Introduction: The Ober2 system uses infrared reflections to record and analyze eye movements made during reading. The system's ability to analyze data from normal subjects, and the reliability of the data produced by subjects who read standard paragraphs were investigated in this study.
Subjects: Forty-two college students and 20 junior high students participated in the project. All were self-reported normal readers.
Methods: Subjects read 5 different paragraphs during each of two sessions. Ober2 analysis was attempted for each paragraph; analysis of all 10 paragraphs was successful for 38 percent of the college subjects and 20 percent of the junior high subjects. Use of manual calibration procedures did not allow any additional data to be analyzed by the Ober2 system.
Results: Data from 30% of the paragraph presentations could not be analyzed by the Ober2. When analysis was successful, grade equivalent scores based on fixations, span of recognition, regressions, fixation duration, and reading rate were provided. Using mean grade equivalents from the 16 college subjects for whom all 10 paragraphs could be analyzed, significant differences were found between results for two of the test paragraphs. Split-half reliability coefficients for grade equivalent data from the two sessions ranged from 0.84 to 0.95.
Conclusions: Although the Ober2 can provide valuable information on eye movements made during reading, problems exist with respect to its ability to analyze data. The analysis failures that occurred for approximately one-third of the paragraph presentations were frustrating and time consuming. With respect to the standard paragraphs, significant grade equivalent differences were found between several of them. These results suggest that caution be used when interpreting data from the Ober2 reading analysis system.
Files are restricted to Pacific University. Sign in to view.