Devil is in the details
Reading lengths: 15min;
Method: Affinity Clustering;
Content: Experiance; Reflections
Complexity: Easy;
Keywords: #clustering; #grouping; #tendencies; #sets and subsets; #characteristics;
#categorization; #unique themes; #games
The affinity clustering method or the task of ordering and reordering post-it notes might sound monotonous and boring. Something that is not very inspiring to do. Maybe, but only if you ignore all the creative potential within it. If you look at this activity as a technical task with a sole purpose – rearranging the post-it notes - I agree it will be pretty mundane. However, if you face this task with the prospect of discovering something new, it already feels more inspiring. It really can become as enjoyable as playing detectives. "Devil is in the details," they say, and this method proves the phrase.
Later in this article, I'll give two examples of how this method can boost creativity
and make new discoveries. The first example will outline the game you can play
with children to have joint entertainment and encourage them to think outside
the box. The second example will explain using this method to find connections
and essential themes in an unsorted data pool.
But, before we get to examples, here is a short description of the affinity mapping method for readers who haven't read the detailed method description article.
Method in short
In a few words, affinity mapping (also called grouping and clustering) is a method used to find similarities or connections in a set of diverse opportunities or options. Each identified similarity or connection acts as a combining factor to consolidate influenced options in one group.
>Why do we need it? We, as humans, can't split our attention and remember multiple things simultaneously. Scientists have discovered that people, on average, can remember seven things at once. And it becomes a problem if you finish the creative workshop and have one hundred different ideas in front of you. You can't remember them all. You can't evaluate them all at once and choose the subset for subsequent actions. Moreover, some of these ideas will overlap, but some will be heading in different directions. What you need, in fact, is a good approach for structuring these ideas and understanding the unique themes. And affinity mapping is an excellent method that will bring this structure in. Bonus: you get a chance to practice your imagination and creativity.
First example: playing the game with children
Who doesn't want their kids to be innovative, think outside the box, and be equipped with creativity skills further in their lives? Well, of course, everyone does! Do you also know that absolutely everyone can be creative, that creativity is like a muscle that gets stronger when you practice it? Trust me or google, but it's true. Therefore giving our kids another chance to train this "muscle" is the best we can do for our children. Here is an example of how.
Setting the stage
Choose the number of different options you will involve in the game. It can be ten, fifteen, or whatever you choose.
Next, decide what will be "options" in your game. It can be different things like pictures, toys, Lego bricks, etc.
Then place the chosen number of options in the play area. For example, I've used fifteen animal-printed play cards. I've selected cards with elephant, lion, snake, wolf, horse, dog, cat, penguin, bear, fox, giraffe, monkey, goat, reindeer, tiger.
Starting the game
-
Come up with the grouping rule. Assess if and how each option corresponds to
the rule and, based on that, create smaller groups. For example, we choose
to group animals by size. We made the first group of smaller animals - snake,
dog, cat, penguin, fox, monkey, goat, and wolf, and another group with bigger
creatures - an elephant, lion, horse, bear, giraffe, reindeer, and tiger
in the other group.
Note that you can come up with any rule you can imagine. Alternatively, it could be the color of animals, animals that bite and that don't, animals that move around using four legs and those that don't need four legs, animals with antlers and those without, etc. You can also use the role to make more than two groups. All possibilities are open and depend only on your imagination. -
Now pick one of the groups and ask children to look for more similarities and
connections between various group members. What other characteristics do
some of them have in common? How can these be rearranged into smaller groups?
In our example, we choose to follow the four legs criteria. We split our first group into two smaller ones. One with animals using four legs to move around - dog, cat, fox, goat, and wolf. Another with animals moving around differently - a snake, penguin, and monkey. -
Now look at all groups and decide, if to continue splitting them further in the smallers groups.
You can continue in this manner until you are out of ideas or have groups you can't really split further. - And then start over from the beginning. Now invite your children to come up with some criteria. Let them come up with interesting ideas from the very beginning.
- Or, even better, they can do the grouping without loudly expressing the criteria. And then - you try to guess it. It's fascinating to see what sort of things our children can come up with.
Professional example – understanding essential themes in interview results
Grouping is not only fun but also valuable in professional activities. Further
in this section, I'll give one of the examples where affinity mapping has helped
me.
The context
I've conducted multiple interviews with people regarding one of my research questions - patterns of providing and consuming product or services feedback. I needed to understand what motivates people to give feedback or quite the opposite – what demotivates them, what are good and bad practices, what are the tools they use, etc.
Once I've completed the interviews, I had quite a list of opinions and habits people told me about. As you might already guess - the list and diversity of insights were quite extensive, so none of the trends clearly stood out. Yet, I used this method to normalize the information since I know affinity mapping.
Looking for the proper grouping criteria
Option 1
I could start with what seems the most apparent – grouping all the insights around the questions I've been asking. But this quickly would prove itself a useless approach. Since prepared questions are more a guiding, not limiting, the experiences people are talking about also often tend to shift from the initial question. It's ok since I'm still interested in experiences. Yet it doesn't help for structuring.
For example, some people talked about their motivation for leaving feedback when I've asked if they do it at all, while others - only when asked explicitly about motivators. So from this perspective, grouping by question doesn't often give expected results.
Option 2
The second option, obviously, would be to group by similar themes. Like insights describing emotional aspects, routines of leaving feedback, tools they are using, etc.
This approach gives more specific results. Yet, in my case, this also is not ideal.
Since themes are generic and overarching, there are a lot of insights I will include in these groups. Since I was looking for unique ideas and themes, I needed more detailed groups and information.
In the article title picture, you can see it visually in the middle example. Yes, it's structured, but it doesn't fit my purpose. I can't clearly get the information I'm looking for.
Option 3
I continued the grouping by splitting each of the bigger groups into subgroups.
I've set a limitation of a max of five ideas per group as a guiding principle and started reorganizing post-its again. Comparing one by one and evaluating which one fits best with others. By the way, limited clustering was the exercise when the grouping turned from a monotones activity to a pretty creative one. Since you not only read but try to find different forms of connections behind the words.
Yes, this took me considerably more time to do, but it was worth it.
I've got different products that people use to provide or read feedback, some known but some not. I've got multiple sub-groups hinting at what motivates people to give feedback and multiple sub-groups with hints to what they dislike in current practices. I've got sub-groups with positive surprises and sub-groups with annoying limitations.
And here, the mission is accomplished. The information I have in front of me is now normalized, clustered, and providing me precisely the level of understanding I was looking for.
Conclusion
Grouping and clustering diverse ideas are essential for understanding unique themes and hints. We, as humans, can't remember more than ten things max, so one hundred options for sure is not operatable.
To work with such an amount of information, we have to normalize it first. We have to structure and shape the result in such a form that makes sense for our unique purpose.
Grouping can be tedious if you look at it from a monotonous perspective. Or it can be really creative and boost your imagination.
Did you like this method and the description of it?
Do give me a note about it.
I will appreciate it if you include in your message what you did like, what you think should be improved, and of course, any other ideas that would be worth experimenting with.
Maybe you want to join me in some of the method experiments?
Do you have interest or experience in this method usage?
I would love to hear your example. Let me know.