How can you make a design decisions based on "future behavior" of users?

Question

I am working on designing of mobile application that will be used by field nurses in a developing country. The application guides the nurse through a series of questions and then gives her a diagnosis and treatment advice. The app is part of a research project and will be used by a limited amount of nurses. I have a major decision to make:

1) Put each question on a separate screen

2) Put similar questions in a group on a single screen, and have much fewer screens.

The advantage of the first decision is that it's much clearer to the user what needs to be answered to move to the next screen. However, in the long run (for an expert user) it seems to me that it will be very daunting to flip through 20-30 question screens, especially in time-constrained setting. Therefore, I'm tending to the second solution.

How can I verify this assumption (and decision)? It seems to me that if I conduct a comparative usability testing now, there is certain chance that as a first-time users, the nurses will prefer the single-question-per-screen version. However, I know that time is major factor for them, and am afraid they will skip using the application in the long run.

Answer 1

Personally, I would choose individual cards. There are multiple reasons for why.

As users become more expert, they develop some mechanical memory of where to tap for a particular answer, assuming they come across the same starting point of the decision tree often. This is both a weakness and a strength.

It is strong, because they will be faster at using it. Also, when scrolling you might accidentally scroll past a question. Not so if they are on individual screens.

Its weakness is that if the order of the questions changes or the answers are laid out differently, the user may make a mistake.

So always have an easy way to correct that mistake (e.g. "back") and an "always there" way to review what has been recorded so far (e.g. "view answers so far"). It is good practice to allow jumping to an answer to change it.

There are three things to consider:

1 - Are any question dependent on answers from previous questions? If they are, you will need to alter the questions further down the page to reflect that higher answer. Instead of enforcing a fill order on a single page, which can be annoying, it would be easier to split questions into cards. . e.g. "patient gender" will impact showing "could you be pregnant".

2 - Do you have triage questions at the start, in batches of critical importance? e.g. "visibly bleeding" is a triage question, and should be asked on a first screen, as it should bypass all others and move on to emergency care. Prioritising triage questions would be a challenge, and I'd expect this to be one card, which auto-forwards to "go do this thing" if a very critical one is checked. Can't imagine still looking at the device if someone with a bone sticking out has just walked in though...

3 - Do you need to add explanatory information alongside a question? If so, it will be easier to do when following the one question per page rule, because saving progress will be easier, and there will also be more space for that content or triggers to that content.

(please disregard any medical-sounding information, as I am not a doctor and do not know what the levels of triage are. I just know they exist in all areas of life.)