For my thesis I'm doing a mixed method analysis of the portrayal of AI in contemporary science fiction films. One of my methods will be a critical discourse analysis of 30 film scenes. I wonder how I can justify film samples, bearing in mind I want scenes where there is textual dialogue between a human and a robot or AI?
This project sounds wonderful. I hope you're enjoying working on it!
From my perspective, it is a good idea to write honestly about the challenges and realities of selecting films and why you did/didn't choose certain things, and it sounds like you have a good starting point for this already with your textual dialogue bit (though remember to define what you mean by textual dialogue). Some questions to reflect on: are they all from a specific country? Are they in different languages or one? Are they studio films, independent films, a mix? So you might say, for example, you are focusing on Anglophone studio films if you are, or a mix if you're not. Again remember to define, eg what is a studio film. Are they on streaming services or did you see at festivals, cinemas etc? You might be honest about what you've been exposed to/able to locate and recognise there will be more examples you've not come across or been able to see due to access challenges (not being available to watch or not subtitled in a language you speak, for example). Reflect on how you did actually identify them - did you ask around, or had you just seen these 30 films already? Did you Google search?
So basically, describe how you selected them and reflect critically on this process and you should be fine. It is likely to be a question that comes up in a viva if you're going to have one for this project so be prepared to chat about it too, and remember there isn't a 'right' way to select texts so reflecting on how/why you did what you did is key.
Best of luck with your project!
Bethan Michael-Fox, thank you kindly for the answer, I will keep it in mind to have a clarified description of all my steps in the sampling process. I was also wondering how to justify selecting 1 scene from each film for the qualitative part? Is it okay for example to select 1 scene from each film with the same narratological function?
I short, yes I think so. You have to select something and it sounds like you have a clear approach here. I am sure your supervisor will give you feedback too on what you draft.
The simplest cop-out is 'availability'. At the end of the day, there will be only a limited amount of films where AI systems have a speaking role and, amongst those, not all will be human-AI interaction. Another good reason to make your own selection - "this has not been done before". Maybe it has - but people have looked at AI-human dialog interaction for other reasons. Still, any selection by such prior research should be looked at to see whether it is worth using - or, if it is indeed not useful - explain why you cannot employ it for your own research.
You already have one criteria: textual exchange between a human and a robot, perhaps of a scene/conversation of a certain lengh? a ne to one conversation or more participants humans/robots. I dont know what is itthat you want to analyse, what are you looking for.
Your reseach sounds great!!!! best wishes
You can built your own criteria based on: genre and subgenre, time period, author, topics, geographic context. If your criteria is contemporary you have to stablish time period: The last five or ten years? just the XXI century?
You are not really selecting a sample. You are purposefully selecting film texts that meet the criteria you have established. Only if you identify many more than 30 films that meet the criteria would you need to "sample" or select from among them. Maybe you do an initial selection and then add or tighten criteria to get a "corpus" of 30 films.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude