Assessment

There are three forms of assessment for this seminar. They break down as follows:

Assessment Due Date Worth
Portfolio (Sem 1) Weekly 15%
Presentation (Sem 1) Arranged 0%
Essay 26 January 2024 40%
Portfolio (Sem 2) Weekly 15%
Presentation (Sem 2) Arranged 30%

Good Scholarly Practices

All work submitted for assessment in this course should be carried out following the University’s policies on Good Scholarly Practice.

Portfolio

Over the course of each semester, you will build a portfolio of reflections on required papers. Each week, you need to submit two (~100-200 word) entries, one for the Tuesday class and one for the Thursday class. An entry should:

  • assume a reviewer role (see Lecture 1)
  • briefly summarise the main idea of the article (1-2 sentences),
  • comment on the article

The goal of the weekly entries are to ensure that you are prepared to participate in class discussions. In class, you will may also be asked to comment on the discussion and state if you revised your belief and why. You only need to pick one of the assigned readings to reflect on with your portfolio entry (although you are welcome to comment on both, or take multiple reviewer roles if you like)

Portfolios will be marked at the end of the semester. Five of the 20 entries can be missing, no questions asked and no detriment to the marks. Otherwise, no extensions.

Essay

Content and Format

Your essay should be based around one or two papers of your choice on a topic in cognitive modelling. It must discuss empirical findings in the context of a computational model or models to a non specialist audience. The easiest way to do this is to find a single paper that includes both empirical and modelling results, but the paper in question should be very meaty, i.e., fairly long and including many experiments or a thorough comparison with other approaches. If you cannot find a suitable single paper, you can look at two shorter papers covering different models of a single phenomenon, or additional experiments testing the same model. Marking will take into account the difficulty of the chosen article(s)/theme(s), but we find that stronger essays tend to be those that focus on a single article. It’s okay to choose an older paper, but in that case you should explain how it relates to more recent work.

You must have your choice of paper(s) approved by 11 November 2024. Post your intended paper(s) on the Piazza note for essay topics. For help choosing papers, see the tips on the website in the first instance. If you are still having trouble, feel free to contact one of the instructors well before the deadline. Essays may not focus on papers that are the subjects of other student presentations. So make sure to look at the essay note to avoid overlapping topics.

Whether you choose one or two main papers, your essay should discuss the context of the work (i.e., behavioural findings or philosophical questions addressed), summarise the model(s) and experiment(s), and critically evaluate the work. A good essay will include some material you get from “reading around” the topic. You may also wish to address one or more of the following questions: how does this work relate to other models/approaches we have studied in class? What questions are raised by this work? What further experimental or modelling work might help to address these questions? Your goal is to demonstrate that you can read a cognitive modelling paper, understand its methods, evaluate its claims and place it in context in the field.

Your essay should be between 2000 and 2500 words, including headers and figure captions but not the bibliography, and should be written at a level that an interested but non-specialist reader would understand – someone who has some background in cognitive science but not necessarily in the specific area you are discussing. Articles in the journal Trends in Cognitive Science, many of which have been included as background readings in this course, are good examples of this level of writing (although the format/content of these articles is different from what you will write). Do not include an abstract.

Marking

Basic Criteria

  1. Provides a clear description of the phenomena motivating the work and relevant empirical data from this and/or previous works.
  2. Clearly states hypotheses being tested by the model(s) and their relevance to the field of cognitive science.
  3. Clearly summarises the model(s), including any key technical details and assumptions that are made.
  4. Summarises key experiments and results obtained in the chosen paper(s) and explains how these bear on the hypotheses being tested.
  5. Demonstrates a detailed understanding of the chosen paper(s).
  6. Discusses strengths and weaknesses of individual model(s) and evaluation(s).
  7. Well written

Additional Criteria (beyond chosen papers)

  1. Includes original discussion.
  2. Explicitly compares alternative approaches or competing hypotheses.
  3. Reviews additional behavioral evidence.
  4. Reviews other uses of the model.
  5. Relates the topic/model to broader isues and themes from the course/cognitive science or society.
  6. Proposes useful extensions to the model or further ways to test them

Presentation

Format

  • Plan on presenting for 30 mins, followed by discussion/questions. We will give you time indicators and cut you off after 35 minutes regardless of whether you are done; if you aren’t, it is likely to affect your mark.
  • You may use slides or whiteboard, but should be organized and use visuals clearly and effectively (see AV section below).

Topic Assignment

For the first semester, topics have been prepared for you. You should have a look at the topics on the website and fill out the survey on the first Thursday of class.

For the second semester, you are encouraged to propose your own topic (the website will be updated at the end of Semester 1).

To propose your own topic:

  • you must select one/two papers for the seminar to have people read, and have them approved by instructor.
  • typically these should also be the topic of your essay
  • you must draft 2+ questions under discussion

Content

Presentations should include:

  • An introduction to the topic: what is the psychological phenomenon being modelled, and what question(s) are being addressed in the papers? How do they tie in with some of the big questions we’ve discussed?
  • An explanation of the models presented and results obtained.
  • Some analysis or comparison of the models, e.g. What are their strengths and weaknesses? Do you find one more believable, and why? Are there remaining issues that neither addresses?

Good presentations will also normally include examples, figures, or diagrams to illustrate important concepts or results.

NOTE: 30 minutes is not a long time! You will need to decide how much detail is appropriate, and which information is most important. Please do not try to present everything. Remember, all students in the class should have read the assigned papers. Prepare your background material with this in mind.

Logistics

  • If you don’t want to use your own laptop, you must email us us your slides (include the [scm-pres] tag in the subject of your email) by 7 p.m. the night before your presentation. It is a good idea to do this anyway, to ensure a backup copy in case your laptop/USB fails, and also for marking purposes. If you do not do this, and you have a technical problem that prevents you from presenting, you will receive a 0 on the assignment. We have intentionally set the deadline on the previous night so that you will not be tempted to stay up late working on your slides. Not sleeping enough is a great recipe for a poor presentation.
  • When sending slides, PDF files are preferable.
  • If you plan to make extensive use of the whiteboard rather than slides, you must speak to us well ahead of time (at least 4 business days). Some use of the whiteboard is fine without advance notice.
  • Please arrive five minutes early on the day of your presentation, to allow time to set up.

Delivery

Part of your mark will be based on your delivery. If you wish to do well on this part of the assignment, one word will tell you how: practice! Practicing your presentation out loud will not only help you figure out how long it will take, but will also make you more confident and will allow you to work out how to explain difficult concepts ahead of time. See the marking sheet for the criteria we will be using to assess your delivery.

Having a well-organized talk is a very good start to keeping your audience engaged and your message understood, but it’s important to understand that people have limited memory and attention. That means they will not remember everything you said in the beginning of your talk by the end (or even the middle). You need to help them out by providing verbal cues (rarely, if ever, are bullets needed!) reminding them what you have already told them, what’s coming up, and how everything fits together. I’ve heard these referred to as “signposts”. Examples include: “OK, so that’s the overview of the system, now I will talk about each of those parts in more detail”, “I’ve now explained how they use X and Y to produce Z. Remember that what they are going to do with Z is A, so I’ll now give you some more detail about that.”, “I’m not going to talk about this part of the paper in any detail, but if you have questions about it I can answer them in the discussion after the presentation.” In other words, think about your transitions, not just your slides. It is also often a good idea to include actual summary slides after each major section of your presentation, recapping the main points before moving on. You should not need more than one or two of these.

Audio-visual aids

You will also receive a mark for your use of audio-visual aids (slides, whiteboard, or any additional media). Here are some tips on appropriate use of these tools:

  • Perhaps the most common mistake for beginning presenters is to put too much information (especially text) on the slides. If you practice your presentation (see above), you should know it well enough that you do not need to use the slides to remind you of every single thing you want to say. You should especially try to avoid the trap of reading text directly from your slides. Slides are better used for highlighting key points and showing pictures (examples, figures, etc.)
  • Make sure your slides are easy to read. Fonts should be no smaller than 20-24 point. Different colors can be used for emphasis, but should be used sparingly. Remember that color-blindness is fairly common, so try not to rely on color to distinguish things.
  • If you reproduce a figure or table from a paper, make sure any labels or text are still visible, or add your own labels in larger fonts.
  • Sometimes it can be easier to go through an example or draw a picture on the whiteboard than on the slides. This is ok as long as you are prepared: know how the whiteboard is set up in the room and how to switch between the projector and whiteboard without taking lots of time. And know what you are planning to draw!

Citations

  • You may find that it is useful to include images, figures, or tables from the papers you are discussing or other sources. If you do so, you must cite your sources, even if the source appears to be clear from context. Papers should be cited as (Lastname, year) on the slide, with a full citation on a separate References slide at the end. Web sources should be cited by URL.
  • Direct quotations may also be used if cited (as above), but note that long quotations are usually a crutch used by students who do not understand the material enough to restate it in their own words. Use quotations with care.

Tips to avoid common mistakes

The following list of pointers is intended to help avoid common pitfalls in presentations.

  • Explain all important technical terms that won’t be obvious to your fellow students.
  • Be sure to provide motivation for the models and experiments. Why did the authors do what they did? Why should we care?
  • Make sure that fellow students who haven’t read your specific paper can still follow the presentation.
  • Don’t necessarily describe every experiment in a paper if it would come at the cost of clarity or coherence.
  • Always give citations for figures, quotations, and anything else you get from an external source, even if that source is the paper you’re presenting.
  • Practice your presentation, to ensure good pacer and transitions.
  • Don’t read directly from your slides.
  • Only include figures or tables copied from the original paper if you can clearly explain their contents. If you’re only going to talk about a small part of the figure, it may be better to create your own.