A Handbook for Personalized Competency-Based Education: Ensure all students master content by designing and implementing a PCBE system by Marzano Robert J. & Norford Jennifer S. & Finn Michelle & Finn III Douglas

Author:Marzano, Robert J. & Norford, Jennifer S. & Finn, Michelle & Finn III, Douglas [Marzano, Robert J.]
Language: eng
Format: azw3
Publisher: Marzano Research
Published: 2017-05-19T04:00:00+00:00

Note: The standard deviation of this test was 0.42, and the upper and lower limits have been rounded. For a detailed discussion of how to compute these intervals, see Marzano (in press).

Table 4.2 depicts how the accuracy of a single score changes across four levels of reliability: 0.85, 0.75, 0.65, and 0.55. Note the observed score (in the second column) is always the same—2.5. Consider the precision of an observed score of 2.5 when the reliability of an assessment is 0.85 (the first row). This is a typical reliability one would expect from a state or standardized test (Lou et al., 1996). The third and fourth columns represent the 95 percent confidence interval—a range of scores in which one is 95 percent confident that the true score actually falls. Examining the first row, we see that an observed score of 2.5 on an assessment that has a reliability of 0.85 has a 95 percent confidence interval of 2.19 to 2.81. In other words, one can be 95 percent sure that the accurate or true score for a student with an observed score of 2.5 is somewhere between 2.19 and 2.81. This is a fairly large spread, which becomes even larger as the reliability goes down. Consider what happens when the reliability is 0.55, which is what one might expect for a teacher-made assessment that was rather hastily put together. The 95 percent confidence interval is much larger: 1.95 to 3.05.

This example uses scores on a proficiency scale. The situation is the same when a one hundred–point scale, as opposed to the four-point scale, is used to score assessments. This is depicted in table 4.3. As indicated, an observed score of 75 on an assessment with a reliability of 0.85 has a 95 percent confidence interval between 68.63 and 81.37—a difference of 12.74 points. An observed score of 75 on an assessment with a reliability of 0.55 has a 95 percent confidence interval between 64.06 and 85.94—a difference of 21.88 points.

Table 4.3: Ninety-Five Percent Confidence Intervals Using the One Hundred–Point Scale

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.