Last month I started thinking about slicing, my instinctive approach to looking for perspectives on a problem, an opinion, an observation, or anything else. This time around I've got an example to talk about.
On Fridays I ensemble with a group of medical quality engineers and medical knowledge engineers. We learn from each other, about testing and about the domain. On and off recently we've looked at a project of theirs which aims to understand better what work they do, how they do it, and why it's that way, and then write it up for internal and external consumption.
In one early session, with a wider group in the company, there was an extremely open and exciting conversation about what should be covered in this effort.
It was the kind of discussion that greenfield projects often have, before scope is nailed down, where the world seems ripe with possibility, no difficulties have been identified, and there is no talk of who will taking responsibility for the implementation.
Partway through I was asked to facilitate so I shared a mind map I'd begun to make while people were talking. I then opened the map to everyone to add their own ideas, encouraging them to actually do it, and reassuring them that it can be helpful, when we don't understand the scope, to get down as much as possible that seems plausible and then choose.
That data was deep and broad but lacked organisation and cohesion, so after the meeting I spent a while arranging it. Mostly this consisted of taking similar ideas and clustering them, and adding categories and sub-categories as I felt they made sense.
I did this on a copy of the original map to make comparison possible, help tell the story of what will be a medium-term project, and give an informal audit trail.
In the next session, a few days later, a group of three of us looked for a way to extract some concrete scoping proposals from the sorted mind map. This is a slicing problem: where do you put the knife? Categorisations help at this point, because they provide natural edges to cut along.
We had created a category containing possible audiences for the report or reports. It consisted of three entries, which seemed like a manageable number of slices, so we started there.
For each audience type my colleagues provided a couple of areas of interest, and we annotated the map to show which nodes would be relevant to each topic, producing six slices through our data.
In my previous post, trying to think about how I do this, I ended with these words:
... collect, arrange, slice. I don't know quite what I mean by it yet ...
I'm still not sure they are the right terms, or everything that I do, but you can see that structure in the work we did on this project: collect by casting the net wide, arrange into structures that expose some potentially useful seams, slice down the seams.
I feel like this is obvious but perhaps that's just because it's what I've learned to do. I'm in reasonable company thinking about it, though. Hillel Wayne wrote Collecting and Curating Material is Good and we Should do it More a few days after my last post.
He breaks things down slightly differently to cover his research process but there's clear similarity:
- Collection: gathering material that’s out there and putting it in one place.
- Curation: identifying which gathered material is useful for knowledge-building.
- Analysis: taking the curated material, breaking them down, and studying what they’re "saying".
- Synthesis: taking the analytic information and processing it into an overall idea.
Note that on our project as I've described it so far we're not doing "the work" yet. Rather, we're trying to choose which work to do and how to do it, in a proportionate way.
Although in the retelling it feels linear, this process is exploratory. At each step we took a perspective and tried it, judging how far to go before switching to another approach. The skill, as with exploratory testing, is to try something, observe something, conclude something, and repeat for as long as is reasonable given the constraints. There's no general formula for that.
Collect, arrange, slice itself looks linear when written but don't be fooled. You can collect a little, arrange a little, collect some more, slice some, see whether that seems productive, rearrange, slice again, collect again. Or anything else: whatever seems like it will best serve your mission.
Writing all this down has moved my thinking along a little and I've collected more stuff for later arrangement:
- Writing blog posts is often collect, arrange, slice.
- In this case of this post, I rewrote it three times until I was happy with the slice.
- Fieldstones involve collection and arrangement.
- Arrange the work to get to a conclusion (of some kind) in the time available.
- This may mean renegotiating the mission along the way.
- Perhaps this is slicing the mission.
- Slices can be thicker or thinner. Breadth and depth.
- Might a slice create a pivot point, like in a pivot table?
- The process is recursive, naturally.
- We are trying to get to a point where there is something that can be attacked directly.
- Mnemonics like SFIDPOT are pre-sliced arrangements.
- They can be helpful to bootstrap collection ...
- ... or to just abbreviate the whole process if time is important.
- Wide reading or collaboration with others of diverse expertise expands the collection and arrangement possibilities.
- It's OK for there to be contradictory data in the collection and arrangement ...
- ... we're not necessarily trying to make a single model of a domain ...
- ... we're trying to make a helpful model of our thoughts about the problem.
That'll do for today.
Image: https://flic.kr/p/qYXP6A