Best Way to Digitize Table for Quick Data Collection- Survey123 or Other?

[KatMaloneyODEQ]

Hi all,

My agency is working on transitioning to Survey123 and other products to digitize our data collection process. I am working on a survey for one of our programs, I am trying to figure out the best way to digitize the following table used for quick data collection. Parameters are measured by group 1 (1ST) and group 2 (2nd) about every 3-5 minutes and recorded on this table. It is imperative that we are able to quickly record these 2 values accurately so we can compare them later on.

What is the best way to digitize this using an ESRI form? Is there a way to add this into a Survey using Survey123, or would it be better to create this form in Field Maps or Quick Capture (or another app)?

Thanks so much!

[DougBrowning]

Grids are about as close as you can get.  https://community.esri.com/t5/arcgis-survey123-blog/survey123-tricks-of-the-trade-groups-grids-and/b...

Yours may be a bit too many fields to fit but you could try some variations.

[abureaux]

I have a similar chart for GW monitoring... Unfortunately, a linear list is the best solution I could find.

Here is an example from one of the low-flow sample sheets. Likely similar to what you are doing. The parameters are select-able on the previous page (I only selected a few for the sake of screen shots).

[KatMaloneyODEQ]

This could work- did you use the grid functionality listed above? The data we will be collecting with this form are also groundwater data.

Thanks!

[abureaux]

Not quite. I am not a fan of the grid function (at this time). Too messy IMO (you can't hide the grid lines at all).

Fun fact about the grid implementation though! You can actually use the syntax in non-grid surveys. You are limited to columns of identical width, and can only add columns to groups, but it still works wonders. Just add w# to a group's appearance (e.g., w3 will add three columns to a group). This is what I am doing for the "Overview" section:

EDIT: It can be applied to Groups and Repeats. I forgot to mention repeats above...

[KatMaloneyODEQ]

Hi again,

Can you provide some detail as to how you achieved the Overview table below the repeat grouping? I think something like that would help our crew out in the field. Did you use the join functionality? Thanks!

[abureaux · Accepted solution]

There are two components to the Overview table. Depending on you needs, you may only need one of the two:

1. Layout. Generally speaking, this is a grid layout designed to help end users see data at a glance. Once Esri implements native table layouts (think Excel data entry for repeats), I may replace this. I don't expect this functionality any time soon though, as they haven't even announced the Maui version of S123 yet (current S123 is built on QT, which is being deprecated. S123 is moving to .NET Maui).
2. Parameter stabilization. For our purposes, we require the last 3 parameters be "stable" prior to gw sampling. This is a part of our Low-Flow SOP. I won't go in depth to our SOP, but I will show the general calculations/principal.

Layout

1. Create a group
2. Give that group an appearance of "w#", where # is the total number of columns needed. In my case, I am using "w9"
   

   NOTE: This is important. I am not using the grid survey style. In my case, I am using the "pages" survey style, but you could just as easily leave that field empty. When I say "survey style", I am referring to the "style" column in the "settings" tab of the XLSX.

3. Add to this group up to # visible fields, where # matches the value in step 2 (above). In my case, I have 9 notes.
   NOTE: I say "visible fields" because fields that are not visible don't count. E.g., I probably have >80 calculates in this group. Because they aren't visible, they don't mess anything up.
   NOTE2: If you are using the Grid style, be aware that invisible components can change the appearance. Using "dynamic-grid" should fix that issue. 
4. Optional: Use the "relevant" column to dynamically show content in the group. In my case, I allow the end user to customize which data is visible (see image below). To save time in the field, I have defaults selected, according to our SOP.

Parameter Stabilization (super general guide)

1. Our SOP requires that the last 3 samples be "stabilized" prior to sampling. Because we need to extract that information from a repeat, we will be using indexed-repeat()
   

   indexed-repeat(question, repeat, index number)

   Returns the value from a specific question in a repeat record. For more information, see Repeats.

   indexed-repeat(${room_no}, ${floor}, 3)

2.  Indexed-repeat() supports math. That means you can hard code stuff like the image below. Here, I am extracting conductivity values from my parameter repeat with different indexes.
   
3. I like to show dynamic text and colors to visually show the user that parameters are stable. I've included the calculates, the label formatting, and the final product for reference. As you can see from the label formatting, I include the title, sub-title, and all the values (gw_conductivity_group_sum)
   
   
4. To get your values to appear, you need two calculates
   1. One within the repeat
      
   2. And one outside the repeat.
      

Hopefully I remembered all the important things. If not, let me know.

As a side note, here is everything (outside the repeat) that goes into testing if a single parameter is sable. If this is something you require, I can go into more depth where needed.

[KatMaloneyODEQ]

Thank you, this is super helpful!

[LaurenceTait]

When you say you want to digitise this, do you mean you want it to look exactly like the paper form, or do you mean that you want to capture the same information.

I've attached an example of how to capture the same information.

There isn't enough detail of your work flow to know how best to do this. When you say "Parameters are measured by group 1 (1ST) and group 2 (2nd)", are groups one and two different groups of people measuring the data at different points in a process chain? Do they currently physically pass the sheet of paper between the groups, and do they need to do that with the tablet? Can you afford two tablets?

Without that sort of detail it's impossible to know what the optimal way of doing this is. It's perfectly achievable under any scenario, but the optimal way of doing it is likely to be very different from the way you are appear to be currently doing it using paper and manual entry.