.svg)
BackAddressing the threat of potentially harmful online video content
EXPERTISE
Behavioural and Communications
A series of immersive experiments for Ofcom
Video Sharing Platforms (VSPs) - and social media in general - have the capacity to bring an extremely wide range of content direct to any user in a way that encourages immersive engagement.
In some cases, the content may be illegal, and users should not be exposed to it. In others, the content could be legal but presents a risk of psychological, physical, or financial harm to particular groups. As the UK communications regulator, Ofcom needed to know how different types of intervention might encourage VSP users to skip or report this kind of content, thus helping to protect themselves and others.
Verian was commissioned by Ofcom to run a series of experiments to measure real behaviours within a natural VSP setting. Using our platform for online experiments, Behaviour Change Lab, we created a simulated video sharing platform which was similar in appearance and functionality to YouTube and Vimeo, and which allowed users to interact with content in a realistic fashion.
In each experiment, we made changes to the interface’s features – for example altering the ‘choice architecture’ to encourage skipping or reporting potentially harmful videos, or showing guidance on how to report content.
The results indicated that showing alert messages and including an ‘auto-skip’ function will be effective ways to encourage users to skip harmful content, and that increasing the salience of the report button and seeing an interactive micro-tutorial can both increase the rate at which users report harmful content.
Impact
These were the first online randomised controlled trials commissioned by Ofcom. They have established a robust, ecologically valid framework for experimental design and analysis that can be carried into their future work. Results from these trials have further served as evidence to underpin Ofcom’s engagement with video sharing platforms.
Ofcom has recently published multiple discussion papers based on the findings of this programme. The primary discussion discussing default effects and alert messages is accessible using the button on the left. Additional discussion papers are available below: