Evaluation and grading of MSc theses at IIK

Each master's thesis is to be jointly evaluated and graded by an external and an internal evaluator, both nominated by the department. The internal evaluator is also the administrator of the evaluation committee. The internal evaluator must not have been involved in the supervision of the student·s. If any of the evaluator has a conflict of interest that the department has not been made aware of during the nomination process, please inform the department's administration.

The evaluation and grading is based on the thesis and a brief report from the main supervisor. The work is to be assessed according to the criteria that are normally used for scientific technological works such as conference contributions, journal papers and similar. The assessment criteria are detailed below and are communicated to the students at the beginning of the semester by the course coordinator. The evaluation should take into account the technological and scientific challenge posed by the task, as well as what can realistically be expected to be achieved within the duration of the master’s thesis work (20 weeks, full-time).

There are guidelines for the use of AI Tools.

The evaluation and grading must be documented by filling in an evaluation form. The evaluation form must be submitted to the department's administration and the grade must be registered in Inspera – see below – within 12 weeks after the submission of the thesis. The form will be made available to the student·s and the main supervisor

Evaluating, grading, filling in an evaluation form and submitting it, and registering the grade in Inspera in time is the joint responsibility of both the external and the internal evaluator.

Assessment criteria

Scientific and technological challenge and results

• Challenge: Is the task the application of known methods on a standard class of problems, or are any of these new? Have similar tasks been dealt with before? What are the requirements for abstraction and innovation? A simple task must never be given a top grade.
• Substance of the results: Has the student built a (working) implementation? Were simulations carried out and/or analytical results obtained? Has the student produced a specific, unambiguous design, or is a thorough trade-off between alternatives based on sound criteria performed? It should be taken into account that working towards concrete results may require time to get equipment, software, etc. configured and up and running.
• Broadness: What are the prerequisites in terms of maturity and overview of the field? How broad are the issues addressed and what is their scientific/technical/disciplinary span?

Significance and originality

• Novelty: Does the work give new facts, ideas or insights? Are there innovative elements?
• Relation to the state-of-the-art: Has the student shown sufficient insight into and overview over the problem domain? Does the manuscript include representative references to other work within the domain? Is the student able to put their own work into a wider context and the work of others? The references and bibliography are important in this context.
• Utility: Is the work practically or theoretically useful? With respect to this criterion, be aware that it is the relevance of the work with respect to the formulation of the task that is evaluated, not the relevance of the (given) task itself.
• Autonomy: What has the student achieved by themselves from the given task/problem? For a fair evaluation, the evaluator must be informed about the evolution of the problem formulation throughout the project, as well as the support provided to the student.”.

Methodological quality

• Method: Is any method used in the work, and is there an explicit choice of method (if relevant)? Are suitable formalisms chosen and used?
• Methodological reliability: Are the methods used and the investigations done sufficient to support the conclusions?
• Logical consistency: Are there contradictions in the report? Do the observations support the conclusions? Are there alternative interpretations of the observations? Is the basis for the conclusions (e.g. the observations) complete?
• Procedural quality: Is the working procedure well-documented and is any quality assurance with respect to the work/results carried out? When the task is to develop a (software) system, the SW quality (in the broad sense; documentation included) should be taken into account.

Presentation

• Structure: Is the report written in a manner that makes it easy for the reader to get an overview over starting point/objectives, what is done and the conclusions/results, and to maintain this overview throughout the reading? Does the report contain the necessary elements such as abstract/summary, table of contents, introduction, etc.?
• Clarity: Is it easy/possible/difficult/impossible to follow and understand what is written? Are proper references given with regard to the source of claims and information? Are the references in the list complete?
• Information density: Is the relation between content and volume satisfactory? How long is it between the “golden nuggets”? Are information about details unnecessary for progression of the reading put into appendices?
• Style: Does the student make a distinction between essentials and details? Is the work interesting or boring to read? Is there an overview chapter/section (cf. structure) which makes the work more accessible?
• Illustrations and tables: Are illustrations and tables clear, reasonably “self-contained” and informative, or is there a duplication between text and illustrations/tables. Could some of the text be better presented/conveyed in the form of illustrations or tables?

Summary and overall evaluation

All experience shows that the final evaluation is best based on an overall judgment of the quality of the work. We recommend that equal weight is put on the four main groups of criteria. There is no similar recommendation for weighting of the sub-criteria. These recommendations are embedded in the evaluation form.

With respect to the sub-criteria such as challenge, novelty and utility, it is mainly the responsibility of the main supervisor to ensure that the problem/task formulation is relevant and that it will be possible for the student's/s' work to meet these criteria. The student/s is/are normally not in a position to do this when the problem/task is chosen.

In preparing this part of the form and in establishing the grade, please consult NTNUs generic description of MSc thesis grades for the NMT area. The use of concepts and terms should be consistent. A guideline for converting an evaluation in percent points to the A-F scale is available here.

How to proceed

It is common practice for the evaluators to first go through the thesis individually and afterwards discuss and agree upon the evaluation and grading during a physical or on-line meeting. This can also be handled by e-mail, but it might be less efficient.

As for filling in the form, this can be handled by both preparing comments that are merged after the coordination meeting (this is usually done by the external evaluator); alternatively, one of the parties can make a draft that is supplemented by the other.

Quality assurance

The evaluation should be quality assured according to the following procedure for quality assurance of the evaluation of MSc theses (pdf) (in Norwegian).

Registering the grade in Inspera

Once system access is clear, both evaluators are given access to Inspera Assessment. From there, you will receive a link to the thesis you will evaluate. From there, you can access both the thesis and additional submitted documents (if any), and the registration page.

Both evaluators must register the grade in Inspera, when the thesis evaluation is done. When both are done registering the grade, the evaluators must also verify the final grade. We recommend conducting this registration together during the meeting between the two evaluators, to avoid mistakes and save time. Note: It is important that both evaluators register the grade first, then verify it. Do not verify the grade before both evaluators have completed the registration.