Hello. I'm new-ish to both Tableau and to this forum. I'm looking for the best way to display a specific set of data but first I need to explain a little background about what I'm attempting to do.
I'm in the process of building a Management of Change (MOC) system for my employer. As a company we have 5 different "types" (think categories) of changes:
- Equipment
- Chemical/Hazardous Material
- Process
- Procedure
- Other
Our MOC program has approx. 18 checklists that are used to evaluate the impacts of each one of these changes; and each of these checklists are discipline-based (think one for accounting, one for engineering, one for HR, etc.). I sent out a spreadsheet (similar to the one I attached) to all of my stakeholders and asked for their feedback on which checklists should be "default" for each type of change. My goal is to build a "template" for each category of change so that our MOC software will automatically select the proper evaluation checklists once the user has picked the category of change.
When I sent out the spreadsheet to the stakeholders with the current default checklists "X'd"...I asked them to either add a green X where they think a certain checklist should be added to a template or to put a strikethrough red X where they think a checklist should be taken away from a template. The data I wish to display is where disagreements exists between Stakeholders and maybe even "how much" disagreement. My thought process is that in areas where most or all of the stakeholders agree to the current default, we probably got it right; but, where there exists some "threshold" of disagreement, we need to have further discussions.
Anyone have any ideas about some creative and/or best ways to display this data? Hopefully I've given enough information but if not, please feel free to ask. Thanks in advance for your help.
The first thing you might do is change the values. In the future you might change the requested responses to make your life a bit easier Strikethrough text isn't actually a different character than an X (at least I don't think ASCII/Unicode recognizes the default strikethrough as such), nor is a different color X actually different to Tableau. So we need to change the underlying text before we bring this in or everything will just be imported as an X.
That being said, you might do a search on visualizing survey data, as that is more or less what you're doing (albeit without the typical Likert Scale. In a way, though, your responses are in 4 buckets:
- Original Default (almost like a 'Somewhat Agree' response)
- Original Non-Default (these are the null/empty values - almost like a 'Somewhat Disagree')
- Added as Default (almost like 'Strongly Agree')
- Removed by Default (almost like 'Strongly Disagree')
There are obviously some cases where the above metaphor/correlation breaks down, but maybe it's a good start. Treating data in this way, I've mocked up a few options for showing this in simple bars - sometimes treating anything that adjusts the defaults as the "negative" or "disagreeing" options, and doing the opposite for the responses that leave the defaults in place. Other times, I treated any "remove by default" OR "original non-default" (I now realize this could be named a bit better) as the "negatives". It's up to you in what message you are trying to convey, and you may want to swap out a few of the colors for one another, or move away from the red/green palette altogether. No matter the case, either the bars which have equal red/green values are the most dissenting, or the ones which have bars going left vs right are what we would consider different sentiments. I wrote a couple calculations to quantify this level of agreement. 0% agreement would indicate as many people voted for something vs. against something. 100% would mean everyone either voted for or against.
Hopefully the attached workbook helps get you in the right direction.
