Personality research of today applies basically inventories having neither unambiguously interpretable items nor responses. The substantive process of generating the test answer is rarely investigated and thus the possible field of meanings, out of which the answer is created, remains hidden. In order to investigate the possible array of spontaneous answers to personality test items, a situative open-ended personality inventory was developed to determine individuals' ways of interpreting personality test items and relevant personality descriptions for individuals. The children's sample (N = 704 of 10-13 year olds) answered five free-response contextualized personality test questions, each related to one of the Five Factor Model personality dimensions. It was revealed that there is no universal interpretation of an item. First, different children's answers to same question described different personality dimensions - substantial number of the respondents' answers did not reflect the personality domain assumed in an item. So there are several ways to interpret test questions; answers may refer to different personality dimensions and not necessarily the one assumed by the researcher. Second, a number of children mentioned more than one personality trait for one item, indicating that even within one person there may be several relevant interpretations of the same item. Considering personality traits as occurring one by one and mutually exclusively during personality test answering may be artificial; in reality trait combinations may reflect actual reaction. In sum, the results suggest there is no single predictable interpretational trajectory in meaning construction process if semiotically mediated constructs, e.g., personality reflection, are assessed.