Measurements and Descriptions
There are endless ways to measure and describe the world around us.
When you boil data down to the essentials, what you get are…
- Descriptions of the measurement(s)
It doesn’t matter what industry you’re in. It doesn’t matter the process that captured the data you’re looking at. At the end of the day, you’ve got numbers and classifications as measurements, and you’ve got the context describing the who, what, when, where, why, and how of those measurements.
“Dimensional Model” - Breaking it Down
Let’s unpack the term “dimensional model” to see if we can make heads of tails of what the words mean when they’re put together in this term.
I’ll start with the noun (model) and then look at the adjective (dimensional).
What’s a model?
The first images that come to mind for me involve cars, trains, and rocket ships. When you think about the hobby of model-making, what you have is an activity where you take [something] (like a car, train, or rocket) and create [something else] (the model) that represents the original.
Let’s keep that word: represent
Another image that comes to mind is a person who is a model. You think about the man or woman walking down the runway, showing off what a particular designer’s outfit should look like on an “ideal” person.
Let’s keep that word too: ideal
The world is complicated. Sometimes, in order to shut out the noise, we create something that portrays the general concept of something that would otherwise be difficult to understand.
General concept is the last word we’ll keep for now.
So what do we have?
A model is a representation of data that portrays its general concept(s), in a way that is ideal for understanding the patterns and information that the data contains.
That last part, “ideal for understanding the patterns and information that the data contains” leaves the door open to the possibility that there might be different kinds of models that are more or less ideal for different analyses.
The first half remains true to all models. Models are representations of data. And models boil things down to the general concept(s) that exist within the data. The goal for models is to help humans see, understand, and interpret the patterns and information within data.
What does the word “dimension” being to mind for you?
I immediately thought of shapes (2 dimensional or 3 dimensional shapes).
In a 2D shape, the dimensions are length and width, aren’t they? A square or rectangle has sides that are [x] units long by [y] units wide. 3D shapes add a dimension: depth.
In the case of shapes, a dimension is [the thing] that answers the question “how long?” or “how wide?” or “how deep?”.
Outside the world of geometry, the word “dimension” carries the idea of the characteristics or qualities of something.
If I say, “When I look at the diamond from this angle, it explodes with brilliance!”, I’m substituting “from this angle” for a way to describe the diamond’s facets… its dimensionality, if you will.
If I say, “Wow, when you add jalapeños to the salsa, it adds a whole new dimension of flavor!”, you see how the word dimension gets at an essential feature of the salsa’s flavor.
Coffee from different regions of the world have distinct flavor profiles - different dimensions to their taste.
The same goes with wines and cheeses and teas and on and on with just about anything you can think of.
The word dimensional tells us that the characteristics, qualities, features, and facets of something are what’s important.
Dimensions of What??
So we’ve got a representation of data (a model). We’ve got characteristics…qualities…features…facets… (dimensions).
The question now becomes… dimensions of… what?? The answer gives us the distinguishing characteristic of dimensional models that makes them ideal for understanding the patterns and information that the data contains.
The Defining Characteristic of a Dimensional Model
The defining characteristic of a dimensional model is that it captures and highlights the characteristics, qualities, features, and facets of the event that generated the measurements.
Dimensional models make the numbers visible, true… but they also make a point to bring meaning to the numbers by making the who, what, when, where, why, and how behind the generation of those numbers into the forefront.
Qualitative data, in and of itself, is descriptive by nature. It’s the qualitative data about the situation that caused a measurement to be recorded that gives dimensional models their “ideal” edge when it comes to understanding the patterns and information that the data contains.
There are a few things that have to happen for that to be possible:
- You’ve got to define a “measurement event” – the action or process that causes you to capture and save data in the first place
- You’ve got to detect when that event has occurred, either physically or in some automated way (like with a sensor)
- You’ve got to capture and save the measurements that matter, and the 5 W’s + How + any other qualitative information that you think are important to describing that event
Putting it all Together: “Dimensional” + “Model”
“Model” is the noun; “dimensional” is the adjective. When you put them together, you might get something like this:
A dimensional model is a representation of data that…
- Portrays its measurements in a way that highlights the characteristics, qualities, features, and facets of the event that generated the measurements (the who, what, when, where, how, of the data collection situation)
- Therefore, a dimensional model is ideal for understanding the patterns and information that the data contains in a way that is widely approachable by analysts of all kinds