive. The second hypothesis, that an electronic/manual search will locate, overall, more relevant items than a manual/electronic search, was also not statistically supported, but did show supportive direction of effect. Small sample size may be the cause of the statistically insignificant findings. The study suggests that for naive searchers, an electronic search followed by a manual search may be the best approach for covering a large and complex topic with an expanding literature. © 1986 John Wiley & Sons, Inc.