Level 5 — Advanced · Technical & Evidence-Based Writing
Task 56: Building a Quantitative Argument
Use a number as genuine support for a claim rather than decoration for one — stating what changed, what it's compared with, and what a confound stops you from concluding, the writing side of Lesson 56's data-argument framework.
Learning Objectives
By the end of this task, you should be able to:
- State a data observation and a separate interpretation, without collapsing the two into one sentence
- Name a specific confounding factor that limits how confidently the data supports a causal conclusion
- Write a conclusion that matches the evidence's actual strength, not a stronger claim the data doesn't earn
Introduction
Lesson 56 taught you to push a single number through three levels — observation, interpretation, causal claim — and to stop before the top level unless the evidence genuinely earns it. Writing a quantitative argument means building that same ladder yourself: stating what changed, what else might explain it, and exactly how far the number is entitled to take your conclusion.
The Rule, for Writing
Core Concept
A weak quantitative argument jumps straight from a number to a cause: "Costs fell 18% after we automated invoicing, so automation cut costs." A strong one separates the observation from the interpretation, names at least one confounding factor the data doesn't rule out, and ends with a conclusion that is honest about what the number does and doesn't prove.
Example
Average berth turnaround time fell from 9.5 hours to 7.2 hours in the quarter after the terminal introduced automated crane scheduling — a 24% reduction. That's the observation. The terminal also increased night-shift staffing over the same period, so the drop can't be credited to the scheduling system alone; the honest conclusion is that turnaround time improved after the change, not that the change caused all of it.
Your Turn
Your Turn
Take a number from work, study, or a field you know (a cost, a delay, a conversion rate, an error count) that changed after something happened. State the observation as a plain fact, name one thing besides the change you are crediting that could also explain the result, and write a conclusion that only claims what the data actually supports — resist the urge to declare a clean cause.
Vocabulary in Context
confounding factornoun phrasetechnical
something else that changed at the same time as the thing you're crediting, making the true cause hard to isolate (বিভ্রান্তিকর উপাদান)
Often with control for a confounding factor, possible confounding factor, identify a confounding factor
“Weather is a confounding factor that affects both delays and productivity.”
attributable tophraseformal
able to be credited as the cause of a result (যা কারণ হিসেবে ধরা যায়)
Often with be attributable to, largely attributable, directly attributable to
“Most of the increase is attributable to higher fuel prices.”
associated withphraseformal
showing a relationship with something, without claiming it is the cause (সম্পর্কযুক্ত)
Often with be associated with, strongly associated with, associated with delays
“Longer dwell times are associated with incomplete paperwork.”
Model Answer
Show a model answer
Customer complaint volume fell from 340 to 210 per month in the three months after the support team introduced a new ticket-routing system — a 38% reduction. Over the same period, the company also hired two additional support agents, so the drop cannot be credited to the routing system alone. What the data supports is a genuine, sizeable improvement in complaint volume following both changes together; it does not support the stronger claim that the routing system by itself caused a 38% reduction, since staffing increased at the same time and no comparison team was tracked without the new system.
Self-Check
Self-Check
- Did you state the observation as a plain, separate fact before interpreting it?
- Did you name a genuine confounding factor, not just gesture at "other things"?
- Does your conclusion stop at what the data actually supports, without overclaiming causation?
- Did you avoid writing "after X, therefore because of X" as if the two were the same thing?
Next: Writing 57 — Defining and Classifying a Concept: building a category tree with a single, consistent criterion instead of a vague list of types.