Since developmental and other personality research hinges on structured measurement, in which Q-data is most used of the three media, a check on the nature of factors stated by present theory is urgent. Programmatic research—indispensable in factor analysis—over 20 years has indicated 18 to 20 primary factors, for 16 of which at the adult level and 14 at the high school level scales have been constructed (the 16 PF and HSPQ). Taking 560 items in the 56 10-item HSPQ scales (Forms A, B, C and D for each factor), and administering them to 399 girls and 401 boys of mean ages 13·6 and 14·3 years respectively, the experimenters set out to answer the following: (1) How many primary factors appear? (2) Are they measured by the scales as now labelled? (3) If so, to what validity? Regarding (1), the scree test showed 14 matchable factors. At a blind Oblimax and Rotoplot simple structure, reaching, in both, 73 per cent in the ±0·10 hyperplane, all but factors C, J and Q2 reach Bargmann significance of structure. Congruences of resolutions across the samples were beyond P < 0·01 on all factors. Validities, lowest on E and O, were above 0·9 for more than half, and 0·8 for the rest. Second-order results and theoretical indications for research are given elsewhere.