1993). Multiple choice tests require a non-verbal response and allow Iimitedanalysisof
the representational status of the lexical item. Furthermore, no repeated measurements have
been included to examine word learning over time.
4.3 Aims and rationale for the design of Experiment 1
Experiment 1 was designed to extend the understanding of stories as word learning contexts
by assessing child based and contextual factors. According to Crais (1987) using stories to
introduce new words and observing which aspects are stored and recalled can illuminate what
and how a child Ieams from context. The use of stories to look at fast-mapping skills can be
informative across a wide developmental range, provides a natural context for word
acquisition, motivates the children to attend to the task, and permits the systematic
manipulation of variables affecting the fast-mapping process.
Particularly, the aim of Experiment 1 was to investigate the nature of the linguistic input
(inference, definition, analogy, lexical contrast) in relation to child based factors
(phonological memory and existing vocabulary knowledge) and their age for word learning.
Furthermore, most of the studies demonstrated word learning from stories by using only a
single assessment, mainly a multiple choice test. Experiment 1 aims therefore to demonstrate
word learning in more depth using multiple measurements (Naming and Understanding tasks)
particularly designed for the purposes of Experiment 1.
Additionally, most of the studies measure word learning immediately after the story session,
without measuring whether that learning remains after a longer period. Experiment 1 will
investigate word learning over time by measuring word knowledge both in an Immediate and
a Delayed post test. The rationale for choosing the above parameters is presented in the
following subsections.
Child based factors
Phonological working memory
The role of phonological working memory was investigated since several studies have found
that phonological working memory is related with word learning (Gathercole and Baddeley,
1990; Michas and Henry, 1994; Gathercole et. al. 1997). Nevertheless in these studies, the
89