# Average wage can deceive you

--

Average for you, median for me, mode for him, p-value for others !

TL;DR (for statisticians)

For all the statisticians, whose constant screams to help people appreciate or understand data distribution, sample vs. population, correlation vs. causation etc — is still profound to most folks who don’t work in information jobs. This article discusses some of the most fundamental concepts and how the world still needs to work on understanding just that, so skip to conclusion if you’d like.

For non-statisticians, continue..

# Measures of central tendency

Let’s say your data points are 1,3,4,5,7,9,9

Mean = 5.4, Median = 5, Mode = 9

Thats stats 101 intro ! Not so interesting and sometimes, it gets boring as we grind through formulae and get into permutations and combinations, and gotchas trying to make sense of data and its distribution — ultimately to make inferences and decisions, that model real world.

There are situations where mean/average matters more than median, or vice versa OR that our understanding of data distribution needs multiple measures . The stakes are not so high when you are in a class room and crunch numbers and argue with each other around hypothesis/conclusion/invalidation loops.

What if the data points are salaries — 10k, 30k, 40k, 50k, 70k, 90k, 90k

Think about scenarios (when you don’t have access to all data points) and your employer/new employer/whoever you are negotiating with tells you the below

• Hey the average pay for this job is 54k, but we can pay you 55k
• Hey the median pay for this job is 50k, but we can pay you 52k

You are quite happy in both the scenarios right ! Because you are made to believe you won (Philosophically happiness is sometimes found in ignorance, but lets discuss that for another time)

Until….

You accept the offer, and then one day you discover the data points and mode being 90k i.e. most of them got paid 90k

Funny how the stakes become high, once they are salaries/wages and your life is dependent on them !

# Eye to Details — Scenario 1

• In real world, think about how many times have you really paid attention to details when you look at data points ?
• When buying or selling real estate (home, land whatever)
• When buying or selling stocks
• When negotiating salaries/wages

Data and information access (at sample and population levels) is NOT so symmetric in the real world, especially when the stakes get high. And with more data produced with technology, the bigness of it makes it non-palatable for a human, hence we have to rely on systems, and that means stats/math/code etc. to aid with comprehension of such big data.

And in all of that — there is a great opportunity to be manipulative OR be fair and do real good in the world — The former or latter relies on the karmic decision of the person/entity who develops insights on top of data and presents to readers (I do say Karma, because it is almost impossible to put checks and balances around every data point with real-time detection unless it matters to the big guys)

As a simple example, going back to our data set

salaries — 10k, 30k, 40k, 50k, 70k, 90k, 90k — has states in which those salaries are paid

If you live in CA and all of your team members got paid in between 70k–90k, the average (54k) and median (50k) numbers told to you at the negotiating table is not good faith — in fact your lack-of-access to data played against you.

The above scenario is when you know the average and median (mode is still not yet discussed). Think of how many times you knew the wage data during your negotiations.

If you live in NY and only average (54k) is shared with you and lets say you are in the top 0.1% of achievers , wouldn’t you want to know that mode is 90k, and that is what you truly deserve (assuming you are confident of your skills)

How many of us know the average, median and mode salaries for the employer, job titles, state, city that we are hired in/at ?

# Eye to Details — Scenario 2

Again now lets say the same data is now shown as below

• If you live in NJ, the above average (54k), median (50k) salary is an awesome wage (At the negotiating table you are told company wide average is 54k)
• If you live in CA, the same average (54k), median (50k) numbers told to you so as to negotiate your wage down is pretty unfair (At the negotiating table you are told company wide average is 54k)

What differentiates the above — is the cost of living in the state you are living in.

Above we compared data only across states.

• What if you are getting paid the same salary over years ?
• What if you compare across cities, job titles and so on ?

Think about it. Isn’t inflation always going up ? Shouldn’t your wage go up in general when you start comparing over years

# Epiphany

If you had the epiphany by now — congrats ! The above two scenarios are still one of the most fair cases that happens in the industry and if you happen to get such information, you are lucky i.e. most of the times you don’t have access to average or median or mode. Symmetrics is the only platform out there, that provides such information at this point. An example is as below

With average, median, mode, min and max. information on wages, you are still at the beginning of your wage negotiation — but at least you can be fully invested in the interview process and give your full to it. Otherwise any energy (for e.g. that feeling of getting cheated in the last moment) that is not spent in interview is a lost energy and a lost opportunity.

For e.g. the salaries that you see in most user submitted portals are user-submitted. There is no verification for those and even if it were, only a few data points are verified and with few data points, you can tell the story steering the reader in any direction you decide upfront.

# Symmetrics.fyi

Symmetrics is a platform that provides fair metrics for base wages, approval KPIs (for immigrants) across various dimensions — YEAR, STATE, CITY, EMPLOYER, JOB TITLE and so much.

So much information is packed — at aggregate levels and at the same time personalized enough based on your location, employer, job title and so on.

# Conclusion

We discussed only wage, however there are so many other metrics for e.g. approval rate for employers, states, cities etc. for skilled immigrants is an important measure to look at the place they work at.

# More Articles

Do you know your Job’s net worth ?

Things the smart do BEFORE accepting a new offer

The Great Big Resignation and now this!

To Each his domain

--

--