Think Like a Modeler Skill #1: Define

Someone asks you, “Do you think Obama’s proposed immigration policy is a good one?” How would you respond? You might end up talking in circles if the two of you have different definitions of what “good” means. “Good” to one person may be a policy that effectively keeps immigrants out of the country, while to another, it means a policy that considers fairness and opportunity.  Someone in the legal profession might be obsessed with whether Obama overstepped his presidential power limits, while a politician might perceive “good” as whether the policy boosts the public image of Obama’s political party, regardless of whether it is truly effective. These are all questions that force us to define our purpose.

Similarly, models must have a purpose – there should be a central problem that’s being explored so that you aren’t just modeling an entire system. This problem helps guide your modeling – if you reframe the problem, then your model will end up with very different variables and relationships.

Your purpose helps you define the boundaries of your model by forcing you to draw a box around the parts that you deem to matter and ignore the parts that don’t, so that you can focus on the central question. But this is inherently problematic because there are no true boundaries in the real world. Our personal lives affect our work performance just as our work lives affect our mood at home. If you’re feeling tired at work, you might also ask how much sleep you get at night, whether you eat breakfast, and whether your conversations at home help you relieve stress rather than create more of it.

The way we perceive our food consumption experiences shape where we place boundaries on our understanding of food systems. If you grew up on a farm and raised most of the foods you consumed, you’ll likely have a strong opinion about what is considered “good” food based on how it is grown.

But if you grew up in the city and your perception of food starts at the local market, then you create a more limited boundary. You might have more limited knowledge about pesticides use, food transportation, farm subsidies, and farm labor practices. At an even greater extreme, an increasing percentage of families in the U.S. don’t cook at home.

If your food consumption starts as a ready-prepared meal provided by a restaurant or supermarket, you’ll have even more limited boundaries; your food starts at your table when the waiter brings it out to you, just one step away from where it ends up in your mouth. You don’t know how the chefs prepared it or where the ingredients came from. The invisible boundaries we draw can greatly influence our consumption choices and our understanding of systems. And this is where purpose comes in – if you care about food systems and have learned about the health, environmental, and ethical arguments for responsible eating, then you’ll set a wider boundary around your eating habits. If you perceive eating as a daily routine focused solely on enjoyment, relaxation, and short-term bursts of delicious flavors, then you’ll be less careful with eating habits.

A popular quote paraphrased from George E.P. Box goes “All models are wrong but some are useful”. A directionless model that tries to represent an entire system becomes cluttered with too much detail. So you shouldn’t try to model the entire system, but instead, start with a question. For example, if we tried to model the entire political system, we’d likely both fail to capture everything, and also fail to understand the complex diagram in front of us.

So instead, we can start with a problem we want to understand – How does the electoral college affect voting behavior? How do campaign finance contributions impact citizen perceptions of candidates? How does our current three-branch structure create gridlock? you can hone in on variables and relationships that are most relevant to a defined problem, which improves both your ability to represent and interpret the system. “Purpose” describes both the process of understanding a particular problem, as well as the goal to fix that problem.

Practice This: Similarly in our everyday lives, the only way we can evaluate the things we participate in is to have a purpose to measure against – in other words, how do you know you are doing a good job at something?

Try this: For the following arenas, define what your purpose for each of these would be. Or define what you would like your purpose to be. Then, write down whether you are on track to achieving that purpose:

• Work
• Friendships
• Romantic relationships
• Watching the media that you currently consume
• Reading the books that you currently read
• Whatever activity you have planned for tonight
• Having Children
• Life

Hopefully you’ve recognized that none of these items should be the end purpose. You don’t work just because you want to work, but because there’s something valuable you get from working, and when you feel yourself satisfying what the underlying purpose behind work, you feel better about your job. Some purposes for work can be: desiring a stable paycheck, pursuing something you are passionate about, making as much money as possible, meeting interesting people, finding something to occupy your time, climbing a ladder, etc. Once you’ve defined the purpose that matters to you, ask yourself if you’re truly satisfying that purpose at the moment. If you’re struggling to find fulfillment, what are some things holding you back? Are there pieces in the other arenas that impact your performance in this particular one? How can you design an intervention to maximize your purpose?

---

Previous:

• What is Mathematical Modeling & Why We Should All Become Modelers
• Think Like a Modeler Skill #2: Expand
• Think Like a Modeler Skill #3: Aggregate

Further Reading:

• How to Read Causal Loop Diagrams