class: center, middle, inverse, title-slide .title[ # The assessment ] .subtitle[ ## A few tips ] .author[ ### Lincoln Colling ] --- <!-- vim: set ft=markdown tw=80 spell spelllang=en_gb: vim: set conceallevel=0 foldlevel=1: --> # Outline - Your task is to implement a computerized experiment task using Matlab. - The recommended task is a standard Colour/Word Stroop task with *at least* three colour words (e.g., blue, red, yellow). - However, if you'd like to design a task that is relevant to your Master's thesis then you're also welcome to do so **if you check with me first**. --- ## Things to consider The aim is to develop a computerized task that is as realistic as possible. This means that your task should: - Have multiple trials (not just 3 or 4) - If your code is well-designed then it should be trivial to alter the number of trials for particular use cases - You should consider things like trial order - You should think about how the data is saved and **what data** is saved Programming is not just about writing code. It's about *designing* your code so that your code can *achieve your objective*. This means that **most** of the assignment will be about **designing** your code. Only after you've planned out exactly what *your code will do* should you start writing code. *The [assessment page](../../assessment.html#things-to-consider) has more information about the sorts of things you should be thinking about when you're designing your code* --- ## How it will be assessed You will be assessed on the quality of your code... What does this mean? - First and foremost, this means that **your code should run** However, I will also be looking at other things - Do you write your code in such a way so that you make use of *reusable components* that could, in theory, be used as part of different experiment - Do you include documentation for reusable functions that you write, so that, for example, other members of your lab could make use of them - Does your code save files in a sensible way - Does your code save all the information you'll need to analyze the data for your experiment I won't give you a bullet point list of everything that your code needs to do, because thinking about the *design of your code* **is the task YOU need to solve** <!-- ## The task The task should be as realistic as possible. This means that it should include: - Multiple trials - Should record some kind of response Additionally, it should save the data from the experiment. This data should be saved in a way that would allow it to be analyzed later. E.g., if you record responses, these should be matched with a stimulus/condition --> --- ### What you need to submit You'll need to include the following things with your submission. 1. The `Matlab` code for your task - This might be one file, or it might be multiple files (for example if you decided to employ multiple functions) - Please ensure **that your code runs**, because I will be running it! 2. Data from one participant (yourself), so that I can see what the output looks like 3. A description of the task: - What kind of task is it - What are the different conditions - What responses are being recorded / why? I expect this *description* to be fewer than 750 words. The purpose of this description is so that I can make sure that your code *does what you intend it to do*. --- ### What you need to submit - The submission will take the form of a word/pdf document with your description and one or more `.m` files with the actual `Matlab` code. - You **must** zip these together in a `.zip` file because **canvas** renames uploaded files so uploading the files individually will break things and it will make it impossible for me to run the code! --- ### Getting help Peter Scarfe’s PsychToolbox Demos provide lots of good examples of using PsychToolbox. Note, however, that these are designed as **demos** so _please_ don't just copy one of the examples and submit that as your assessment, because these don't do everything that _your code will need to do_ (e.g., they don't have very good participant instructions, and they don't save the data in a useful way etc) This website, together with the stuff that was covered in class will give you all the building blocks you need to put together an experiment. You'll still need to put those blocks together, but you can make use of any resources you can find. --- ### Finally **Keep things simple** but **do the simple things well** Trying to overcomplicate things is always risky, so try to avoid it.