Complexdatacollective Interviewer Versions Save

Simplifying complex network data collection.

v6.2.1

2 years ago

Network Canvas Interviewer 6.2.1

This is a small bug fix for the 6.2.0 release, which addresses a visual glitch in the rendering of the pairing code dialog, and the "other" dialog on the categorical bin interface. It should be installed by anyone running version 6.2.0.

If you are running a version prior to 6.2.0, please consult the release notes for that release before updating: https://github.com/complexdatacollective/Interviewer/releases/tag/v6.2.0

Changelog:

  • Fix a small issue with the rendering of the Server pairing code dialog, and the 'other' option on the categorical bin interface.

v6.2.0

2 years ago

Network Canvas Interviewer 6.2.0

This is a significant feature release, containing numerous changes and bugfixes. It introduces schema version 6, but remains compatible with schema versions 4 and 5. It should be installed by all users currently running version 6.1.0 or above.

We have prepared a video demonstrating the new features in this software: here.

If you are running a version prior to 6.1.0, please consult the release notes for that release before updating: https://github.com/complexdatacollective/Interviewer/releases/tag/v6.1.0

Changelog:

  • Include a new interface: The "name generator roster interface" is designed to supersede the previous roster interfaces and simply the experience of using roster data.
  • Automatic layout mode on the sociogram, which uses a force-directed physics simulation to automatically position nodes in real time.
  • Improved support for using CSV data in rosters. Interviewer will now attempt to correctly match the variable type to any known codebook variables, and can successfully import previous interviews as roster data.
  • A new prompt component that handles overflowing text more gracefully, and includes experimental text-to-speech support, allowing prompt text to be read aloud.
  • Automatic desaturation of categorical values where the option value is less than zero, or the string value is 'other'. This allows for styling missing or other options with less visual emphasis.
  • A new participant 'nudge' on the ego form interface, which reminds the participant to scroll down to answer further questions.
  • A new redesigned data management screen for sessions, allowing access to export options and more space to view and select sessions to delete or export.
  • Numerous other small bugfixes and improvements.

v6.1.2

3 years ago

Network Canvas Interviewer 6.1.2

iPad Users: We must sadly announce that we are unable to continue to distribute Interviewer on the public App store, due to a disagreement with Apple. A policy was applied during the review process which would require us to manually verify and approve all users of our app. We refused to do so, which has resulted in our app being rejected. We now advise all researchers to use Interviewer on alternative platforms while we explore sustainable alternatives. We continue to be able to support iPad users through a private distribution process, but it will likely only be suitable for large studies with access to IT support. Please contact us at [email protected] to explore this option, or if you have questions or concerns. We sincerely apologize for the situation.

This release contains a fix for a small bug that impacts protocols that use the '>' character in ordinal/categorical labels. It should be installed by all users currently running version 6.1.0 or above.

If you are running a version prior to 6.1.0, please consult the release notes for that release before updating: https://github.com/complexdatacollective/Interviewer/releases/tag/v6.1.0

Changelog:

  • Fixed an issue where ordinal/categorical labels that contained the '>' character would not render the character.

v6.1.1

3 years ago

Network Canvas Interviewer 6.1.1

iOS Users: We have encountered serious difficulties with Apple regarding the review process, which have held up the release of this app. At present we advise you to plan to use Interviewer on an alternative platform, or to continue to use the 6.0.3 release. Please contact us at [email protected] if you have questions or concerns.

This release contains a fix for a small bug that impacts updating an already installed protocol. It should be installed by all users currently running version 6.1.0.

If you are running a version prior to 6.1.0, please consult the release notes for that release before updating: https://github.com/complexdatacollective/Interviewer/releases/tag/v6.1.0

Changelog:

  • Fixed an issue where the app would incorrectly inform a user that there were in-progress sessions when attempting to import a protocol.

v6.1.0

3 years ago

Network Canvas Interviewer 6.1.0

This is the first significant feature update to Interviewer since our launch. These changes required a new protocol schema version, although all existing protocols will continue to function. Review the changelog below, and take extra care when updating if you are currently in the process of collecting data.

Changelog:

  • Changed the behavior of boolean variables. Boolean variables were previously assumed to be 'false' unless explicitly set to true during the interview. They are now set to 'null'. This means that skip logic or filtering rules that deal with boolean variables should be reviewed to ensure they are functioning as you intended. This change also means that boolean variables that are never explicitly set to either true or false will not be exported as node or edge attributes. If you require these variables to be exported, ensure they are set to a value.
  • Added a new interface: the "Tie-Strength Census". This interface follows the same paradigm as our dyad census, except it assigns a value to an ordinal variable if an edge is present. This can be used to capture both the strength and existence of a relationship in one interaction. It resulted from a collaboration with Dr. Brea Perry at the Indiana University Network Science Institute.
  • Introduced a new input component: "BooleanChoice". This new component is available for boolean variable types, and differs from the existing toggle switch in two ways: (1) it requires user interaction to set a value (it has no default), and (2) it supports rich labels describing each 'choice'.
  • Introduced markdown support for field labels and option values. This means that all field labels and option values for ordinal/categorical form variables can now contain limited rich text formatting, such as bold and italic text. A WYSIWYG editor has been added to Architect to make formatting these labels more straightforward.
  • Introduced new validation options: unique, differentFrom, and sameAs.
    • sameAs - Requires that a variable value be the same as another variable value.
    • differentFrom - Requires that a variable value be different from another variable value.
    • unique - Requires that a variable value be unique among all entities of the same type.
  • Fixed a regression from the previous release that broke the ability to toggle the display of some preset options.
  • Fixed a regression from the previous release that broke the ability to specify the sort order of nodes waiting to be placed on the sociogram.

v6.0.3

3 years ago

Network Canvas Interviewer 6.0.3

This is a bug fix release that addresses an issue with the Narrative interface. It is recommended for all users.

Changelog:

  • Fixed a regression that caused the Narrative interface to crash when no highlight variables were selected on a preset

v6.0.2

3 years ago

Network Canvas Interviewer 6.0.2

This is a bugfix release that resolves two issues that were identified by our users following the previous release.

Changelog:

  • Fixed a regression that caused stage level node filtering for the Sociogram and Narrative interfaces to stop working
  • Fixed an issue with entering text on some iOS devices

v6.0.1

3 years ago

Network Canvas Interviewer 6.0.1

This is a small release containing bugfixes for several issues that have been identified since the launch of the stable version.

Changelog:

  • Fixed an issue during installation where the user would have to cancel Bonjour installation even if they had indicated they did not want to install it
  • Added a variety of Linux packages
  • Fixed various console warnings and errors
  • Implemented the ability to cancel edge creation on the Sociogram interface by clicking/tapping the first selected node
  • Changed the behavior of the filter and select functions of the data export section to be more logical
  • Fixed an issue where data exported from protocols without ego specific stages could contain empty GraphML elements

v6.0.0

3 years ago

Network Canvas Interviewer 6.0.0

This is a small release containing bugfixes for several issues that have been identified since the launch of the stable version.

Changelog:

  • Fixed an issue during installation where the user would have to cancel Bonjour installation even if they had indicated they did not want to install it
  • Added the ability to install the app on Linux using the snap store, as well as providing AppImage and deb packages
  • Fixed various console warnings and errors
  • Implemented the ability to cancel edge creation on the Sociogram interface by clicking/tapping the first selected node
  • Changed the behavior of the filter and select functions of the data export section to be more logical
  • Fixed an issue where data exported from protocols without ego specific stages could contain empty GraphML elements

v5.1.0

4 years ago

Welcome to Network Canvas 5.1.0

This release includes the following fixes and improvements:

  • Support for four new input types: Slider, DatePicker, Relative Date Picker, and Text Area.
    • Slider: An input that allows you to implement a Visual Analog Scale (VAS) on a form.
    • DatePicker: This input allows you to capture data information in various resolutions (Year/Month/Day, Year/Month, Year).
    • Relative Date Picker: As above, except automatically bounded by distance from an "anchor date". Allows questions such as "Within the past 6 weeks...".
    • Text Area: A text input for longer format text entry.
  • Support for protocol schema version 3.
  • Fixes for rendering issues that were impacting Android devices.
  • Changed the behavior of the finish interview screen to not default to deleting the interview session on finish.

Known Issues

  • On the ordinal bin interface with devices running Android, individual bins cannot be scrolled and there are rendering glitches. We are investigating a fix targeted for the next release.
  • Switching between protocols on the start screen is stuttery on iOS. We are investigating the cause of this issue.
  • When scrolling on Android devices, scrolling momentum sometimes abruptly stops/reverses. We are investigating the cause of this issue.
  • Information interfaces using video may not be working on Android devices. We are investigating the cause of this issue.
  • Loading CSV formatted data performs poorly, particularly on mobile devices. This operation is CPU intensive, and causes devices to stall, sometimes for multiple seconds.
    • Strategies for mediation: Use alternative data formats, and/or minimise the size of external data files. Use more powerful devices. Use the Name Generator Roster (Large) interface, which performs slightly better.
    • Long term resolution: Parse external data when session starts (potentially to temporary storage), and stream it into memory on stage load.
  • The Name Generator Roster (Search) interface does not work well on devices with software keyboards.
    • Strategies for mediation: Use alternative platforms if this feature is important for your research.
    • Long term resolution: Adjust the styling to better work within the constraints of devices with limited vertical space.
  • Large protocols may cause the app to crash during import on iOS.
    • Strategies for mediation: Do not include large assets in protocols. Avoid using iOS if large assets are essential.
    • Long term resolution: Attempt to implement unzipping within a stream.

This app is part of a free, open-source suite of tools for the collection of social network data developed through the Complex Data Collective, a registered not-for-profit, and funded by the National Institutes of Health (R01 DA042711). Network Canvas is a collaboration between Northwestern University and the University of Oxford managed out of Northwestern’s Institute for Sexual and Gender Minority Health and Wellbeing.

For documentation, further information about the project, and download links for the other apps in the suite, visit https://networkcanvas.com.

Please also support this project by providing feedback! To contribute feedback, email our project team at [email protected].