What are the most essential math concepts that all journalists should know?

Daniel Bachhuber and I want to know — and we’ll compile a list based on the responses. Please comment below and include links, if possible. Thanks!

Skip to content
# Essential math concepts for journalists

##
14 thoughts on “Essential math concepts for journalists”

### Leave a Reply

What are the most essential math concepts that all journalists should know?

Daniel Bachhuber and I want to know — and we’ll compile a list based on the responses. Please comment below and include links, if possible. Thanks!

Your email address will not be published. Required fields are marked *

how to calculate the street value of drugs for police beat reporting

PERCENTAGE. CHANGE. (new – old)/old

Also, calculating rates per thousand(etc.), especially when the populace is small.

Not sure if this is too broad for your purposes, but I’d like a guide to “Here are the questions you must ask when you get a press release about a survey.” Evaluating sample sizes, self-selected versus random polling, whether the results are statistically significant, etc.

This, for example, had no credibility without any info about the methodology: http://www.mediapost.com/publications/?fa=Articles.showArticle&art_aid=155451

As far as math concepts, beyond basic arithmetic I think it’s mostly stats stuff that’s most useful.

Understanding sums, box plots, quartiles, mean/median/mode. Understanding percentages (if there’s X% chance of A happening per year, what percent chance is there of A happening in 10 years) and the difference between percent change and percentage point change. (something increased X% versus X percentage points)

Of course, science journalists or data journalists need to know/be comfortable with a lot more.

How to calculate equal fractions (e.g. a/b = x/c ) Particularly useful in web design.

Also check out IRE’s Math for journalists test.

http://www.ire.org/education/math_test.html

Many of the most damning problems I’ve seen with math and journalism have less to do with inaccurate computation and more to do with a lack of understanding of logic, specifically false analogies, using true figures and numbers to “prove” unrelated claims, overgeneralization or post hoc.

I wrote a guest post about this on Michelle Rafter’s wonderful freelance writing blog, Word Count: http://michellerafter.com/2011/07/12/5-simple-math-equations-every-writer-should-know/. Great topic!

Laura Laing

http://www.mathforgrownups.com

I’ll propose two:

•

Regular expressions: just the basics. It’s a bit daunting to look at, but retest is actually pretty easy if all you need is the basics, and while not exact ally math, it’s amazingly useful for looking at datasets.•

Statisticsmore than just the basics. A good jour no should know enough to determine the statistical significance of anything in the fly. Politicians, corporations, and other institutions throw out meaningless numbers all the time that journos never bother to check. An understanding of bell curves, r values, significance, and probability would be huge.I’d go with the basic concepts of statistics, sampling, and probability, because the logic of these is KEY to what journalists do when they try to describe a whole class of events or people from just a few examples. There are right ways and wrong ways of doing this.

– The basic notions of a “population,” and “statistic” as a number that captures some feature of a population.

– What is sampling? How does it work? When does it fail? What kinds of sampling bias do journalists need to look out for?

– How can we reason effectively about very rare events? Conditional probability, base rate fallacy.

See also:

Understanding errors, biases that affect journalists. A great old paper on how known cognitive biases play into journalism.

Science training for journalists, a blog from the Royal Statistical Society (UK).

I somewhat touched on this topic in a blog post I wrote a few months ago about media coverage of Egypt’s sexual harassment problem: http://laney-lanesworld.blogspot.com/2011/05/every-once-in-awhile-in-midst-of.html.

A basic issue, in many articles and news reports about data, is that journalists fail to say whether a finding comes from a convenience sample or a true random sample. Similarly, few take the time to carefully look at definitions — obviously, *how* a term is defined will affect what findings one obtains and how one might interpret those findings. No surprise — terms defined very broadly result in BIG statistical findings which may, substantively, not be as meaningful as originally thought.

I don’t expect most journalists to understand math beyond basic statistics (though I agree that if their job regularly involves looking at statistics — reporting on politics or on health, for example — that they should know r values, significance and probability). Even if they struggle with those, if they *at least* understand that a convenience sample can never be said to offer more than suggestive data and that definitions can greatly skew findings, they wil be more able to report findings with appropriate caution. Viewers/readers may not know the r values or the tests of statistical significance, but it would be nice if they could trust that journalists who say a finding is significant are actually referring to statistical meaning and a well-designed study rather than simply to substantive interest.

All the suggestions above are great, but what we really need is a j-school class that gets students passed the I-don’t-do-math-and-I’m-proud-of-that attitude. Too many journalists, including those who teach, carry that with them. You can no longer be a journalist without basic, fundamental understandng of just about all of the above.

And, besides, none of it is that hard to get your head around.

One more late entry: order of operations.

Basic concept, but it will bite you if you start putting formulas into Excel (or a calculator, or Python) and expect things to just go left to right.