Understanding the basic principles of statistics transforms a
good manager into a great manager. Using basic business statistics,
you can turn data into knowledge and increase your decision making
accuracy. In the time it takes to make a coffee, you'll know the
most useful basic business statistics and become fluent in number-speak.
"Lies, damn lies and statistics".
But first, a cautionary note. Statistics are generated by everyday
people who have underlying motives, both good and bad, for presenting
numerical information. Your job is to question and question some
more until you're satisfied you understand;
What question the data is answering?
Who paid for the research?
What's the data source and how was it captured?
How has the information been manipulated from its raw form,
to its current form?
Conversely, if you are presenting information, it's good practice
to answer those questions up-front, and you'll save yourself a major
grilling from your audience.
Basic Business Statistics Terms
Mean - the arithmetic mean is the average value of a
set of numbers. E.g. the average in this number series, 5 6 8
1 2 7 5 6 4 9, is 5+6+8+1+2+7+5+6+4+9 = 53 divided by the how
many numbers in the series = 10, which gives you a mean of 5.3
Mode - the number occurring most frequently, in a set
of numbers. E.g. in this number series, 4 5 3 4 8 4 5 6 4 2, the
mode is 4 as it occurs most frequently.
Median - the middle value of the distribution, that is,
if the numbers were put into order, an equal number lie above
and below the median number. E.g. in this number series, 2 2 3
3 4 5 5 6 8, the median is 4 as there are an equal set of number
both above and below it.
Average - technically speaking, this can be the mean,
mode or median, but usually people expect an average to equal
Standard Deviation - nope, nothing to do with kinky fetishes,
and everything to do with describing the data and how it varies.
If a set of data has a high standard deviation, it means that
the numbers include teeny numbers and massive numbers, but if
the standard deviation is small, the range of the numbers is small
Outliers - if you want to mislead and draw wonky conclusions,
it's best to keep your freaky way off the scale numbers in your
study. But often, those radically different numbers are because
you've made a mistake. However, check them out before you cast
the outliers aside, because you may have stumbled upon something
genuine and worth further investigation.
Significance - Numbers geeks get a bit sniffy if you
don't understand statistical significance. In short, if you are
comparing the results of a test and want to confidently say one
result is better than the other, you have to test for statistical
significance. This means doing a bit of number-crunching to figure
our whether the difference in your results is genuine or is just
due to something random, like the weather, day of the week, or
errors in your data capture method.
Normal Distribution (Bell shaped curve) - so named because
when you plot the frequency of the numbers in a data set, you
get a bell shaped line. A classic example of a bell shaped curve
is the height of a population, where there are small numbers of
extremely tall and extremely small people, and the majority residing
somewhere in the middle.
Skewed Distribution - if you did the same height exercise
but only used basket-ball players, you're distribution curve will
be skewed towards taller people. Non-bell shaped curves are valid
descriptions of data too - don't be fooled into thinking everything
has to be a perfectly symmetrical bell shape.
Quartile - you'll probably come across this term in relation
to salaries. Companies often compare who pays what for similar
jobs and put this onto a graph. This will usually give a normal
distribution, centred about the average salary, and ranging from
the lowest to the highest pay rates. If you split the normal distribution
into 4 pieces, or quartiles, you see what the bottom 25% of people
get paid, the next 26% to 50% etc etc.
Some companies have a policy where if you are new into your job
you will be paid in the lower quartiles, and as you gain skills
and experience, your salary will increase and head into the upper
quartiles. For example, a salary might range from $40,000 to $80,000
with the average at $60,000. You will start in the lower quartiles
somewhere between $40,000 and $60,000, and progress over time
to a salary of between $60,000 and $80,000.
That's probably enough basic business statistics for now! Embrace
numbers and you will transform from a good to great manager. I'll
leave you with a quote to ponder;
"I often say that when you can measure what you are speaking about,
and express it in numbers, you know something about it; but when you
cannot measure it, when you cannot express it in numbers, your knowledge
is of a meagre and unsatisfactory kind." Lord Kelvin (1824-1907) Scottish physicist and
If It Moves, Measure It! How To Create a Histogram
When you know how to create a histogram, you move from knowing
a little bit about a process to knowing a lot. When you know your
process, you've gone a long way towards making big improvements.
Learn how to make a histogram in this short article, and discover
hidden information about your process