Abstract
The PhD. thesis was made within the line of research of the GICEC (Group of Research of Concepts in Science Teaching) and concentrates on the methodology of analysing texts.
Textbooks are the carriers of information, of the contents on the subjects and therefore of the concepts. They are one of the basic pillars of the teaching practice, at any level. They are designed to be the mediators between educational policies reflected in content standards and teachers, and have therefore a great impact on what happens in a classroom. They are written by knowledgeable professionals. The content analysis of texts will therefore provide the network of concepts on which the teachers should focus.
The GICEC has developed a software, PAFE (Software for Frequency Analysis and Environment Studies), to be used in the content analysis of texts, centred in the analysis of words that express concepts. This research focuses on the improvement, justification and limitations of the methodology used for determining frequent vocabularies, specific vocabularies, collocations and relations between lexical units (words or groups of words). This requires the use of the software PAFE and techniques based on corpus linguistics for detecting specific vocabularies, frequent vocabularies and collocations. The methodology has been tested applying it to a particular text of more of 1000 words: a text of Astronomy about stars made by the Hubble European Space Agency Information Centre (ESA/ESO) for Secondary Education.
The frequency data of the lexical units obtained with the PAFE software is used to propose a numerical way of obtaining frequent vocabularies based on the analysis of the absolute frequencies distribution. Specific vocabularies are found comparing the relative frequencies of a given text with the ones of a general corpus following techniques developed in corpus linguistics. Collocations are found comparing frequencies and relations between lexical units. Frequencies and relations are then used to build the semantic networks enriched with the previously detected information.
The research concludes the need of: removing all functional words; unifying synonyms spellings only for words which are in the high and medium intervals; not removing the mathematical language due to its importance in frequencies and semantic networks; an expert in the subject being involved in the analysis in order to correct the data obtained when finding collocations and specific vocabulary; the mathematical determination of the frequent vocabulary; determining the optimal system analysing the conservation of relations for each lemma and the conservation of total relations within the system; and enriching semantic networks with the data about specific vocabularies and relation conservation.
Moreover possible errors in the methodology are established, the limitations that affect the results are analysed and possible errors in each of the phases of the methodology are quantified.
References
Hansen-Ruiz, C. S. (2013). Conceptos relacionados con estrella. Lingüística de corpus de astronomía. Universidad de La Laguna. Spain. Available in Spanish at Dialnet database: http://dialnet.unirioja.es/servlet/tesis?codigo=24261
Hansen-Ruiz, C. S. (2012). Conceptos relacionados con estrella. Lingüística de corpus de astronomía. Enseñanza de las Ciencias 30 (3), pp. 325-326.
Ceballos, J. P., Hansen-Ruiz, C. S., Galotti, A., & Stengler, E. (2005). What Science Concepts Appear in Popular Science Magazines? How are They Related? The Case of Stars. In T. J. Mahoney (Ed.), Communicating Astronomy (pp. 199-203). La Laguna, España: Instituto de Astrofísica de Canarias.
Ceballos, J. P., Hansen, C., & Stengler, E. (1999). Las estrellas en las revistas de divulgación científica. In B. León, A. Izaguirre, & A. Martínez (Ed.), Divulgar la ciencia. Actas XIV Jornadas Internacionales de Comunicación (pp. 293-303). Pamplona, España: Ediciones Eunate.
Correspondance
Cristina Silvia Hansen Ruiz
GICEC – Departamento de Didácticas Específicas
Universidad de La Laguna
38200 – La Laguna (Tenerife)
SPAIN
e-mail: chansen@ull.es , cs.hansenruiz@cantab.net
