The Effect Of Attention On The Guide Dog Training Success Rate

Objective: To establish and improve the methods of ten behavioral tests and evaluation criterion for 33 behavioral variables of DMA(dog mentality assessment). Using the attention related behavioral variables of Ghosts test, Chase and Distance-play tests, Play and Passive situation tests, and by analyzing the test results of the dogs from Chinese Guide Dogs Training Centre of Dalian(referred to as Training Center), to explore the effect of attention on predicting the training success rate for guide dogs.Methods: The methods of Social contact, PlayⅠ, Chase, Passive situation,Distance-play, Sudden appearance, Metallic noise, Ghosts, Play Ⅱ, Gunshot tests and the related evaluation criterion of 33 behavioral variables in total were improved and established. 9 guide dogs and 8 eliminated dogs from Training Centre were used to examine the predicting effects of the 5 behavioral variables of Ghosts test on the training success rate for guide dogs. Besides, 12 guide dogs and 12 eliminated dogs from Training Centre were used to examine the predicting effects of the 9 behavioral variables of Chase and Distance-play tests on the training success rate for guide dogs.And 17 guide dogs and 17 eliminated dogs from Training Centre were used to examine the predicting effects of the 5 behavioral variables of Play and Passive situation tests on the training success rate for guide dogs.Results: In the process of verifying the behavioral variables of Ghosts test,according to the results of statistical analysis, the “Aggression” variable in Ghosts test showed extremely significant difference between guide dogs and eliminated dogs(P = 0.00),and the “Avoidance” variable showed significant difference(P = 0.02). No significant difference was detected from the scores of the rest three variables(P > 0.05), moreover,guide dogs and eliminated dogs both scored high in the “Staring at ghosts” variable. Toanalyze the distribution of the numbers of guide dogs and eliminated dogs in different score grades, we divided the scores of each variable into 3 grades(1~2 points, 3~4points and 5 points). It showed that for Aggression and Avoidance variables, the guide dogs and eliminated dogs were mainly concentrated in the 1~2 points. For Staring at ghosts variable, the guide dogs and eliminated dogs were mainly concentrated in the3~4 points. For Exploration variable, the guide dogs were mainly concentrated in the3~4 points. For Contact with ghosts variable, the guide dogs and eliminated dogs were similar.In the process of verifying the behavioral variables of Chase and Distance-play tests, according to the results of statistical analysis, the “Play Invitation” variable in Distance-play test showed significant difference between guide dogs and eliminated dogs(P = 0.045). However, no significant difference was detected for the scores of the rest variables(P > 0.05). To analyze the distribution of the numbers of guide dogs and eliminated dogs in different score grades, we divided the scores of each variable into 3grades(1~2 points, 3~4 points and 5 points). It showed that for Play Invitation variable,the eliminated dogs were mainly concentrated in 1~2 points. For Aggression variable,the guide dogs and eliminated dogs were mainly concentrated in the 1~2 points. For Following I variable, more guide dogs distributed in 5 points grade compared to eliminated dogs. For Grabbing II variable, eliminated dogs were mainly concentrated in3~4 points. For Exploration variable, the guide dogs were mainly concentrated in the3~4 points. In comparison, the eliminated dogs were mainly concentrated in 1~2 points.In the process of verifying the behavioral variables of Play and Passive situation tests, according to the results of statistical analysis, the “Activity I” variable in the Passive situation test showed higher significant difference between guide dogs and eliminated dogs(P = 0.003), and the “Activity II” variable in the Passive situation test showed significant difference(P = 0.014). However, no significant difference was detected for the rest variables in Play test between guide dogs and eliminated dogs(P >0.05). While, there were clear trends showed that the scores of eliminated dogs were higher that that of guide dogs in variable of Interest in play. The distribution analysis ofthe numbers of guide dogs and eliminated dogs in 3 different score grades(1~2 points,3~4 points and 5 points) of each variable showed:for Interest in play variable, more guide dogs distributed in 1~2 points and less guide dogs distributed in 5 points compared to eliminated dogs; for Grabbing variable, the distribution of guide dogs and eliminated dogs had a similar trend; for Tug-of-war variable, more eliminated dogs distributed in 1~2 points and 5 points grade compared to guide dogs; for Activity I and Activity II variables, more guide dogs distributed in 1~2 points and less guide dogs distributed in 3~4 points compared to eliminated dogs.Conclusions: We have successfully established and improved the methods of ten behavioral tests and the evaluation criterion for 33 behavioral variables of DMA. The“Aggression” and the “Avoidance” variables in the Ghosts test, the “Play Invitation”variable in the Distance-play test, and the “Activity I” and “Activity II” variables in the Passive situation test are confirmed as effective attention testing variable indexes to predict the training success rate for guide dogs.