Friday, December 7, 2012

Appreciating the role of data in Atc - Part II: Using Empirical Studies in Atc

Take a look at the illustration below. A 12th century engineer is proposing the construction of the tower of Pisa. Can you recognise the link between data evidence or empirical studies and the engineer's photo of the proposed Tower? What do you think he is trying to suggest by his words at the side of the photo?

The engineer did not recognize the need for data analysis. He apparently was more concerned with saving costs. 
We know that the Tower in Pisa, Italy leans to one side. The message presented here is the usefulness of empirical studies. If the engineers had done more of the appropriate data analyses, they would have surmised that the Tower will tilt and they would have built a stronger foundation. 
The Aviation sector has many opportunites for using data evidence. All issues in Atc should be accompanied by empirical studies. Yes, there should be some supplementary data evidence to act as a vehicle for the issue under investigation. The hearer will at least acquire a more accurate perception or a direction of the issue. True, some topics in Atc might be boring but visual aids add variety to the Atc topics requiring our attention. We will at least be less inclined to nod off at a seminar. 

Today, we will examine the issue of Fatigue using the guide from the previous post. Remember the   illustration of the TV purchase from the last post? To help us we could use the acronym ORACLE. Our ORACLE methodology will form the basis for our examination of case studies. From our acronym, O stands for objective, R stands for research, A stands for analysis, C represents confirmatory tests, L represents the stage where we make linkages or inferences from our analysis and THEN we exit the process with an informed decision.  
O represents our objective or aim or purpose of the empirical study
In this case, the objective was to determine the factors that contribute to fatigue and the organizational economic implications. In my readings, I noticed several perspectives on Fatigue:- 1/ it is a conditional construct and thus difficult to measure; 2/ it is related to sleep loss; 3/ it is linked to safety. I had not seen anywhere with the exception of sleep loss that a study was done to confirm that specific factors cause Fatigue. 
Being a controller, I was also concerned about making an application to Atc. I wanted to do a study that confirmed specific factors in Atc contributed to fatigue. I also wanted to determine the organizational implications of the prevalence of these factors in keeping with the organizational theme of my postgraduate studies.
R stands for the research that we do
We need data evidence or an experiment or an example to help us better understand our objective. That research is based on knowledge. Also important, in this context are the preferences of the Atc unit and not necessarily the preferences of the administrators of the Atc unit. Inadequate training of administrators/managers in Atc units pave the way for the influence of narcissism and personal bias in decision-making. It is possible that the preferences of the Atc unit may be far different from the preferences of an Atc administrator who is not fully trained to meet the demands of the Atc unit.
We use the research to aid us in collecting evidence in groups or variables that will serve as our data. In doing an analysis, the names of the employees disappear into a chasm of nonexistence and the statistical principles now reign supreme. The analysis is concerned with figures, not names.

In the research on fatigue, from at least 50 studies, I selected several benchmark papers and a book as the foundation for the study. My main references were: an engineering paper that justified the quantifying of fatigue via Bayesian methods*; a paper that looked at age and shift work**; a report by the NTSB, USA(2007) ; the psychosocial effects of shiftwork***; the proposal of a Fatigue Index Tool**** and a book on organizational concepts*****. 
Since it was a global preference of ICAO at the time, I considered safety as a preference for the Atc units of the Caribbean and I set up as my assumption that fatigue is inversely related (that is, an increase in one aspect causes a decrease in the other) to safety operations in ANS. 
The purpose of doing research as you may have aleady noticed, is to acquire a basis for formulating the variables that help us to collect the data. We must have a general idea of how we are going to proceed with an analysis. It serves as a guide like a road map or GPS system that we use for direction. 
A is for the actual analysis of our data
I focused on the statistical application of Bayesian theory by reading the engineering paper. Like the GPS system, I used my research to formulate variables that I considered appropriate such as the testing of organizational psychological variables based on the based on the article that spoke about psychosocial conditions. In a little while you will read about the analysis that I used.

Fatigue is a conditional construct which means that it is subject to individual behavior and surveys are excellent in capturing subjective conditions. We can easily capture individual controller views surrounding Fatigue by using a survey. We tend to shy away from filling surveys but they serve useful purposes. We can anonymously supply information that can be analysed to dispel or prove a speculation or to help us identify flaws in a policy or to help us consider the costs and the benefits of an objective.
I have divided this post into 2 so as to keep the information concise without causing distraction by introducing the elements of boredom. I do not want anyone to be fatigued from reading about a study on Fatigue. This evening you will be able to read about making the inferences from linkages and the exit decision as well as the highlights of the study on fatigue, ....à bientot! (soon!)

*(McCallum & Nigam, 1998; Qiang Ji et al, 2006)
***(Mélan et al,2008)
****(Kilner, 2010)
*****(Morgan, 2006)

Reddit Digg Share on Tumblr Delicious StumbleUpon Follow Me on Pinterest Print Friendly and PDF

No comments :

Post a Comment