Assignment
Deliverable
The final deliverable is a research poster, size A0 portrait, both printed and digital (pdf). We will use the printed version for a final exhibition. The content of the poster should communicate your pictorial research which includes the following:
- Introduction: describe the theoretical analysis based on the theory you learned during the lectures. Converge to a research question.
- Methods: describe what research you did in a scientific way, i.e. concise but sufficient to be reproduced by others.
- Results: Visualise your results and report descriptive statistics (no inferential statistics, e.g. t-tests etc, is needed) if applicable.
- Discussion: Interpret and contextualise your results. Focus rather in the implications than the shortcomings.
Instructions
Your investigation starts with selecting 4 pictures from our yearly updated selection of images. You will be provided with a gallery of pictures without any metadata. In order to retrieve background information you can use reverse image search from google, bing, or any other search engine.
After choosing you will perform a theoretical analysis of these pictures. The theory will be explained during the lectures and is summarised by the picture above: it combines the formal elements with context and content. Think ahead of the second part of your investigation where you will perform empirical research. You converge towards one or more research questions.
Next you will perform the empirical investigation, both on humans and machines. For humans you can use the ‘visual paradigms’ which are perception experiments that you can also in other contexts. For example, investigating where people look is often important when you design interfaces, way-finding signs, or advertisements. In other words, the techniques you learn here are certainly transferable to other domains that the art works we investigate in this course. For machines you can use the method of ‘distant viewing’, which is essentially quantifying picture collections with machine learning techniques (e.g. AI).
There is a challenge when it comes to the 4 pictures you start with and the actual research you will perform. We will shortly address it here, and if needed discuss later during the lectures. After your theoretical analysis of the 4 pictures you have to formulate a research question. The first challenge is whether the research question will address all pictures, or only one or two. If you initially choose four very different pictures, you will enjoy the diversity but may end up dropping some pictures for the empirical investigation. The second challenge is that your research question will be inspired by your theoretical research but will not be limited to your specific pictures. You will need to generalise and broaden the set of pictures for your empirical investigation. Very simple example: if your research question concerns expressive colours of an artwork, you need to compare the original to a grayscale version.
Past years examples
As the course grew out a mandatory DfI master course, we used to have about a 100 students per year. To learn from past editions we will publish some previous assignments online. Be aware that the course changed, as Part 3 (doing a p5 sketch) is not part of the assignment anymore. You can nevertheless enjoy it, but do not forget to check out Part 1 (analysis) and Part 2 (empirical evaluation) as these are still in place. We also share all past images in one online gallery below.