Data Diaries

Data Diaries

DataDiaries was a study to find out about online misinformation habits and we created a tool for recall-based diary data collection with special features for sharing URLs into the system and annotating them.


DataDiaries centers around the diary page. It's designed to be the single tool to add, view and annotate entries for the study. We lent into the metaphor of it being an actual diary to make it more familiar to use and easier to operate.

The diary interface The diary interface Filling in a diary entry The profile editor

Design

The interface is composed of a sidebar for navigating between the different days and main view is the diary days and their entries. The sidebar doubles as a guide by showing the days which have entries and highlights days with entries that need annotation.

The main view lays out a day of the diary in chronological order. We split up up into morning, afternoon and evening to make it more friendly and to infer some grouping to make participants think about how their habits spread out across the day.

The invisible part of the diary is the reminder system. When signing up, participants pick a schedule of reminders to prompt them to submit entries or process entries that haven't been annotated yet. Back on the profile page, participants can update that schedule whenever they want to make the reminders work best for their schedule.

To make it easy to use the diary system, participants could send in entries without visiting the website. They could email or SMS to special DataDiaries contacts and the URLs contained in those messages were automatically entered into their diary and marked for additional annotation.

We thought it would be best to capture the link at the time of consumption to reduce the chance it is forgotten and preserve metadata like the time and website information so that they could recall the interaction when they came to annotate the URL later. The inbox system fetched OpenGraph metadata about the link which is often included for social media sharing. This meant the link looked familiar when it came to annotation.

The interface itself and forms specifically were designed to be plain and simple, inspired by gov.uk's design system and it contributed the "forms" section of Open Lab's design system, Alembic.

Development

The initial idea was to create a browser extension or VPN that could automatically track all of a participant's internet use, then create an interface for browsing and annotating that data. We early on decided this was not the way we wanted to go. It would have been extremely invasive, there would be a data overload and it would still miss interactions on devices that weren't set up. Instead as an initial prototype as a basic reminders and recall interface. Participants got reminders to visit the app and write down as much as they remembered since they last visited.

We designed the interface in an iterative fashion. This helped us hone in on the features we needed to develop and drop unnecessary features without wasting too much development time.

The first iteration had too many questions and didn't have the inbox system and we found it hard to use and hard to recall the actual interactions we were interested in.

The second iteration prototyped the inbox system so participants could email or SMS in their links to their inbox rather than having to remember them when they opened their diary. We found it easier to remember the article by revisiting the URL but it took you out of the diary and with lots of links it became difficult navigating between lots of tabs.

The third iteration added the OpenGraph support to make the links look more familiar in-situ and we redesigned the interface to embrace the diary metaphor, creating the interface as it is now.