Criteria for Evaluating GDSS

---

The following case study is an interesting evaluation of a GDSS by an empirical method.

The method is objective and scientific in the sense that it uses regression analysis.

The technique is applied to monthly statistics, about the system, gathered over a fifty-seven month period.

The New Orleans Vessel Traffic Service (NOLA-VTS) is a voluntary vessel movement reporting system operating in the Lower Mississippi River area. It was installed in the early 70s. It is best described as a Local Area Decision Support System. Prior to the system
being set up, the Mississippi River had the worst shipping accident record in the whole of the United States.

The hardware consists of VHS-FM communications equipment and monitors. The system processes and disseminates information about traffic movements in the VTS area to the participating vessels. It is manned round-the-clock by about 50 coastguards using radar-like consoles that show simulated traffic movement, over an area of about 360 square miles.

This is divided into four control zones.

Ships report their position, destination and speed when entering the VTS area for the first time and at subsequent pre-arranged reporting times. This is supplemented by coastguards monitoring ship-to-ship communications as they make individual arrangements for passing one another. In between these positions, the ship movement is simulated. The system provides vessels with information on predicted traffic that they will encounter on moving to the next reporting point.

Also, other critical information such as “no wake” areas, dredging, ships anchored in unauthorised areas and other hazardous situations are relayed to the vessels. Ships can request other information that is not given automatically in order not to
clutter up the airwaves, froe example: low-visibility areas, river stages and bridge clearances.

A regression equation takes the following form in Equation 2:

When carrying out regression analysis, it is found that some of the explanatory variables are not statistically significant and can therefore be ignored.

In our case only those variables that have their coefficients underlined with * or ** or *** are of any significance.

The most important performance measure is the number of accidents per month shown in Equation (1). “After all, that is the purpose of the system: to reduce accidents! “

We can see from this equation that for every 1% increase in the usage of the system by vessels, accidents are reduced by 0.26 per month.

Other factors influence the number of accidents such as the river stage and traffic intensity. However, these are being used as control variables in order to isolate the relationship between “usage” and the “number of accidents”. This shows that the system is effective and gives a quantitative estimate of its effectiveness.

Equation (2) looks at the facts that determine usage. Here we see that all the factors are significant. However, the factor that had the largest effect was the forced closure of the system for a six-month period due to a lack of Federal funding. This had a knock-on effect, so that when the system re-opened, usage dropped by 9%. Another factor that encouraged usage was the onset of bad weather in the two period. It gives a quantitative estimates to the effect that the explanatory variables have on usage.

Finally the R² term indicates how much of the variation in the dependent variable is “explained” by variation in the explanatory variables. For example: Equation (1), 70% of the variation in accidents per month is explained by the variation in the explanatory variables on the right-hand side of the equation. The other 30% is left unexplained and is due to factors not taken into account. A similar argument applies to Equation (2) where R² = 90% (higher the R² the better.)

---

Reference(s)
	Journal		Dolk, D. R. & Kridel, D. J. (1991) An active modeling system for econometric analysis. Decision Support Systems, Volume: 7, Issue: 4, Page(s): 315-328. [doi: 10.1016/0167-9236(91)90061-F]. [Available on: ScienceDirect: http://www.sciencedirect.com/science/article/pii/016792369190061F].
	Book		Pfaffenberger, B. (2002) Computers in Your Future 2003. 5th Edition. Prentice Hall: United States of America (USA), New Jersey (NJ), Bergen, Upper Saddle River. [ISBN: 9780139227820]. [Available on: Amazon: https://amzn.to/3gv8n7D].
	Book		Sprague, R. H., Watson, H. J. & Sprague Jr, R. H. (1993) Decision Support Systems: Putting Theory into Practice. 3rd Edition. Prentice Hall: United States of America (USA), New York (NY). [ISBN: 9780130422354]. [Available on: Amazon: https://amzn.to/3Defw3T].

---

Reference (or cite) Article
	Kahlon, R. S. (2013) Criteria for Evaluating GDSS [Online]. dkode: United Kingdom, England, London. [Published on: 2013-02-10]. [Article ID: RSK666-0000105]. [Available on: dkode | Ravi - https://ravi.dkode.co/2013/02/criteria-for-evaluating-gdss.html].

---