The Demise of Quality - My Wife's Friend, British Airways and Air France

I was vacuuming the living room last evening, when my wife told me that her friend arrived safely in Egypt after a long flight from Washington with a connection in London. She then added that her 10 year old son is walking around in his swimsuit at home and her friend is wearing some old maternity clothes she had from a decade ago. I asked my wife "Did they lost their luggage?", she replied that no a single bag of the eight pieces her friend and her three kids had arrived.

It has been two days since my wife's friend family had arrived on board British Airways with no luggage. I mentioned to my wife that her friend should request a compensation of some sort, may $100 per passenger, so they can use this modest pocket money to go buy some clothes until their bags appear. My wife's response was "we are thankful that they arrived safely". Well, yes of course compared to Air France's latest flight that plunged in the Atlantic, the losses and inconvenience is minimal, and one should always be thankful, regardless of what one goes through". The point however is that the passenger's (customer) expectation is be transported safely along with his luggage from point A to point B. If customer's start changing their expectations, quality will automatically by definition change.

An attorney would seize the opportunity and ask the passenger to file a suit against the airline asking for some ridiculous amount of money as a compensation for damages. Maybe $1 million dollars per lost bag, or the highest that the court will allow. We have various mindsets of customers when it comes to settling disputes with a service provider. At one extreme a client will say no problem, just get me my bags if you can, when you can. On another extreme another customer will say meet me at court, and then there are infinite of perspectives in between.

What would you do? What would a good system's engineer do? Share your thoughts in the comments below.

Probability Distributions for Business Events - Binomial

System engineers and technical managers need to make decisions based on the probability of certain events occurring. For example during a risk assessment practice, the systems engineering might be calculating the risk value of an event, and based on the value of the risk level the organization will take certain steps to respond to the risk.

To accurately calculate the risk value which is calculated as the (impact level * probability of occurrence), the probability distribution selected needs to be accurate.

In some events outcomes could be binary, for example: good or bad, correct or incorrect, successful or failed, conforming or non-conforming. A suitable probability distribution would be a binomial probability distribution, using a binomial calculator the system engineer can calculate the probability of occurrence of one of the two possible events knowing the sample size (n), the rate of good versus bad, or correct versus incorrect (p) and the number of items (x) fitting a particular outcome. For example if we select a sample of six items from a batch which has a defect rate of 3%, we can find the probability that the sample has one defective item using the binomial formula

P(x) = [n! / x! (n-x)! ] p^x (1-p)^(n-x)

In the above example, n=6, p=0.03, x=1

Using the formula above or the binomial calculator available at Texas A&M, we get that P(1) = 0.1546

Stay tuned for other probability distributions that are also common in business environments.

Two Approaches to Use Statistics on Your Project

Most people don't like statistics, and most decision-makers make wild guesses when they need to take a decision. Statistics might not be the most straight-forward science during the college years, but it has tremendous value in providing insights and guidance for making rational decisions.

Using statistics one can show various properties of a set of data such as the mean, mode, median, dispersion and distribution. These parameters can be represented graphically using histograms, charts and other visual illustrations.

Another approach using statistics is to develop inferences, test hypothesis and develop forecasts through the use of sample data from a bigger population. Relationships between variables can be developed and illustrated using a scatter diagram or a regression equation.