Short paper
This paper aims to analyze the field of vocabulary CALL in terms of the influences as they
can be observed through a bibliometric approach, using a method called Inclusive Author
Co-citation. This is based on the assumption that authors who are frequently cited
together work in similar areas. The aim of this method is to make visible clusters of
authors who work within the same subfield, presumably influencing each other’s work.
The focus is thus more on communities of research than on individual papers and their
citation numbers.
-340-
2014 CALL Conference
LINGUAPOLIS
www.antwerpcall.be
The database that forms the source material for the bibliometric analysis consists of over
100 papers that have appeared in the journals CALL, ReCALL, LT&T, and Calico and that
fall into the area of vocabulary teaching and learning within CALL. This includes work that
examines how glosses can improve learners’ vocabulary, whether electronic dictionaries
improve learners’ results, as well as work on vocabulary trainers, and on collocation in
the field of CALL.
If we count the authors of all papers separately, i.e. disregarding the distinction between
single-authored and co-authored papers, we arrive at over 160 names. Of these, the vast
majority appear only once. Only about 20 authors have more than one paper in the
corpus examined, and none has more than four. This pattern is quite similar to what
Meara (2012) found for vocabulary acquisition studies in general. Considering citations
next, the methodology employed for this analysis again focuses on authors rather than
individual publications. So if a paper cites several works by one author, this still only
counts as one citation. This principle is meant to give relatively more weight to an
author’s general influence in the field than to a single much
-cited paper. To give a
simplified example, if author A cites authors B, C, D and E in the bibliography, the raw
data will be a file [Author_A, Author_B, Author_C, Author_D, Author_E]. From this, the
following co-citations can be generated for this one paper:
Author_B : Author_C
Author_B : Author_D
Author_B : Author_E
Author_C : Author_D
Author_C : Author_E
Author_D : Author_E
This is repeated for every paper in the corpus; and then the various co-citations can
simply be counted to see which ones are more or less frequent. A threshold of between
three and five co-citations will typically produce interesting results that can be
interpreted in terms of influences among researchers and groups of researchers. Using
the corpus of vocabulary CALL papers described above, such an analysis can be used to
show that the field of vocabulary CALL has good links to the general applied linguistics
research on vocabulary, but most papers cite a limited number of major authors from the
mainstream, i.e. mostly ISP Nation, B. Laufer and a few more. The psycholinguistic
strand of vocabulary acquisition is much less well represented in CALL. We can also see a
certain grouping within vocabulary CALL, the clearest of which is a group of authors
working on glosses. An example of a graphical representation of this data can be seen in
Figure 1.
The main methodological problems with this approach concern the setting of appropriate
thresholds for inclusion in the graphs, and the one shown as Figure 1 is only one possible
among many. Once the data has all been collected and entered, a certain amount of
interpretation is needed to set the threshold, and then again to interpret the results.
-341-
2014 CALL Conference
LINGUAPOLIS
www.antwerpcall.be
Figure 1: A graphical representation of bibliometric data for vocabulary CALL
Do'stlaringiz bilan baham: |