In our daily interaction with online products and services we generate lots of data containing personal information. Commonly we accept sharing without hesitating to share this data in exchange of using this online experiences but we usually don't really know the future usage of our data.

This is starting to change. With new law legislation such as GDPR or controversial usage and leakages of data present on certain companies, data privacy and protection has become a public topic that has attracted interest specially in the last years.

For that reason, a new way to help citizens to understand and control their data when shared with third parties must be defined.

In this part of the project, we want to collect insights from people in order to understand which are the specific concerns when data sharing and how these concerns can be overcame.

In order to collect data we decided to create an online questionnaire where multiple scenarios are shown to the user who has to decide to donate or not.

Our main goals with our design are to simplify user experience to avoid users mental fatigue and to highlight the variables that compose an scenario to help users correctly understand each context of data sharing.

User flow through application

Previous work on this project helped us to define a limited number of variables to define each scenario:

• Requestor : who demands your data
• Data : which data is required
• Purpose : which usage will be given to this data
• Results : access to the results optained with data shared

These variables contain three possibilities each giving a total of 81 possible scenarios. For that reason, a critical mass of answers must be collected to ensure a correct extraction of insights. Each unique user will be given at least 10 random scenarios to evaluate. Later on, more scenarios can be evaluated when finished the test.

A total of 16 variables have been defined on this experiment

Moreover, we also defined two more global variables, risks and incentives. to understand their possible influence on previous scenario decision.

• Risks : willigness to donate if data is compromised and user can be identified.
• Incentives : possible compensations in order to encourage data sharing.

We decided to separate this two variables for two reasons: simplification and correlation avoidance. In terms of simplification, user mental overload is already high because of multiple scenario answering. So, adding two more questions to answer in each scenario will increase user mental fatigue a lot, ending in possible biased answers while advancing the test.
Secondly, in terms of correlation, users can also be biased by the given scenario to decide if risks or benefits influences them. We decided not to analyse in this level of detail and have a global overview of this two variables.

Finally, users are requested to complete a profiling section to segmentate their answers for future analysis.

Final screens of the application

This experiment is inspired by Moral Machine and My Goodness of MIT Medialab where people must take moral decisions related to autonomous cars and charity donations respectively.
As in previous projects, given a certain number of answers we can extract data related to the variables and better understand the main concerns on data sharing. Moreover, our final questions for profiling our users allows us to segmentate and notice differences between groups and individuals.

Once designed the experiment, I developed a responsive web based application with Node.js and Express.js. We stored the data on Firebase, a cloud-based service where we can track users answers in real time. This platform is already deployed and can be tested here

In this project I was responsible on the definition of the design and experience while interacting with the platform and also the data collection. Once defined, I developed the platform to be accessed both in mobile and desktop. Later on, we defined the database to store all user ansers for the final analysis.

The results of this project will be presented on MWC