PROBLEM Cattell ( 4 ) recently compared the second-order personality factors found in children with those in adults and firm support was again obtained towards the identity and replicability of the two major secondary factors found in the questionnaire realm. I n earlier work Cattell(2) identified four second-order factors based on a large sample of Air Force young men in training and male and female college students. I n an attempt t o replicate these factors in a different sample, Karson and Pool@) identified six second-order factors in a clinical sample of 71 U. S. Air Force flyers. There was substantial agreement with regard to the two major factors which emerged in both of these studies, namely, Anxiety-us.-Dynamic Integration and Extraversion-vs.-Introversion, but none of the remaining factors were successfully replicated. The replication of the two major factors was, however, quite convincing in view of the differences in samples as well as in the actual methodology of the factor analysis.The present study seeks t o clarify the meaning of the primary factors in the 16 p . F. test, and t o identify the second-order personality factors found in a sample of physically and mentally healthy U. S. Air Force airmen undergoing recruit training. It was believed that this work with late adolescents would help provide a bridge towards an improved understanding of developmental personality factors in children(4), young adults(2) and adults (9). The sample of 96 basic airmen had a mean age in years of 18.5 and a mean Otis I& of 98.59, S. D. 15.00. This sample was randomly selected from a large pool of several thousand available basic airmen. It is believed that the sample used in the study was representative of the airmen in training a t Lackland Air Force Base, Texas in 1956 with respect to age, I& and major personality traits.
METHODA psychologist-officer administered the Otis Self-Administering Test of Intelligence, Form C and Forms A and B of the 16 Personality Factor Questionnaire. He reported excellent cooperation on the part of the airmen as judged by their relative speed and apparent interest in completing the tasks. The tests were scored according to directions given in their respective manuals. The raw scores for Forms A and B were converted into standard scores by using the sample of male college students and high school seniors whose average age is 21 as the standardization group. These sten scores were then intercorrelated using Pearson product moment correlation coefficients t o achieve the matrix of intercorrelations in Table 1.The second-order centroid factor analysis was completed using machine methods for factoring the matrix of inter correlation^.^^ Of the unrotated factor matrix, 'Grateful acknowledgment is expressed to Kenneth B. Pool for his aid in deriving the original test score matrix of intercorrelations. This data was collected while the author was stationed at the School of Aviation Medicine, U. S. A. F. T h i s study would not have been completed without the interest and cooperation of Prof...