I saw a headline today in an online publication cautioning PR practitioners not to embarrass themselves by making mathematical statements that didn’t make sense. I believe the article was focusing on talking about equations like ROI or ROE. I agree that one’s credibility with management will suffer immediately if you misuse such expressions.


However, a more common way to sacrifice your credibility in communications is to do a mathematical analysis of some kind and make mistakes. I see this regularly because I do survey research, which has lots of numbers in it. Sometimes you might want to combine two different respondent sets — for example those who have graduated college and those with advanced degrees, or those 18-24 and those 25-34 so you can describe those 18-34. These are not difficult calculations to make, especially if you know your way around a spreadsheet. But it pays to check and recheck, because if, during your presentation, someone questions your number in one instance and demonstrates you have miscalculated, it throws all of your data into question and everything else you or your team says.

I have been hired a number of times by agencies and communications departments of corporations, that have been called on relatively simple mistakes, to go through all their data and make sure it was air-tight before they presented it again.

One very useful trick I use is the idea of cross-checking. For example, let’s take the case in which we combined those 18-24 with those 25-34. Let’s assume the survey polled adults 18 and over. So if we take our new combined number of those 18-34 and subtract it from the total number or percentage of respondents, that should equal the sum of all the other age categories. If it doesn’t, you need to find out why. It might be you made a mistake with your original calculation, it might be a rounding error, or something else. But you need to know why there is that difference so you can defend the numbers, should someone call you on them.


Comments

Numbers Count — The Magic of Cross-Checking — No Comments

Leave a Reply

Your email address will not be published.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