The Designing And The Equivalence Analysis Of The Basic Flying Ability Items And Construction A Model Of Predicting Academic Performance

Flying ability is an organic combination of all kinds of psychological characteristics which can help finish the fly successfully. They include ability constructions, personality characteristics and flying motivation. And between ability constructions, there are basic ability and special ability. As a method for detecting the flying basic ability, the flying basic ability tests on SubsystemⅠare the most important content of the psychological examination. The flying basic ability tests include nine items, including"Addition calculation","Comparing simulated scales","Detecting special graph","Assembling figures","Simulating keyboard input","Recognition regulation","Distinguishing direction","Selecting paired word","Regularity array". They mainly inspect the intelligence level such as sensibility, perception, memory and thoughts. Our research is using the nine flying basic ability tests as the research contents. Firstly, designing four new items for each test based on the former items, analysising the difference, solving the problem of leaking items'possibility induced by using the tests again and again. Secondly, To establish a model of predicting academical performance by using basic ability tests of pilots'selection and 16PF for eliminating the learning disability students in advance, and for saving expenses for pilots'selection. Meanwhile, This research also compare the equipment difference between DXC-W and DXC-V for extending the equipment in the air force or other army forces, and compare the sex difference between the female and male for giving information to air force when they enroll the female pilots.There were mainly two parts in the reaserch:1.The preparatory research of constructing new items of basic ability test on subsystemⅠObjective: To compare the new items with the former item; to contrast the DXC-W to the DXC-V; to analyze the gender factor in basic ability test. Methods: Using completely random design, one thousand two hundred and twenty three-month college students were investigated by taking nine basic ability tests. The difference between the new items and the former item were examined by multiple comparison method; the equipments difference and the gender factor were discussed after t-test. Results:Except"Distinguishing direction"(T3&T4),"Selecting paired word"(T1&T2)and"Regularity array"(T2&T3&T4), there were no significant difference between the new items and the former item(P>0.05). No significant difference was found between the two types of equipments(P>0.05). Gender factor had significant difference on tests which were"Addition calculation"(t=2.760,P<0.01),"Comparing simulated scales"( t=-2.629 , P<0.01 ) ,"Assembling figures"( t=-2.524 , P<0.05 ) and"Distinguishing direction"(t=-3.673,P<0.01). Conclusions: The new items of basic ability test need further consummated. DXC-W could be generalized. The sex difference of tests should be paid more attention. 2.Construction a composite model of predicting academic performance(1) The relativity research of cognitive ability,Cattell's personality traits and academic performanceObjective: From the aspects of cognitive ability (CA) and personality traits (PT), the psychological factors which affect the academic performance (AP) of the college students are to be discussed. Methods: 3-months College students were investigated by using 9 ability tests and 16PF. Results: 4 ability tests which are"Addition calculation","Detecting special graph","Simulating keyboard input"and"Selecting paired word"are correlated substantially with academic performance; When factor B is partialled out, Cattell's personality traits increase 2.1%(14.6%-12.5%) of the variance on mathematics, 0.7%(9.9%-9.2%) on literature, 0.9%(13.0%-12.1%) on whole score. Conclusion: CA and PT are both significant predictor, and maybe there is a little overlap between them.(2) Construction a composite model of predicting academic performance using CA and PTObjective:To establish a model of predicting AP by using pilots'selection basic ability tests and 16PF for eliminating the learning disability students in advance, and to save expenses for pilots'selection. Methods:Using completely random design, 3-months College students were investigated by using 16PF and 4 ability tests which are"Addition calculation"(JS),"Detecting special graph"(TX),"Simulating keyboard input"(TJ),"Selecting paired word"(XD). Making use of Pearson correlation, which analysis correlations among the CA,PT and AP. Then, using multiple linear regressions, we constructed a model both of CA and PT predicting AP, and analyzed its predictive effect. Results: Using CA as independent variable, the regression equation is Y∧=526.75+1.00JS+1.59TJ+ 0.14TX+1.35XD,which CA can explain 17.2% of variance on AP;Using PT as independent variable, the regression equation is Y∧=589.23-2.179A+0.633C- 1.666E-1.267F+1.517G+1.677Q1,which the PT can explain 12.4% of variance on AP; Combining CA and PT as independent variables, the regression equation isY∧=549.0-1.469A+0.402C-0.813E-1.217F+1.420G+2.769Q1+0.705JS+1.984TJ+ 0.873XD,which both are explain 28.5% of variance on AP. Conclusion: There is a little overlap between CA and PT, and the research method of predicting AP through CA and PT is available.