search

HWR-llc / ett

MassBays Phase I data explorer for data from the EPA WQX Portal
MIT License
1 stars 1 forks source link

Confusing habitat graph #70

Closed walkerjeffd closed 2 years ago

walkerjeffd commented 2 years ago

Can't tell what's happening here:

image

There's a target of zero AND no data?

HWR-llc commented 2 years ago

There are a number of these. Danvers River is an example of an estuarine assessment area (i.e., it should have a target) where MassBays has not assigned a target yet. My understanding on leaving the call yesterday was the same as it was two weeks ago when we discussed this with them. They are going to determine the goals for this group of assessment areas before the next meeting.

Having said that, I believe that a goal of 0 and no data to graph is still a possible/likely outcome for some of the assessment areas.

walkerjeffd commented 2 years ago

So is the goal actually 0 (implying we don't want ANY eelgrass here) or does 0 mean there is no goal? I assume it's the latter, and I feel like that is an important distinction. If there is no goal, I think it would be better to not show a line, and instead have a message saying something like "no goal has been set for this assessment area".

Also, I think we need some kind of message saying "no data available" on or below the graph, otherwise it looks like something is broken. That was my first assumption when I saw this.

HWR-llc commented 2 years ago

I believe it would mean the goal is 0, not that there is no goal. MassBays is drawing a distinction between inter-estuarine areas where there is no goal because they have not done analysis yet and estuarine areas where they have done analysis and concluded that the goal is 0 because eelgrass will not grow in the area.

I think that it looks like a negative goal on the ETT, like: "we are trying to eliminate eelgrass from this area". I don't know think that is their intent, but we've sent email summaries highlighting this issue and not gotten a response. Or at least I have not seen a response.

My read on the overall approach is that there are holes in the logic and framework they are using for the habitat analysis. ETT is the first time they are getting to explore the dataset they have created, and I think that is useful to them. So I'm inclined to show what they give us, and have them figure out the issues on their end. That makes the website look unfinished, and that is sub-optimal, but I think it is the best option for now.

Once I get the final-final-final set of goals, I'll re-evaluate the situation of a 0 goal.

walkerjeffd commented 2 years ago

Ok, feels like this is a bit of a sticky situation. I hear what you're saying about not having sufficient guidance. I've been in the same spot many times before, and its not easy when you need some kind of direction on how things should look when the data are a bit ambiguous.

In cases like this, I usually default to thinking about the end user, someone who is new to the site and has some interest in exploring the data. What would they think if they saw this graph? I would try to make that audience my main focus, as opposed to the client (MassBays) or us (developers) who may know (a priori) what a goal of 0 actually means and why there is no data on the graph.

Ideally, MassBays would have some idea how to communicate this kind of situation (no data/no goal). But expecting them to come up with the solution is asking a lot as they are not experts in web design/data visualization, and they don't have a lot of experience in thinking about their data from the perspective of a new user. The hardest part of this job is often figuring out how to translate the holes in the logic and framework into some kind of information that informs the new user so they understand why something is the way it is, especially when it can be confusing at first glance.

Anyway, just some pearls of wisdom from an over-tired father of two toddlers who is up way past his bedtime... I'll leave it up to you to sort out the solution when the final set of goals comes through.

HWR-llc commented 2 years ago

I think this is all set now. The lack of a goal needed to be addressed, but the lack of data was a mistake that I had made. I assumed areas with no evidence of a particular habitat type had no data. That was wrong... mostly. Almost all of the survey datasets we are using are statewide. In that case, no evidence of a habitat type should be assumed to be 0. The only exception to that for us is the 2006-2007 eelgrass survey.

Changing all of these no data to 0 means that we should never have a habitat graph with no data. There are many graphs that are only 0s, but that makes much more intuitive sense and is easily understandable by the viewer (IMO). So I have fixed the dataset and added the note that says "No goal established yet."