Features

Customise & Optimise Your Grading

SM Grader is simple, easy to understand (both for instructors and students), and most importantly, 100% transparent in the auto-generated feedbacks to minimise student complaints as the students will be able to see justification for any points lost together with feedback on how to improve their performance.

Once auto-marking is completed, SM Grader generates rich feedback to be pushed back to the students' LMS accounts.

Set question types

Automate the assessment of various types of questions: multiple-choice, short text, essay, image, numerical, and programming.
‍
Assisted by AI, SM Grader first auto-marks student submissions, and after your review, provides individualised rich feedback to your students.

SM Grader can handle answers in a mix of lower and upper cases, excessive white spacing, and inclusion of quotation marks - making custom marking criteria creation a breeze.

Non-numeric answers for numerical questions? No worries, SM Grader can clean students' answers before grading them!

SaiGee can handle answers in a mix For numerical or programming questions, SaiGee can detect any hard-coded answers and give zero (or partial) credit and tell students in the feedback!

The grading of images and essays can be streamlined and aligned to rubrics, and if the basic conditions are not met (eg no image loaded, or essay is too short) can be automatically graded by the auto-marker.

SM Grader Features

Seamless experience

SM Grader allows for hybrid marking between automated and manual marking for all question types. That is, if required, you can always override any automated marking results for any student submission before publishing these results.

Complete control of outcome

SM Grader maintains a balance between automation and human control in the assessment process, ensuring ultimate control remains with the assessor.

Marking criterial & rating management

Assessors can create unlimited marking criteria and ratings aligned to rubrics, ensuring consistency in assessment scoring and standards.

Manage multiple assessors

Feature to oversee and manage permissions for multiple assistant evaluators, enabling centralised governance to ensure fair grading across the cohort.

Informative insights

Insight reports provide an overview of answer distribution and question-by-question analytics, providing valuable information for both students and assessors.

Similarity checking

Assessors can verify answers for originality across cohorts, encouraging critical thinking in students' answers and reducing copying.

Question-by-question views

Assessors can choose to review individual questions side-by-side, (not just individual students) without context switching, enhancing efficiency in assessment processes.

Rapid iterative updates

Allow iterative criteria adjustment and updates to feedback and scores, significantly improving assessment speed and overall fairness across all student responses.

Rich informative feedback

Provide comprehensive and partial scoring and rich feedback for answers, ensuring students understand the rationale behind their scores.

Student histograms

Optional capability to enable students a deeper understanding of their performance across the cohort with their scores position on a histogram.

Hybrid marking - between manual and automated

SM Grader allows for hybrid marking between automated and manual marking for all question types. That is, if required, you can always override any automated marking results for any student submission before publishing these results.

Full or partial credits and custom feedback

SM Grader facilitates full control of the marking process via rule-based automation of partial credits and custom student feedback, as well as utilisation of state-of-the-art artificial intelligence (AI) algorithms for auto-marking of essay-type questions.

Flexible marking feedback rules

The marking and feedback rules are simply Python expressions (or Python functions in case a single expression would be insufficient).
In addition, SM Grader marking & feedback criteria are case-, space-, and quote-insensitive. This feature greatly simplifies the amount of work for defining marking and feedback criteria.

Programming add-on features

Programming answer computation

SM Grader allows for hybrid marking between automated and manual marking for all question types. That is, if required, you can always override any automated marking results for any student submission before publishing these results.

Reject hard coded answers

SM Grader maintains a balance between automation and human control in the assessment process, ensuring ultimate control remains with the assessor.

Jupyter notebook edition

A SM Grader notebook edition has all the functionality of SM Grader, but also seamlessly integrates with Jupyter notebooks, allowing for direct submissions.

Programming languages supported

SM Grader currently supports both Python and R, and enables you to import relevant libraries or files to support the computation and assessment of the answer.

Key Principles

SM Grader is built on three key principles:

Control

As educators, we would like to have complete control over how we conduct & evaluate our assessments. This control and flexibility need to apply across the following four aspects of an assessment:

  • Question types: We would like to include various types of questions for a 360-degree view of student performance.
  • Partial credits: We would like to award various types of partial credits for different types of questions and different answers. We also would like to be able to award some partial credits for a correct follow-through even if the answer itself is incorrect.
  • Feedback: We would like to be able to provide rich and informative feedback under a wide variety of circumstances.
  • Batches of submissions: Rather than marking the submissions all at once, we would like to have the flexibility to mark different batches of student submissions in different times.

Speed

Simply having complete control over our assessments is not enough: we would like to do that as quickly as possible in the most effective and efficient manner:

  • Unlimited auto-marking criteria: SM Grader provides a flexible structure for defining multiple marking criteria with partial credits and custom feedback for each question. In SM Grader, you can even write your own computer code for evaluation of answers, providing total control over student scores and feedback.
  • For essay-type questions: SM Grader utilises state-of-the-art artificial intelligence (AI) algorithms for auto-marking of student answers once you manually mark a certain number of answers yourself. For example, you can manually mark about a hundred student answers and let the AI predict scores for the remaining students within just a few minutes. But don't worry, you can still review and override any AI-produced score recommendations.
  • Parallel processing: You can perform batch auto-marking using parallel processing, which can dramatically speed up the process.
  • Ultimate say: SM Grader allows for fully automated rule-based marking of assessments as well as manual override of these automation results.

Quality feedback

  • We would like to give our students timely, rich, detailed, and individualised feedback for their assessments.
  • Furthermore, we would like to receive feedback ourselves from our grading tools we use on how the students performed overall and what the common mistakes were for each question.
  • In essence, we would like to provide quality feedback to help our students see what mistakes they made and how to improve upon their performance in a timely manner so that we avoid belated feedback overlapping with future assessments.