Collect Dataprotocol and data workflows

Protocol and Data Workflows 

Summary:

This tutorial describes a protocol and data workflow for using Network Canvas tools.

Prerequisites:

To follow along, you should:

  • Be familiar with Network Canvas and its data collection process.
  • Have access to the devices you will use to run interviews.
  • For the Interviewer Classic workflows, have the ability to install software or configure your devices, or the support of someone who does (e.g., institutional Information Technology (IT) department). The browser-based Interviewer and Fresco need only a supported web browser.
  • Create or obtain the protocol file that you want to deploy (see building a protocol).

Duration:

15 minutes

Introduction

This guide walks you through three possible workflows for conducting research using the Network Canvas tools: how your protocol gets onto each interview device, and how your collected data comes back to you.

  • The Interviewer workflow — conduct in-person interviewer-guided sessions in Interviewer, which runs locally in a researcher's web browser with nothing needed to install. Protocols are imported and data is exported from within the app itself.
  • The Fresco workflow — author protocols in Architect, upload them to a Fresco dashboard, and let participants complete interviews remotely in their browser, with all data stored centrally and exported from the dashboard.
  • The Interviewer Classic workflows — for studies running Interviewer Classic, the downloadable desktop and tablet app. These involve copying the protocol file to each device and collecting a separate export from each, either through a cloud file sharing service or using a USB flash drive.

Each workflow starts with a protocol you have already designed: use Architect for Interviewer and Fresco, and Architect Classic for Interviewer Classic, which runs schema 7 protocols only.

With Interviewer and Interviewer Classic, you move files on and off each interview device yourself. Interviewer Online and Offline Workflows shows how, with or without an internet connection.

The Interviewer workflow

Interviewer runs in your web browser, or can be installed to a device you control as a Progressive Web App. It keeps everything it stores — protocols, interviews, and settings — on the device, with no server component, so the whole protocol and data cycle happens inside the app:

  1. Import your protocol from a .netcanvas file on the device. Protocols are validated on import, and protocols built against an older schema are migrated to schema 8 automatically.
  2. Conduct your interviews. Sessions are guided by an interviewer and stored on the device — encrypted at rest when you set an app lock — and can be resumed later if interrupted.
  3. Export your data from within the app by selecting sessions in the Data view. Interviewer packages them into a single .zip archive of CSV and/or GraphML files, using the formats and options you configure once in Settings → Data export.

Using Interviewer walks through each of these steps in detail. Because collected data stays on each device until you export it, export regularly, and follow the data organization and backup best practices below when deciding where exported files should end up.

Information:

Good to know:

You can rehearse this whole cycle before collecting real data: Interviewer can generate synthetic interview sessions (from Settings → Synthetic data) that export exactly like real ones. See Piloting Your Study.

The Fresco workflow

If you want to conduct interviews remotely in a web browser without installing software on each device, you can use the web-based tooling: author your protocol in Architect and run your study using Fresco.

This workflow removes per-device logistics entirely. There is no protocol to import on each interview device and no separate export to collect from each — protocols are deployed once by uploading them to your Fresco dashboard, and all collected data is exported centrally from that same dashboard.

Information:

Good to know:

The Fresco workflow requires a deployed Fresco instance. Fresco is self-hosted, so you (or your institution's IT department) will need to deploy it before you can run a study. See the deployment guide to get started, or explore the sandbox to try Fresco without deploying it yourself.

Deploying the protocol to Fresco

Rather than importing the protocol file on each interview device, you upload it once to your Fresco dashboard. Fresco validates the protocol, uploads any assets it contains, and makes it available to participants.

  1. Open your deployed Fresco instance and log in to the dashboard.
  2. Navigate to the protocols page in the dashboard.
  3. Click the Upload protocol button.
  4. Select the .netcanvas file you exported from Architect.
  5. Click Upload.

Once the protocol is listed on the protocols page, you can add participants and share a participation URL with them. For the full set of steps — including adding participants, choosing a recruitment strategy, and generating participant URLs — see the Using Fresco guide.

Conducting interviews

Participants complete the interview in a web browser by visiting a participation URL — there is no app to install and no per-device protocol deployment. Participants can use any computer or tablet with a supported browser; tablets should be used in landscape mode. Smartphones and small-screen devices are not supported.

Exporting data from Fresco

Interview data is stored centrally in Fresco's database (PostgreSQL), so there is no per-device export step to manage. Once participants have completed interviews, you export all collected data directly from the Fresco dashboard.

  1. Navigate to the interviews page in the dashboard.
  2. Either select the checkboxes next to the interviews you wish to export and click Export selected, or click Export Interview Data to export all (or all unexported) interviews.
  3. Confirm the file types and export options you wish to use (the same CSV and GraphML options described in the Data Export guide).
  4. Click Start export process.

Because data is collected and exported centrally, the per-device transfer logistics described in Interviewer Online and Offline Workflows do not apply. For full details on exporting and monitoring your study, see the Using Fresco guide.

Pulling data programmatically with the Fresco API

If you would rather automate data collection than download files from the dashboard by hand, Fresco also exposes a read-only Interview Data API. It lets you list and fetch interview and protocol data directly from your deployment over HTTP — returning JSON, including each interview's full network — so you can build repeatable export and analysis pipelines (for example, in R or Python) that pull new interviews on a schedule.

The API is off by default; you enable it and create an access token from Settings → API Tokens in the dashboard. For one-off downloads of formatted CSV or GraphML files, the dashboard export above is usually simpler. See the Fresco API reference for the available endpoints, authentication, and complete R and Python examples.

Information:

Good to know:

Fresco has some known limitations compared with Interviewer. Review these before choosing the Fresco workflow, especially if you intend to compare data collected with Fresco against data collected with Interviewer.

The Interviewer Classic workflow

If your study runs Interviewer Classic — the downloadable desktop and tablet app — the shape of the workflow matches the Interviewer workflow above: the protocol is imported on each device, interviews are stored on that device, and each device produces its own export.

  1. Install Interviewer Classic on every interview device, and configure the devices before data collection begins.
  2. Import your protocol on each device from a .netcanvas file. Interviewer Classic runs schema 7 protocols and works fully offline.
  3. Conduct your interviews, guided by an interviewer.
  4. Export your data from the Manage or Export Sessions card on the start screen, choosing your file types and export options each time, and save the resulting .zip to a location you control.

Interviewer Online and Offline Workflows covers steps 2 and 4 in detail, for connected devices (via a cloud file sharing service) and for devices with no internet connection (via external storage).

Data Organization and Backup Best Practices

Regardless of which workflow you choose, it is important to consider how you will organize and back up your data. The following are some best practices to consider:

  1. Within the folder where your exports are collected — local, cloud, or external storage — create subfolders to categorize and organize your Network Canvas data based on project, date, or any other relevant criteria.
  2. Regularly back up the data by copying or syncing the local folder to an external storage device or a different cloud storage service. Putting the data on the cloud provider is not a backup, as it is still only in one location.
  3. Consider implementing a version control system or maintaining backups at different time intervals to ensure data integrity and minimize the risk of data loss.
  4. Use the access control features of your cloud storage service to ensure that only authorized individuals have access to the data, and that they only have the level of access required to perform their role.
  5. Give clear instructions to your interviewers about how and when they should export data. Consider creating a checklist or other documentation to help them remember the steps. This should include information such as how many/which sessions should be exported together, and which export options should be selected. Remember that the result of the export process is a single zip file, so this information will not be visible to you once the export is complete until the data is extracted.

File types and export options

Whichever workflow you use, the export offers the same two file formats — CSV and GraphML — along with a small set of export options, and we recommend exporting data in both formats while you determine which suits your analysis tools. For a complete description of what the exported files contain and what each export option does, see the Data Export guide.