The Cause Of Erroneous Statistics

Skyscrapers of Singapore at Central Business District near Marina Bay Financial Centre.

It is the time of the year again, when universities across the island publishes the starting pay of their recent cohort of graduates. Turns out, despite the pessimistic economic outlook of 2016, people are getting higher median and mean pay. (Read Here)

Now, the use of “median” and “mean” sounded scientific and anyone who studied Mathematics would be doing a mental check on how realistic the numbers are and whether the news was sensationalised etc; Because news are mainly fake these days.


Reach for the sky, they said.

I have no doubts the numbers were obtained through scientific means and through respectable research companies.

When my younger friends and colleagues were lamenting about how their current pay is equivalent to the starting pay of fresh graduates, indirectly erasing years off their experience, I told them to take those numbers with a pinch of salt.

Anyone who read “How To Lie With Statistics” by Darrell Huff would remember the same example used to illustrate how properly conducted surveys could still contain bias; as long as a system involves human to provide inputs, that system can never be perfect because humans are inextricably biased.

The logic is, if you are earning less (knowingly, because you can always refer to last year’s numbers), or is struggling to make ends meet, how likely are you gonna reply to a survey that essentially wants you to declare how pathetic you are?

It’s not that no one will reply, but due to the low probability of reply rate, it makes for a smaller pool of people disclosing that their pay is on the lower end. On the other hand, those who are making plenty are filling up the replies almost immediately.

This slight bias in human nature skews the statistics enough to push the median and mean higher than it is supposed to be.

There may be a variant of information collection, by perhaps, surveying employers instead. Surely an employer will not be as emotional as that graduate who is receiving lower-than-average pay, ya?

Now, if I were an employer, I will not want potential candidates to know that I am paying lower-than-average, which will therefore shrink my pool of candidates. In fact, the number should be kept secret, so that if I do come across an outstanding one, I can still up their starting pay, without inducing others to demand the same.

On the other hand, those employers who can afford to splurge will definitely announce how much they are willing to pay, so that they can attract an even greater pool of highly-qualified candidates.

Therefore, at the end of the day, it is all about seeing things in the correct perspective.

If we are really unhappy with the pay we are getting now, we should progress and look for better-paid jobs than sitting their and lamenting.

That is, if we are really worth than what we are being paid.


Everyone has a long way ahead of us.

Disclaimer: I am a casual social commentator, who speaks because it is logical and rational, so my points are not necessarily based on established research and studies. As much as possible, I will refer to theories and knowledge I picked up through formal education. If you do agree to my ideas, please feel free to contaminate others in your social circle. If you disagree, though, it would be good to provide sources and credits, so that I can learn from my mistakes.

Leave a Reply