Referencia: Aliaga, Francisco M. (1999). Correspondence Analysis: a bibliometric study on their use in the educationa research . RELIEVE, vol 5, n. 1. Consultado en http://www.uv.es/RELIEVE/v5n1/RELIEVEv5n1_1.htm en (poner fecha).
This
article has been consulted sometimes
from the date of its publication
RELIEVE -Revista ELectrónica
de Investigación y Evaluación Educativa // 1999 // Volume
5 // Nunber 1_1
ISSN 1134-4032 // D.L. SE-1138-94
CorreSpondence analysis: A bibliometric study on their use in the educational research
Análisis de
correpondencias: estudio bibliométrico sobre su uso en la investigación
educativa
by
Francisco M. Aliaga
University of Valencia
Resumen
Revisamos
la difusión de las técnicas de Análisis de Correspondencia entre distintas
disciplinas y medios culturales, a partir del análisis de diversas bases de
datos académicas. Los resultados muestran una muy escasa implantación en el
ámbito de la investigación educativa, y muy en particular en la española.
Palabras
clave: Análisis de Correspondencia, bibliometría, datos
cualitativos
Abstract
We revise the diffusion of the techniques of Analysis
of Correspondence between different disciplines and cultural environments, analysing
several academic databases. The results show a very scarce installation in the
educational research, and specially in the Spaniard.
Key
words: Analysis of Correspondence, bibliometrics, qualitative
data
Introduction
One
first elementary approach to the knowledge of some topic of researchconsists of
making a bibliométrico approach to the same one. With it we were not only
able to know the importance in specialized Literature, but also to make a
pursuit of its evolution, particularly interesting with regard to the last years,
inasmuch as it can reveal the increase or diminution to us of the interest waked
up by this subject.
In
our case we are interested in knowing the diffusion a concrete technique of
analysis of data, the Correspondence Analysis . One is a procedure that allows
to study the relation between data or qualitative variables (Henry, 1997). One
is a type of analysis created by Benzecri (1963) whose purpose is to graphically
represent the originating data of a contingency table.
With
the objective mentioned one, we will analyze the use of this technique in
different contexts, linguistic-cultural as in disciplines as much varied, in
order to see its implantation differential.
Since
we are working in surroundings of educative research, any approach of this type
has to count, necessarily, with the use of Base de Datos ERIC, who is the one
that has a greater record number on electronic publishings (RIE
and CIJE). It is evident that in the case of the researchon a topic as is
the one of Correspondence Analysis , a procedure that had their origin and main
impulse in the context of the French school of analysis of data, to resort to a
data base of marked slanted Anglo-Saxon can imply certain limitations.
Nevertheless, we will try to diminish this problematic one, as we see more ahead,
with the joint use of other sources of intelligence that can complement the
extracted information of ERIC.
Results
For
the search in the data base ERIC we have used the WebSPIRS program, that allows
an efficient interface with the data,
simultaneously that its use in network, so and as has organized the access to
ERIC in the University of Valencia. In concrete we have limited our researchthe
data of the ERICS who go from 1982 to the update of June of 1998, because the
information contributed by older data would little give an excellent information
at the present time. In any case, the reference older than we have foundÂ
talks about publications of year 1985.
We
have made in ERIC a search of the term "Correspondence Analysis"
that has shown a total to us of 23 references
Graph
1 - ERIC - References on "Correspondence Analysis"
1982-6/1998
As we can verify, the document number that includes references to the Analyses of Correspondences is certainly low, and enough uniform, if we excepted the 1988 tip.
The
fact that is specially significant something more than half of publications
(52'2%) has appeared in an only magazine of noticeable methodologic character, Psychometrika.
If to it we added that, although to enough distance, second source by importance
order is also another methodologic magazine (Multivariate
and Behavioral Research, with three references), we have them two third
parts of publications referenciadas in the ERICS who use or analyze the
technique of the multiple comparisons confine themselves to periodic
publications of character almost exclusively methodologic of the predominant
scope of psychology. On the other hand, only three of the references have been
published in magazines that could be considered applied (Child
Development, Rehabilitation and
Counseling Bulletin and the Journal
of for Marketing Higher Education), being the unique ones that presents/displays
empirical works applying the technique of the Correspondence Analysis . It is
possible to emphasize other two references of communications to Congresses (one
of them originally in French, and the other an empirical study) and other three
of documents of different institutions (some in Dutch), of which only one (South
African) refers an empirical study.
The
use of the Multiple Correspondence Analysis is still more little,
appearing references to this technique only in four of the 23 studies selected,
treating the subject from a methodologic point of view. Of them three appear
have been published in the Psychometrika magazine, and peculiarly, three
of the four are written by authors of Dutch origin
Of
the panorama that we finished reviewing, so and as us ERIC raises, we can
conclude that she in the last exists a use very little in applied works of the
technique of the Correspondence Analysis twelve years, at least in the
Literature (mainly Anglo-Saxon) that is considered in this data base. The most
common publications talk about theoretical developments of the own analytical
technique, although its number also very is restricted, and are practically
limited a same publication (Psychometrika). The reference to the Analysis
of multiple correspondence is still more little, limiting itself four
publications, all of them theoretical ones, and most of nonAnglo-Saxon origin.
Practical applications of the procedure reviewed to any properly educative study,
but rather of psychological cut almost do not exist (excepting one on reasons
for election of the university from the point of view of marketing and another
one on determination of points of cut in the access to the university).
Due
to the shortage of properly educative studies, and to that most of originating
references of our search in ERIC they had a psychological profile rather, we
decided to review a more specific data base of this discipline, to try to
clarify the type of use that is made in Human Sciences of the Analysis of
Correspondence. We went to PsycLIT, commercial name of it data bases PsycINFO,
sponsored by American Psychological Association (APA), that includes
information coming from several sub-you base, such as the
Psychological Abastracts and the
Dissertation Abstracts International, and that empty more than 1300
magazines, in addition to another gray Literature (communications to congresses,
administrative documents, etc.). The version that we have used includes from
1988 to September of 1998. We have looked for the term "Correspondence
Analysis" in anyone of the fields (titles, description, summary, etc.),
having obtained a total of 110 references.
Graph
2 - PsychLIT
References
on "Correspondence Analysis"
1988-9/1998
First
which it is possible to stand out is the much more frequent use of the Correspondence Analysis
, either like object of theoretical development or like applied procedure, in
the scope of the magazines of Psychology that of those of PedagogÃa,
occurring the circumstance that, if we did not consider to the Psychometrika
magazine (of noticeable methodologic character and from one more a psychological
perspective), there are years in which PsycLIT finds the more references that
ERIC, the greater educative data base, in all one decade. Since both are
centered fundamentally in the Anglo-Saxon cultural scope, the result would have
to make us reflect on the very moderate diffusion that this procedure has
reached in the educative research.
Has
to do to notice, in addition, that of 110 references found in PsycLIT, only 30
are of works methodologic, whereas the rest (almost 73%) is applied works in
which this analytical technique has been used, frequently along with others,
values that also contrast with hardly 33% of references of works applied in the
scope of the education, so and as it is deduced of the data that we have
presented/displayed previously of ERIC.
Graph
3 - PsychLIT
Relation
between methodologic and applied articles
As
we can verify in Graph 3, except the first year, exists a notable (and even
increasing) predominance of the works applied on the theoreticians the
psychological texts, like on the other hand it would be to hope in a discipline
that is not predominantly methodologic. Since we have commented previously, the
referring data to year 98 are partial, reason why they do not have to be
considered to establish the tendency. The results obtained in PsycLIT on the
superiority of the studies applied on the theoreticians contrast vividly with
whom we have commented previously originating of the educative scope (ERIC).
Since
the previous data bases, although are the greater ones of the world in their
respective scopes, are centered in a cultural scope something distant to ours,
we have seted out to study the use of the Correspondence Analysis in our
country. For it we have gone to the data base ISOC, produced by the Superior
Council of Scientific researches through
its information center and Documentacio'n Cienti'fica (CINDOC), being centered
in Spanish magazines of Humanities and Social Sciences, draining more than 1600.
One is a compound data base as well by the information of diverse specialized
data bases more, such as ECOSOC (that it covers subjects with economy, sociology
and policy), ISOC-DC (on subjects of scientific Documentation), HISTORY (on
auxiliary history and sciences) or the PSEDISOC, most interesting for our
scope of work, that covers subjects with Psychology, Sociology and Sciences of
the Education.
The
search on the subject "Correspondence Analysis " has given a result of
32 references, that per years present/display the tendency that is in Graph 4.
Graphical
4 - ISOC
Publications
on Correspondence Analysis
As
we can see, there is a slight ascending tendency in publications on Correspondence Analysis
, that undergoes a slight reduction in years 96 and 97, after the great tip
produced in year 95. This apparent reduction can be explained, partly at least,
because the referring publications to these last years have not been considered
completely still.
The
analysis of the use that makes of the Correspondence Analysis the
different disciplines from Humanities and Social Sciences shows Graph 5. In the
same one we can observe that there are eight different Social or Human Sciences
(although two of them closely are related to Natural Sciences) that have gone to
this procedure of analysis of data, which gives idea of its versatility.
The
specialty that uses of more habitual way the procedure of Correspondence Analysis
is Psychology (we have including in this discipline a publication of clinical
psychiatry, to the being the scope closest in Social Sciences), something
totally coherent with the ample use revealed in the PsycLIT. Second the science
social that more uses east type of procedures is Economy, that in addition is
one of the first disciplines that used it, already in the decade of the 80.
Graph
5 - ISOC
Publications
on Correspondence Analysis
by disciplines
Next
is Arqueologi'a-Antropologi'a, with the same number of publications of Correspondence Analysis
that the Education, although we have put it in the first place for antiquity
reasons, since publications exist that use this type of analysis of data from
1986.
Other
disciplines that use this type of procedures, althoughÂ
to a lesser extent, are Geography (2), the linguistic one (1), the
ecology and the statistic (1). This ample variety of disciplines that make use
of the techniques of Correspondence Analysis demonstrates to their
flexibility and adaptation to a great variety of circumstances, data and
purposes.
Reason
why it makes reference to the publications of education that include articles on
or with Correspondence Analysis ,
since we have already said are limited five (although the data bases of the ISOC
specifically do not collect on the
used analytical techniques, reason why all the cases have perhaps not taken
shelter). They are distributed temporarily at the rate of a publication per year
from 1990 to 1995, except 1991, in which no of them was published (Carballo,
1990; Grandson, 1992; EtxeberrÃa, 1993; Gil, Garci'a and Rodriguez, the 1994
and Greater one, 1995). That is to say, the use of this type of procedures does
not respond to no "precise fashion", but that shows a very gradual
diffusion. Nevertheless, it also emphasizes that in the most recent years
articles have not been published on
or with Correspondence Analysis ,
With
respect to the publications in which they have appeared these articles, it
stands out excessively that they are concentrated exclusively in two titles, the
Revista
de Investigación Educativa (Review
of Educative Research),
the three older, and Enseñanza - Anuario
Interuniversitario de Didáctica
(Education - Interuniversitary
Yearbook of
Didactics),
both more recent. First it is the organ of expression of the Interuniversitary
Association of Pedagogic Research ( Asociación
Interuniversitaria de Investigación en Pedagogía, AIDIPE), the one that
probably is the professional association in which they appear a greater amount
of metodologists of educative research. The second publication is located in
Seville, locality that is parallelly the center of one of the most active nuclei
of our country in qualitative researchapplied to the education.
Conclusions
Scientific
Literature has been only reviewed using data bases on periodic publications,
which mainly leaves of side works that can be important, from the theoretical
and/or didactic point of view (for example, Cornejo, 1988). Nevertheless, we
think that the vision that contributes east type to us of information, with
respect to the use of the Correspondence Analysis , is sufficient to do an idea
to us of its diffusion and use, that is objective of the present work.
In
summary, we can affirm that the publications on
or with Correspondence Analysis
in the Spanish scope educative are extraordinarily little and that, in addition,
they are centered in very few publications (only two) very specialized, and
oriented, in good average, to specialists in educative researchmore than to
professionals. These data are coherent with the results derived from the
analysis of Anglo-Saxon data bases, in which a little use of this type of
procedures in the educative researchis revealed, in any case far below to the
made one by other near disciplines. We considered simultaneously that he would
be advisable to spread the use of this type of analytical techniques of a ampler
way, than the knowledge of the same ones would be due to plan, at least of its
interpretation, between the professionals of the education, in order that these
also could take advantage of their evident advantages. The diffusion of simpler
computer science procedures to make the analyses can fulfill, in this sense, a
facilitador effect that there is to consider very.
Bibliography
Benzecri, J.P. (1963). Cours de linguistique mathématique.
Rennes: Université de Rennes.
Carballo, Rafael (1990). Algunas aplicaciones del análisis de
correspondencias a la interpretación de tablas de contingencia en la evaluación
de un programa de formación profesional ocupacional. Revista de Investigación Educativa, n 16, pp. 537-550.
Cornejo Alvarez, J.M. (1988). Técnicas de investigación social: el
análisis de correspondencias: teoría y práctica. Barcelona: PPU.
Etxeberría, J. (1993). SPAD.N (integrado): normas de utilización. Revista de Investigación Educativa, nº
22, pp.167-178.
Gil, J., García, E. y Rodríguez, G. (1994). Análisis estadístico y
muestreo en grupos de discusión. Enseñanza
- Anuario Interuniversitario de Didáctica, (12), 213-231
Henry, G. (1997). Correspondence Analysis. En J. P. Keeves (Ed.), Educational Research, Methodology, and
Measurement: an International Handbook. Oxford (UK): Pergamon.
Mayor, C. (1995). Cómo ven los
profesores de distintas titulaciones las condiciones profesionales de la
Universidad de Sevilla: el análisis de correspondencias múltiples. Enseñanza - Anuario Interuniversitario de
Didáctica, (13): 63-88.
Nieto, S. (1992). El "análisis de contenido" como técnica de
investigación documental. Aplicación a unos textos de prensa educativa, y su
interpretación mediante "análisis de correspondencias múltiples". Revista de Investigación Educativa, nº
20, 179-200.
Return to the index of the Vol. 5 N. 1