THE EFFECTS OF VARIOUS FORMS OF DATA ON THE ABILITY TO ANALYZE TRENDS IN STUDENT PERFORMANCE.

摘要

Special education teachers are routinely taught to collect and analyze student performance data. Data collection and analysis are assumed to be necessary for accurate analysis of trend in student performance. The possible types of trend are an upward trend (when performance improves), a level trend (when performance remains the same), and a downward trend (when performance worsens). There are a few studies which have examined either teachers' use of observation to analyze trend or compared observation with other forms of data. Some possible forms of data are: raw data, the number of correct and incorrect responses; raw data on graph paper; and, another method of graphic analysis labelled the 6-day line of progress. The present study examined the effects of observation and the three forms of data described above on the ability to accurately analyze trend. The following three research questions were investigated: (1) What form of data permits the most accurate analysis of trend? (2) Does trend analysis differ in accuracy according to type of trend? and (3) Is there an interaction between the form of data and type of trend? A two-factor experimental design was used with four groups of ten subjects each. All subjects were randomly assigned. Subjects analyzed fifteen samples of student performance. Observation was provided through the use of videotaped segments of simulated student performance. All subjects viewed the videotapes with or without additional forms of data. The subjects made judgements regarding the trend of student performance following exposure to each sample. The data were analyzed using a two-way ANOVA. Both the main effect for form of data and the interaction between form of data and type of trend were statistically significant beyond the .01 level. Inspection of the results revealed that all subjects were able to accurately analyze upward trends in performance, but data were required to accurately analyze level and downward trends. These results are interpreted as support for the practice of training teachers to collect and analyze student performance data.