Documentation

Hey, there! Great that you found your way here into the documentation pages of Data Foundry, or DF in short. Below you will first read about how to get stated, then about the different use-cases. Then talk about how to run a study with Data Foundry.

๐Ÿคจ General information

Data Foundry is an online platform for design researchers to collect, store, process and export data in an easy and structured way. This platform is part of the ID data infrastructure that also contains the OOCSI network. In this documentation site, we explain general things about the Data Foundry platform, how to use it in different use-cases, and then we dive into topics like running a study with DF or scripting. Come along...


๐Ÿš€ Getting started

We designed Data Foundry to be really easy to use, so you will have no trouble with the basics: How to register, create a project, or add a dataset. These steps will happen in all use-cases.

Dataset types

When you decide to store all kinds of data here, you first need to think about the type of dataset to add to your project. We have made a large overview chart to help with that. In a nutshell, you will likely use the IoT dataset for data from prototypes, Fitbit or GoogleFit datasets for commercial tracking devices, Diary datasets and others for qualitative data from participants, and Existing or Media datasets for uploading data or media. If your prototype needs a database, look no further than the Entity dataset. And then there are even more datasets.

Managing participants, and more

You can manage study participants in Data Foundry, which is really easy and also convenient, because you don't need to worry about data protection on this platform. We show you how to add participants, add/connect wearables and add devices to a project. Participants, wearables and devices can be organized in clusters and you can use these clusters to export the collected data.


๐Ÿƒโ€โ™€๏ธ Use-cases

The use-cases show how to use Data Foundry to collect data in a structured way, process it and export it. Some use-case will directly follow this pattern, others use Data Foundry more as a middle ground to store data that is exchanged between different prototypes or components of a system. Once you get used to Data Foundry, a lot of possibilities open up.

Now a list of use-cases that we have prepared for you:

Any ideas what we should add here? Let us know!


๐Ÿ‘ฉโ€๐Ÿ”ฌ Running a study

Data Foundry is made for running studies that involve physical prototypes, participants, surveys, and even digital prototypes.

Setup and onboarding participants

When you create a project and datasets for a study on Data Foundry, please make sure that you add good metadata to the project. This includes a clear title, description and other fields like relation and organization. Consider that others will see this information and perhaps decide whether to participate based on how clear and descriptive this is. For example, the title and description will be shown to participants when they sign up for a study.

You have two options to onboard participants in the system: (1) you pre-register them in the system and send each participant an individual link to sign-up, or (2) you open the sign-up and generate a general link that can be distributed to a larger audience (also as QR code). For the first option, you need to ensure that you have consent from your participants to store their contact information (name, email address) on Data Foundry. The second option is less strict because participants decide themselves whether to enter this information on the platform. Let's say, we prefer the second option, but the first option can be useful if you have a very limited participant group.

Using Telegram to communicate

We have built a Data Foundry Telegram bot (check out the use-case) that can be really helpful in several ways: you can notify your study participants of new developments or new items to check and respond to, participants can reach out to the researchers (you) in case of questions or extra information, participants can share photos, images and even their location, which will be automatically saved in respective datasets. This is great for collecting rich image data, e.g., about food and diet, health and care, or other topics.

Going remote with digital prototypes

You can let your participants interact with their data or interactive digital prototypes on a participant page - and collect data from these activities in same or new datasets in your project. The participant page is hosted on the Data Foundry and basically a HTML webpage with CSS and JS as you need it. We have built a small API for accessing participant-specific functions from this website. This way, you can store data, retrieve and visualize data, build participant choice profiles and more.


๐Ÿฆพ Prototyping with Data Foundry

This is a new topic and wokr-in-progress for now. You can already use scripting, and several interesting APIs in your prototypes.

Scripting

We have built a really cool feature into Data Foundry: scripting. This allows you to do a lot of interesting things from automatically processing incoming data from a prototype to sending message to participants on Telegram or communicating with your prototypes via OOCSI.

Scripts are written standard JavaScript and receive data from OOCSI or from a Telegram message. Then they can filter the data, read and write to datasets, and send messages out again. We are continuously improving this and adding functionality. Check out the function reference for all current options.

Designing with APIs

Do you want to use chatGPT or GPT-4 in your prototypes? Or use speech to control a device, or even let devices speak out? This is the right section for you! Check out the designing with APIs page for more information on how to use Data Foundry APIs in your prototyping.


๐Ÿ‘ฉโ€๐Ÿ’ป Data tools

We have built all kinds of tools into Data Foundry, which help in working with data in datasets and beyond. You can find them on the left sidebar under "data tools". In short, we have a data export tool that allows to select and combine different datasets and export them exactly how you need them. Then we have embedded a version of RAWgraphs for super easy data visualisation. And finally, we have added a media transcription tool that can extract text from audio and video files, and also extract text from images. This is great for auto-transcribing interviews.


โš”๏ธ Data protection

Data Foundry was designed to protect data and encourage responsible data practices. We store participant data only with explicit consent and also separately from the actual study/design datasets. We de-identify participants by numerical IDs and use public parameters to allow for labeling participant cohorts. There is more to tell here, coming soon.


โ˜•๏ธ Special topics

When designing and building Data Foundry, we thought about a lot of things and how to make them easy or better. This section is clearly work-in-progress. Why? Because we always have new ideas.

Any ideas what we should add here? Let us know!


๐Ÿ‘€ Support

How to get support? Check this documentation first, then our FAQs, and you should definitely check out the Teams community. You also contact us directly, but the Teams community gives you better chances for quick help. ๐Ÿ™ƒ


๐Ÿ”ฅ Develop + Contribute

Data Foundry offers an API that is available to access most of the functionality. With this API you can design and build your application on top of Data Foundry. Focus on the core of what you want to do and leave all the details about authentication, data storage and protection, scaling and scripting to the platform.

The documentation is available and you can contact us for an API key. We are happy to talk about new ideas of how to connect to and from Data Foundry, what else to do and how to make things better.