ESLORA is a corpus of Spanish made up of semi-directed interviews and spontaneous conversations recorded in Galicia between 2007 and 2015. The design and construction of the corpus meets three objectives: to register the use of a variety of Spanish which to date has been scarcely documented, to gain additional insight into the methods for the construction of spoken corpora, and to develop computational tools for corpus search. The paper presents the main characteristics of ESLORA and the criteria followed in the corpus building process. It also includes a brief description of the tools used to build the corpus and how they work together to achieve the project needs and, moreover, it shows that the decisions taken at various stages of the compilation of the corpus are closely related to the wide range of possibilities for retrieving the lexical, grammatical and contextual information provided by the materials.

Observer's paradox has been pointed out as the main problem when it comes to gathering oral materials in an informal colloquial register. The free interview has been proposed as an alternative to the traditional one, where the objective is to minimize the effects of the paradox, through methodology based in the relaxation of the question-answer structure, interviewer-interviewee distance and differentiated roles. The objective of this study is to carry out a lexical corpus analysis, in order to show the influence of the observer's paradox on the materials obtained through interviews, and to evaluate if the strategies of the free interview are effective in obtaining a register similar to that of a conversation.