• Global Pulse HunchWorks Project: Week 3 Challenge

    Imagine you're a researcher working with farmers in Uganda. Your years of experience tell you that something is not right this year. It's not unusual for the harvest to be a week late, but this year, it’s three weeks later than usual. If only there was a way for you to share this hunch and investigate it with your peers.

    This is where HunchWorks comes in…

    This week we begin sketching out some ideas for some interface screens for the creation and management of hunches. To help with this exercise, we’ve created a few scenarios to help us put some context around how people may use the tool.

    Beginning of hunch…

    Nema, a Global Pulse lab developer in Uganda, submits evidence in the form of a picture of a lake 20 feet below normal.

    Collaboration yields more evidence and more confidence in the hunch…

    John posts the hunch: Citizens are starving in Uganda because there is a drought and no irrigation systems available to water crops.

    Further collaboration builds on the evidence and confidence…

    Nema’s contribution spurs UN analyst, Patience, to gather more evidence.

    Other agencies act on this hunch to help alleviate the problem…

    In an effort to address growing concerns with dehydration, UNICEF begins shipping high volumes of water to the area. A UNICEF member posts this response as evidence to support the hunch.

    Which in turn spurs the hunch to further evolve..

    However, additional hospital reports from Patience continue to describe unusually high rates of dehydration and starvation. This reduces confidence in the UNICEF action evidence.

    What makes up a hunch?

    A hunch is an anecdote, musing, observed connection, or hypothesis which may or may not have evidence or research backing it. Hunches are claims about something.

    Evidence is something that supports the claim made by the hunch. It can be data, a photo, a tweet, or a report contributed by people working on the hunch. The purpose of evidence is to add credibility to the hunch and to enrich the discourse surrounding the hunch.

    Confidence is a measure of how the collaborators and author of the hunch view its apparent accuracy. It increases with the number of contributors as well. The more credible the hunch, the more likely people should be to take action, whether that means research, aid, or preventative measures.

    The challenge

    What does a hunch actually look like? What are the key aspects of a hunch?

    How does a hunch begin, evolve, become verified or ever cluster with other hunches to become a new hunch?

    As a hunch changes over time, how is that reflected in the interface?

    We hope this gives you some insight as to what a hunch could look like. Here are a few more resources to get you started:

    Wireframe outlines (ZIP) – 11”x8.5” and 17”x11” – some sketching templates we often use

    Hunch scenario (PDF) – A pdf with the scenarios and some basic sketching templates included

    Feel free to post your thoughts and links to your ideas and sketches (from Flickr, Instagram, Dropbox, Google Docs, TwitPic, etc.) in the comments below or alternatively, email them to todd (at) adaptivepath (dot) com. All the thinking and output generated in this project will be publicly available and Creative Commons licensed, and you can provide your submitted designs under a Attribution, Share Alike (CC-BY SA) license.

    The best ideas will be included in the final set of recommendations.

    There are 8 thoughts on this idea

    1. Danny W

      I wonder if a seed company who donates seed would be interested. It would make sure that their seed is being used effectively, and therefore bolster their Corporate & Social Responsibility metrics?

    2. Raffaella Roviglioni

      I love the way you’re using storytelling and scenarios to focus on users, on possible interfaces and on the objective for this issue.

      By the way, it has a lot in common with typical storyboards used in the past decades with farmers in the developing countries for health, agriculture and other rural trainings.

      I’m a UX consultant currently working for the FAO (UN organization for food and agriculture) and I like to see that this UX/UN connection has a future, and a bright one! 🙂

    3. Jo Anne Wright

      What are the uses of the hunch information beyond adding supporting or detracting information?
      Can anyone start one?
      If anyone can start one, is there a moderator to prevent “black hat” uses?
      Are all hunches equal?
      Are all contributors equal?

    4. Todd Elliott

      Jo Anne – These are all great questions! Developing trust is an important element to the service. As part of that, we imagine different contributor levels to help address these issues.

    5. Nathaniel Calhoun

      As you work out your definitions you might find that hunches fall into two categories: predictive hunches and causal hunches.

      1. Predictive Hunches (forward looking)

      These sorts of hunches seek corroborating evidence for something that may be happening. They are observational and they seek other observations. They don’t need to address causation and they don’t need to be weighed down by theories/hypothesis.

      It would be natural for people to respond to a predictive hunch by generating related/clustered causal hunches.

      2. Causal Hunches

      These sorts of hunches posit a causal relationship and earn their proof in conventional ways.

      A causal hunch would give rise to predictive hunches, which would then become evidence for the original causal hunch.


      An advantage to creating a few types of hunches is that when you are visually displaying related hunches, it could be clear from colors/icons which hunches are playing a supporting role and what sort of role they provide.

    6. Todd Elliott

      Nathaniel – that’s a really interesting idea, thank you for sharing!

    7. Janice

      I also like the storyboard/storytelling approach for this explanation. I can follow something so much easier that way than just reading it. I really like this idea of making potentially very important hunches more public, to turn into action. Jo Anne brings up great questions though that would have to be hammered out for this type of user interface to be effective.

    8. Susan

      This hunch user interface seems like a really effective way to turn anecdotal observations into successful action. Very cool!

    Comments are closed.

  • Close
    Team Profile