Middle school science assessments: comparison of three grading methods

Abstract

What do grades really mean? This study was completed to compare three different grading methods for middle school eighth grade science assessments: Traditional Grading (number correct divided by number possible in a percentage), Conrad Proficiency-Based (rubric based on learning targets 5-10 score, 0 (missing)), and Standard-Based Score (rubric based on Next Generation Science Standards (NGSS) standards 0-4 score). Learning targets developed from NGSS standards were the main scoring method for proficiency-based and standards-based grades. Comparisons were made over five science units. Pre- and post- assessments, three surveys (pre, mid and post), and individual and group interviews were used to collect data on student's assessment scores, confidence levels, perceptions, motivation and student attitudes towards the various grading methods. The proficiency-based and standards-based scores were graded per learning target and each unit had between one and four learning targets. Traditional grading was completed over the full unit. The results showed a significant difference between the average scoring of the proficiency-based method and the other two methods. Proficiency-based scoring averages were higher than the other two methods, possibly because proficiency doesn't always mean that students have to get all the questions correct. To be considered proficient a student must have an overall understanding of the learning targets. It is possible that the proficiency-based scoring method could be inflating student grades when compared to the other two methods. Reported student confidence didn't show any changes when using learning targets to clarify scientific topics on the summative assessments. Letter grades were important to the students at the beginning of the study, but not as important towards the end of the study. This demonstrated that students were understanding that a score per learning target was giving them more information about what they actually understand about each science topic. Students also reported wanting the ability to re-test, which is an important curricular methodology for both proficiency-based and standards-based scoring.