Hacker News Reader

Comment by svantana | Hacker News Reader

>Any normal statistician would just report on the mean of all the scores they collected from respondents

What? No!!! Any self-respecting statistician would know that the mean of a quantity where addition is not well-defined is meaningless.

View on HN

Wow, yeah, agree. Its not the only point in the article and some of the other ones are fairer criticism, but this line was a bit painful.

Use the median!!!

I winced as well! But at least he grabbed an actual mathematical concept used to describe a set of data.. as opposed to whatever NPS is using.

It's not just mean and median - it's proportions. The metric is about growth, so it's specifically concerned with the top of the distribution.

> For reasons never fully explained

The reason is obvious. The entire theory is that for your business to be successful you need to have exceptionally satisfied customers who will PROMOTE your business.

All you have to do is read the name of the thing to understand exactly how and why it works, which apparently everyone besides this author can do.

Author is arguing it doesn't work. You're begging the question.

I've never heard this before, and a skim of the wiki page doesn't mention it as a prerequisite. Mind explaining? The scores are just integers, so the addition is well defined. So you're saying that the context is what's relevant?

Not that the mean is the only (or even the most useful) statistic.

Well, take decibels for example. They are a log scale physical intensity, so averaging them make no sense whatsoever (e.g. absolute silence is negative infinity dB). I would argue these scores are more like labels than actual numbers (i.e. a 3 plus a 5 doesn't really equal an 8 in any real sense). You can of course take the mean of any collection of numbers, but I've heard many a statistician lament such careless practices. The median is at least more easily interpreted for cases like this.

Ah, thanks. The decibels example makes sense (an alternative would be to take the log first, and then convert it back after averaging?), and I can see how the 0-10 system can also be viewed as categorical rather than discrete.

It's because the NPS rating numbers are ordinal, meaning that you can put the rating numbers in order, but the likelihood gap between the numbers may not be equal.

For example, 6 on the NPS scale would be less likely to recommend compared to 7 and 7 would be less likely than 8. However, the gap between 6 and 7 and the gap between 7 and 8 may not be equal. If you were to get the mean of 6, 7, and 8, you would get a value of 6, but there is no guarantee that the average of the participants' likelihood to recommend was actually equal to a 6.

Measuring the mean would work if the interval between the numbers were equal. See here for details: https://en.wikipedia.org/wiki/Level_of_measurement

> Any self-respecting statistician would know that the mean of a quantity where addition is not well-defined is meaningless.

There goes the Dow Jones average.

> the Dow Jones average

The Dow Jones was invented in the 1890s to be easy to calculate. It is widely panned and is not used by anyone in industry for anything serious.

Reply on news.ycombinator.com