We’ve heard from CHWs that they would find it useful to have their tasks organized by area or household, so that they may prioritize other tasks nearby.
We are proposing two options that may solve this design problem:
The ability to choose between sorting options where CHWs may view their tasks in the order they are due, or grouped by household alphabetically. This could potentially also solve for the open issue of sorting tasks by priority.
Adding an interstitial page after completing a task with a list of other tasks within the same household (or relevant grouping). This could potentially also solve for chaining forms.
I’m hoping to get any feedback, specifically answers to:
Which of these two options do you prefer?
Would you change anything to improve these designs?
I’m careful answering a question like “which do you prefer” as I’ve seen too often that the actual users have different preferences, and what makes matters more complex might perform better in a solution that does not correspond to the preference. I think the answer depends a bit on the digital literacy of the CHWs, but probably also on characteristics of the catchment area and service package they deliver.
The second option seems to have “less friction”/requires less explicit interaction and could thus cater better to CHWs with less digital literacy or who do not want to/can pay attention to the different sort or other configuration options. That being said, I’d ideally do some more research on this, e.g. in an A/B test over different programs (I know today this might sound like far fetched, but maybe some day we get to a stage where we can do this across multiple programs implemented by different partners).
Not necessarily an improvement, as it depends on a lot of factors that require more research, but an additional option could be to show “other tasks near this household” or even “tasks near me”. Potential negative side effects are that harder to reach households could get de-prioritized, CHV get tired more easily (the walking break between households might refresh the mind), etc.
Thanks for the reply @hhornung ! We have done some testing on these concepts across a few programs as documented on our blog and are planning on implementing the second option as V1 of this initiative. Once available and deployed we’ll be looking to monitor completion rates and be able to hopefully do some A/B testing as you mentioned .
Excited you also brought up “tasks near me”! It’s a concept we are exploring further that is still in the research phase – more to come on that in the near future.
Here’s the feedback I received from our Program team:
Task Sorting and Grouping: The approach of showing an interstitial page was well received by our program team and they believe that it will help CHVs to organize and prioritize their tasks.
Update on the proposed Action Bar: Our program team thinks that this feature will confuse CHVs because they are already used to the current layout. Also, introducing such a change will require significant amount of effort to train the CHVs.
Update on the proposed collapsible menu: This is also a kind of change which will require lots of training. Hence, our program team believes that the current meu layout suffices.
Thanks for sharing this feedback! Bigger UI changes almost always have groups that favor change and groups that prefer their current familiar UI, so your feedback on points 2 and 3 makes a lot of sense.
In the blog post Nicole referenced about user feedback sessions with PIH, the update to the action bar had the lowest positive feedback rate, with 70% of health workers preferring the new version. There was an interesting nuance in the data though that the blog post doesn’t mention - there are actually two catchment areas where we got this input, one of which has been using their tool longer and the other of which was relatively newer. In the area with more new users, almost everyone preferred the new option. By contrast where health workers had been using a CHT app the longest, just a little over half preferred the new option. So the decision is partly about weighing the preferences of people who are really familiar with the current system relative to new users who are onboarding and might require more training/learning without these changes.
Given the data we have, my sense is that the UI change is going to move forward, so it’s important for us to problem solve about how to accommodate this feedback in the way UI improvements get rolled out. One option of course is to plan an update to coincide with a training that is already happening. In some health systems there’s a monthly meeting where this could be reviewed, but in other cases it might be a quarterly or annual gathering that you plan to coincide with a version upgrade involving UI improvements.
Another opportunity we’ve been exploring has to do with using the app to deliver training materials that help remotely onboard health workers when there are feature updates, so that a new training isn’t needed for every upgrade and/or so that face-to-face trainings are easier and more efficient. There are some demos of this concept in the remote onboarding and learning and care pages on the CHT docs site. My hope is that some earlier adopters will use this kind of digital learning approach when the UI improvements go out, so we have some data on how well the training worked before we’re recommending wider adoption.
@iesmail I’d be curious to hear if either (or both) of these approaches seem relevant for your health system. And thanks again for the feedback, being aware of the training costs/hassles will really help us design the roll out effectively.
I tested the interstitial page functionality and here’s what I observed:
If a task is chosen from a household profile page and submitted, then a landing page appears (containing other tasks for the same household). However, if the same task is chosen from the tasks tab, then the landing page does not appear. Is that the expected behavior?
.