January 11, 2019
Last month on the blog, we talked about skills and tips for writing about data clearly and effectively. This month, we’re following that up with some general writing advice, both related to analytical writing and applicable to writing about pretty much anything.
First of all, watch your tenses. When you start writing a piece, you should pick a tense and generally stick with it. The basics are past (I wrote), present (I write), and future (I will write), and each one obviously has its uses. Tense should be generally consistent across a piece of writing, but tenses may shift briefly depending on what you’re trying to express.
Past tense lends itself best to pieces that describe or analyze changes over time; you’re talking about things that already happened.
Example: The change in the estimated poverty rate between 2016 and 2017 was not statistically significant.
Example: The average high school graduation rate of Champaign County schools decreased between 2017 and 2018.
Present tense describes what is happening in the present. It can also be used to describe a methodology or other type of general practice that is performed in the same way consistently over time. It’s also good for pieces that provide instructions or suggestions, like this blog, and can be appropriate for less formal types of writing.
Example: Residents of group quarters are excluded from the U.S. Census Bureau’s estimation of population below the poverty level.
Example: The average high school graduation rate of Champaign County schools is calculated using the graduation rates of all high schools in the county.
Use of future tense in data analysis is generally pretty limited, but it crops up on occasion.
Example: Some analysts try to predict how the poverty rate will change in the next 10 years.
Example: An updated average high school graduation rate will be posted every November.
Consistency is good, but clarity is more important: start with the tense that’s the best fit for what you’re writing, but make sure to shift as needed.
Correlation vs. Causation
You have likely already heard of this one. Just because two variables are correlated (e.g., both variables increase or decrease together, or one increases as the other decreases and vice versa) does not mean that the change in one is causing the change in the other. This is a fundamental concept of data analysis and we won’t spend too much time on it in and of itself.
But even keeping this in mind, it can be easy to inadvertently imply causation when describing trends over the same time period. In the following examples, pay attention to how changing the punctuation or conjunctions changes the meaning of the sentences.
Example 1: Variable A increased between 2000 and 2010. Variable B decreased between 2000 and 2010.
Example 2: Variable A increased between 2000 and 2010; Variable B decreased.
Example 3: Variable A increased between 2000 and 2010, while Variable B decreased.
Example 4: Variable A increased between 2000 and 2010, as Variable B decreased.
Example 5: Variable A increased between 2000 and 2010, because Variable B decreased.
Example 6: Variable A increased between 2000 and 2010, then Variable B decreased.
The first two examples are absolutely the best options here. Splitting the statements about the two variables either into two sentences or two distinct halves of a sentence clearly separates the two trends. Both occurred over the same timeframe, but they aren’t otherwise connected.
Example 3 is also an acceptable way to go. Using “while” conveys that these two things are happening at the same time, but it doesn’t suggest that they’re related in any other way.
The fourth example is deceptive. Here, it looks like “as” should mean the same as “while,” suggesting that the trends are happening simultaneously. But keep in mind that sometimes “as” can also be causal (e.g., “We turned the thermostat up, as it is cold today.”). This could lead to some confusion about what the sentence is actually suggesting. Since there are clearer alternatives, like splitting the sentence with punctuation or using “while,” we recommend avoiding using “as” in this context.
It should be pretty obvious that Examples 5 and 6 are problematic. In Example 5, using “because” states outright that changes in Variable B cause changes in Variable A. Example 6 is unclear all around. “Then” could be interpreted as causal, implying that the changes in Variable A caused the changes in Variable B, but it could also suggest that Variable B’s decrease occurred after Variable A’s, starting in 2010.
If you’re trying to emphasize that there is no causal relationship between two variables, the word “regardless” could come up as an option.
Example: Variable A fluctuated over the course of the study period; Variable B consistently increased over the course of the study period regardless of changes in Variable A.
This is a completely appropriate use of “regardless,” but, as a reminder, there are no appropriate uses of “irregardless.” It isn’t a word. It’s not up for debate. Don’t use it.
Keep It Simple
Finally, don’t overcomplicate things. Clarity should be the first priority in any technical or analytical piece of writing. The recommended examples for Correlation vs. Causation, in addition to having appropriate conjunctions or punctuation, are short, functional, focused sentences.
There can be a misconception that longer or more complicated sentences make a piece of writing sound more intelligent, more credible, or like it has more substance than a piece primarily composed with shorter or simpler sentences. But the goal of any piece of writing, and especially of technical and analytical writing, is to effectively communicate ideas.
You will absolutely encounter concepts and topics that can’t be expressed fully or accurately in a short, simple sentence structure. That’s fine; in those cases, take as many words as you need. But when it’s possible to express an idea in a simple way, making it more complex than necessary often makes the finished piece less clear rather than clearer.
No matter what you’re writing – data analysis, technical writing, academic writing, or anything else – the quality of writing matters. A sound argument communicated badly isn’t convincing; a good idea or solid conclusion that isn’t explained well might not be received well. Conversely, a good idea with great presentation can generate more interest. So it’s important to take some time to make sure your writing is clear and concise, and expresses your ideas as well as possible.