Information retrieval from the web has increased noticeably and scanning search results from different search engines is a quite usual task. In an information search task, we recorded subjects' eye movements, when they searched for a target in a list of article titles. The lists consisted of twelve news headlines. 75% of the trials were target-present trials in which there was one relevant title in the list. In 25% of the trials the relevant title was absent. The relevant title was located equally often in the beginning (titles 1-4), in the middle (titles 5-8) and in the end of the list (titles 9-12). After finding the target the subjects pressed a button and reported which of the titles corresponded to the question. The preliminary results indicated that the mean fixation durations were longer and there were more fixations per word when the target-title was absent. We suggested that more careful search and processing were needed when the target was absent. The effect of the target position was also significant: there were less fixations per word, and more words per second were read when the relevant title was in the beginning of the list.

Interestingly, there were more fixations per word when the relevant title was in the middle of the list than when it was in end of the list. This might indicate that the subjects had to scan the whole list before they could decide whether they had found the relevant title from the middle.

Supported by the Academy of Finland (202211)