居眠り海難の分析-II : 海難の類型化と居眠り因子の因果関係

概要

論文の詳細を見る
In the previous essay, we obtained the score of items using the result of factor analysis. By means of these score, we classified the accidents into clusters by using cluster analysis in terms of the type and greatness of ship and the area where the accident occurred. The result was 5 clusters of ship types. They are (i) fishing vessel, (ii) cargo ships and tankers, (iii) passenger ships and ferry boats, (iv) towboats and push boats, and (v) pleasure boats. It was found that the accidents can be classified into clusters of different greatness of ship and areas where the ship's accident occurred according to the greatest number of ship types included in each category. Next we assumed the causal relation between two factors where correlation was recognized. We obtained the correlation from factor analysis using oblique geomax method. In order to make clear the structure of watchkeeper's drowsiness, we described the path diagram for causal relation between each factor.
社団法人日本航海学会の論文
1994-03-25