Also, to your point about Pareto – Forrester uses “Wave” methodology – http://www.forrester.com/wave – though technically not a 2×2, it represents a good visual of data that could/would probably be shown in 2×2 if Gartner were to do it.

I do have a list of visual thinking topics I want to cover, interleaved with my queue of less philosophical pieces.

Mind maps/bubble charts etc. interest me a lot, but not in the usual way. People like Tony Buzan have tried to codify and “teach” mindmapping as a discipline, which I think is misguided (not to attempt laying claim to ‘inventing’ the thing, which makes about as much sense as claiming to have ‘invented’ breathing). I have been sort of groping towards my ‘angle’ on the subject, but haven’t gotten there yet. I don’t like to blog a topic until I find a differentiating angle. That’s the ‘refactored perception’ in my blog subtitle after all.

(For convenience of discussing without images, all grids discussed below are assumed to be plotted as two points each on the X and Y axis growing on the right and up respectively. So we have X1, X2, Y1, Y2 where X2-Y2 is the usual High-High on the top right.

Incidentally I think most grids are this way though some show the Y axis growing downward, whereas your discussion started with drawing intersecting lines creating four squares. Interesting to reflect whether this affects the analysis in cases where we have the choice of representation. In Low-High cases, the positive graph is the obvious correct depiction but in Is-Is Not cases we have choice.)

High-High does not *always* represent the preferred or optimal zone. Two of the well-established grids do not proceed from X1-Y1 to X2-Y2.

In the situational leadership grid plotting low-high of directive and supportive behavior, the progression is from X2-Y1 (quadrant labeled Directing) to X1-Y1 (quadrant labeled Delegating). One of many examples can be seen here.

In the learning stages grid the optimal is to move from X1-Y1 to X2-Y1 and not X2-Y2, which is clearly an intermediate stage (conscious competence). While the typical image of this is like the one here, my suggestion is to look at the brilliant improvement (with evocative names) at Conversaction where there are other interesting snippets on visual thinking.

Visual thinking, which you have written about earlier, deserves many more posts as I think it is an emerging new field by itself. One interesting effort is WikIT. Though it currently seems biased towards mind maps, I found useful material such as the comparison between mind maps and concept maps.

Your post talks of the “right” pair of parameters for the axes. That is a clinching criterion for the utility of a quadrant diagram. Unfortunately we cannot always easily determine whether there is a better alternative, or if some other important dimension is left out, though we may be able to find fault (e.g. inter-related, less relevant) with some pairs. The problem is the difference between design thinking and analysis thinking. The mind can more readily compare something to known patterns and categorize/label it than identify something missing. In a way, this probably explains the popularity of anything presented in quadrants (especially with names like analytical framework and mental model) and also hints at why some simplistic two-dimensional grid-ifying leaves us dissatisfied.

I think you are also right about familiarity and juxtaposition.

One thing that occurs to me that needs some thought is that in the common low/hi+low/hi symmetrical case, the top right hi-hi quadrant always represents the preferred “optimal” zone of performance, as in the Schwarz/Loehr case. This immediately suggests that there is always a Pareto boundary in such cases, which makes the hi/hi quadrant hard to be stable in.

And yes; shorter than usual. Am trying to see if I can drive my average down

I was surprised to find no mention of many popular grids, including two organizations making a lot of money for years using 2X2 grids: Grid International and Gartner.

The managerial grid was a useful and simple way to get managers in the 1960s to balance between task-orientation and people-orientation. Now they conduct five-day management programmes around it!?

While Gartner’s use of plotting vision completeness and ability to execute, along with names for each quadrant, may fulfil usefulness criteria you have listed, the only “magic” seems to be how they have “milked this cash cow” (to use a metaphor from another famous old grid, the BCG Growth-share matrix), as a “proprietary research tool”.

From the above two, one could formulate (ahem, readers, you heard this here first) Grid Principle #1: Grids generate serious moolah!

An interesting-looking parody on grids is at http://www.valleyofthegeeks.com/News/GartnerQuadrant.html

Part of its ubiquity and popularity in the softer areas involving open-ended conceptual analysis is its sheer simplicity and ease of drawing. The 2X2 grid pampers to our tendency for either-or binary thinking. By drawing attention on two such dualities and attempting to bring a mini Aha when we look at the four combinations, it is able to create a feeling of gaining new insight to a lot of people.

The Johari window has also survived as a cognitive psychology tool to analyze one’s openness and interaction with other people.

The four stages of learning a skill is often depicted in the form of a 2X2 grid plotting unconscious/conscious and incompetence/competence.

Every Time Management 101 class or book uses an urgency-importance 2X2 grid, now made even more famous by Covey of 7 Habits fame. I can never remember whether I should worry about Quadrant III or Quadrant II because every time I see someone draw it, they plot and number the grid differently

The problem with this grid is that two of the quadrants (high-high and low-low) warrant no attention or explanation. Grids where one of the quadrants is “not applicable” are the worst.

I believe I have encountered 3X3 grids (“tables”? And almost never 2X3 or 3X2–why?) but they seem to cross the limit of simplicity for most people and haven’t endured.

Compared to the high-low of Gartner’s or managerial grid, Johari window and the learning stages grid use absolute opposites, like present/absent. This does seem to favor the grid treatment, though I would add, especially or only when all four quadrants have meaning. Arguable, is/is-not could be seen as an extreme version of high/low.

While high/low could be a warning signal, that may still be a part of a useful grid, one example being some of the variants of situational leadership models.

Methinks, it is optional for evocative naming of quadrants for the grid to have value–the learning stages grid is an example. Johari would be illuminating even without the names it uses like Arena and Facade.

To summarize my list of criteria for useful grids:
1. The axes should be independent parameters
2. All four quadrants should be applicable
3. The mini Aha generated by considering the four combinations should lead to distinct implications for each of the quadrants

In general, if someone uses a grid to aspects already analyzed and understood, it may prove less popular than if a grid happens to be used to focus attention on two important parameters and considering them in conjunction. In most of the famous grid examples above, the latter seems to be the case.