Interpret Gradescope Bubble Sheet Assignment Statistics

Quick Links: | Prerequisites | Instructions | Troubleshooting | 

Introduction

Gradescope provides statistical information about bubble sheet assignments to help instructors evaluate student performance and assessment quality. These statistics include reliability, standard deviation, and standard error of measurement. Understanding these measurements can help instructors identify patterns in student scores and evaluate how consistently an assignment measures student knowledge. Assignments containing multiple versions provide certain statistics separately for each version.

Prerequisites

• Instructor access to Gradescope.
• An existing bubble sheet assignment with graded student submissions.
• Basic familiarity with reviewing assignment grades.

Step-by-Step Instructions

Review and Interpret Assignment Statistics

To review bubble sheet assignment statistics in Gradescope, do the following:

1. Open the bubble sheet assignment in Gradescope.
2. Navigate to Review Grades.
3. Locate the assignment statistics.
4. If the assignment contains multiple versions, select an individual version tab to view its reliability and standard error statistics.
5. Review the measurements using the explanations below.

Reliability

Reliability, measured using Cronbach's alpha, evaluates the internal consistency of an assessment.

Gradescope displays a reliability score between 0 and 1.

Reliability score	Interpretation
Below 0.5	May indicate limited internal consistency.
0.5–0.79	Indicates a level of internal consistency that may be useful depending on the assessment.
0.8–1.0	Often a target range for higher-stakes or research-oriented assessments.

Reliability depends on factors such as the number of questions, how questions relate to one another, and the range of student performance.

To improve reliability, instructors can consider increasing the number of relevant questions, covering an appropriate range of subject material, and reviewing questions with low discrimination scores.

Standard Deviation

Standard deviation describes how widely student scores vary around the assignment's average score.

• A lower standard deviation indicates that students earned relatively similar scores.
• A higher standard deviation indicates greater differences among students' scores.

For example, an assignment in which most students score between 80 and 85 would generally have a lower standard deviation than an assignment with scores ranging from 40 to 100.

Standard deviation describes score variability but does not independently establish whether an assignment is accurate or well designed.

Standard Error of Measurement

The standard error of measurement (SEM) estimates uncertainty in observed scores due to measurement error.

Gradescope uses standard deviation and reliability when calculating this statistic.

For example, if a student earns 60 points and an assignment has an SEM of 3 points, the range of one SEM below and above the observed score is 57–63 points.

This range illustrates measurement uncertainty; it is not a guaranteed range of scores the student would receive on repeated attempts.

A larger SEM indicates greater measurement uncertainty.

Interpret Statistics for Multiple Assignment Versions

For assignments containing multiple versions, Gradescope calculates reliability and standard error separately for each version.

To compare these measurements, do the following:

1. Navigate to Review Grades.
2. Select the tab for the first assignment version.
3. Review its reliability and standard error.
4. Select another version tab.
5. Compare the reported statistics.

Gradescope does not provide combined reliability and standard error statistics under the All tab.

Troubleshooting