U. S. Department of Transportation
Download 2.66 Mb.
|
- Navigate this page:
- Aircraft Information
Descriptive StatisticsThe following section focuses on the analysis of various groups of variables. The groups of variables to be discussed include aircraft information, pilot information, controller information, weather information, and other variables. It is important to keep the overall distributions noted in the previous section in mind when examining these subsets of the data. While formally a regression model, the logistic regressions (logits) presented in this section serve a role similar to the descriptive statistics above, a way to explore, rather than explain, the data. The results presented in this section focus on single variables with some examples of two or three variables at a time. The drawback of using these logit models is that the dependent variable must be dichotomized – destroying some information inherent to the rankings. It was chosen to examine severe (categories A and B) versus non-severe (categories C and D) events. To reiterate, these results serve more as data exploration and as a way to being to quantify the effect of various variables rather than as a formal modeling exercise. More formal modeling results are presented in Section 4.3.
Aircraft information originates from both the Runway Incursion and ATQA OE databases. All of the variables are of a categorical nature. These variables cover information about what the aircraft was doing at the time of the incident. Intersecting Runway Departure or Arrival(Runway Incursion Database) The Runway Incursion database contains information on whether there was a departure or arrival on an intersecting runway. Figure 1 presents the distribution of this variable. Table 9 and Table 10 contain the cross tabulation of this variable by incident severity. Note that category D incursions were excluded (as by definition an event would not be a D if this variable was yes). This table includes the results of Fisher’s Exact test. This is a similar test to the Chi-Squared test indicated above, and tests the same hypothesis (independence of row and column categories), but is applicable when some cells have very small values and the assumptions of the Chi-Squared test do not apply.35 
Figure – Distribution of Intersecting Runway Departure or Arrival Table – Observed Distribution of Intersecting Runway Departure or Arrival by Severity
Table – Expected Distribution of Intersecting Runway Departure or Arrival by Severity
This table indicates that there is a relationship between these two variables. Examining the observed versus expected values indicates that incidents with departure or arrivals on intersecting runways occur more frequently than expected among category A and B incursions than among category C incursions. In some sense, this is not surprising given the definition of incursion severity – if there is an arrival or departure on an intersecting runway it is more likely that the two planes will come into conflict. Given that, this result indicates that these events are more severe than other conflict events. Table 9 indicated that there was a relationship between this variable and severity. Table 11 presents the results in terms of odd ratios. Again, category D incursions are excluded for definitional reasons. Table – Logit Estimate of Impact on Severity, Intersecting Runway or Departure
Future Research
The results suggest that the odds of being severe for incidents with an operation on an intersecting runway are approximately twice as large as those without. Again, this is consistent with Table 9, but is a more quantitative look at this relationship. Table 12 presents the results of a logit where the dependent variable is a flag for an OE incident or not. Again, category D incursions were excluded for definitional reasons. Note that the alternative here is “not OE”; that is, both V/PD and PD incidents are included in the alternative. The odds of an incident being an OE are approximately 4.4 times as high if there is an operation on an intersecting runway. Recall that no V/PD incidents were coded as “yes” for this variable. This should temper the effect somewhat, as seen in Table 13. Table – Logit Estimate of Impact on Incident Type, Intersecting Runway or Departure
Table – Logit Estimate of Impact on Incident Type, Intersecting Runway or Departure, OE and PD Only
Table 14 and Table 15 contain a cross tabulation of the same variable by incident type. Category D incursions are still excluded. Interestingly, intersecting runway departure or arrivals occur most frequently for OE incidents. The Chi-Squared statistic supports the conclusion that there is a relationship between these two variables. The observed values reveal two things. First, there is only one V/PD incident where this is variable is coded as yes. This suggests that airport vehicles are effectively never in a situation where this could be coded yes. Secondly, OEs are over represented while PDs are underrepresented. This relationship holds even when V/PDs are excluded from the analysis, as seen in Table 16 and Table 17. This indicates that intersecting runway departures or arrivals are proportionally less a problem for pilots than controllers. Further research is required in this area to detail why that is the case. Table – Observed Distribution of Intersecting Runway Departure or Arrival by Incident Type
Table – Expected Distribution of Intersecting Runway Departure or Arrival by Severity
Table – Observed Distribution of Intersecting Runway Departure or Arrival by Incident Type, OE & PD
Table – Expected Distribution of Intersecting Runway Departure or Arrival by Severity, OE & PD
Directory: lib -> 46000 lib -> Strategic Plan South Sudan Country Office Context lib -> The World Order of Bahá’u’lláh Selected Letters lib -> 1. naclo 2015 – North American Computational Linguistics Olympiad lib -> Ieee/iee electronic Library (iel) lib -> Lnpa wg position on Porting of Telephone Numbers Used by Voip service Providers lib -> 1. naclo 2017 – North American Computational Linguistics Olympiad lib -> [0] B. Hardware [2] lib -> [1] B. Hardware [1] lib -> Chem I honors Unit 2 Notes: Numbers in Chemistry Measurement 46000 -> Measuring the effects of aborted takeoffs and landings on traffic flow at jfk Download 2.66 Mb. Share with your friends:
|