Costings Reproducibility

[rbroth]

User story:

I am a local government official. I have used the MAPS tool to create a costings proposal for a food fortification intervention and submitted the proposal to the government. Two years later, they come back to me and ask how I got the numbers.

How to solve?

1. Store population, GDP data etc. for past years and allow the user to select a start year and recalculate the intervention for an arbitrary year
2. Only use the most recent data for calculations. If the user looks at a saved intervention, it's automatically calculated using current data. The user will have to export a summary of the costings model and save it themselves for reproducibility. <-- easier for us to code, probably acceptable to @LouiseAnder

[LouiseAnder]

cc @bgsandan - relates to data integrity discussions

[rbroth]

@kpitten

[LouiseAnder]

With point 2, I think it is important for us as the creators of the tool, as well as for the user of the tool, to be able to see the changes made to any saved version, whether by them or automatically by an update being run by the tool.

Agree with @kpitten that we don't want users to come back and (accidentally or intentionally) create a new scenario 18 months after the first and be able to view this in the tool alongside an old one as this would be like apples & oranges, given different populations and IMPACT data being drawn upon. This suggests that it's important for the tool to update scenarios to current year = year 0 when a model is opened/saved, but that we need to alert the user this has been done. Pop up? @pautva

Also allow them to see the data detail of a previous fully saved model for reproducibility and data integrity reasons. Perhaps this could be minimally formatted (not editable) if space-saving is of importance?

[pautva]

When saved intervention(s) are loaded we could inform people what is the intervention start year:

And to make it more explicit in the UI itself, we could include it both on the intervention summary and details pages.

Changing the 'Start year' makes it more complicated. We could add it to our 'Feature request list' and come back to it later.

[LouiseAnder]

thanks @pautva, I like the pop up and the persistent reminder of the start year!

any thoughts @kpitten?

[kpitten]

I also like the pop up window and persistent reminders, though I'm wondering if the label should instead be "base year" or something like that, since "start" kind of implies an intervention is not already started, which will not be the case for many interventions? @LouiseAnder @pautva

[pautva]

"Base year" seems to be more accurate. We could provide a short explanation in the tooltip:

[kpitten]

@pautva Great. The explanation could be something like "The base year is the first year of the 10-year time horizon over which intervention costs and/or effectiveness will be modeled."