examined 5 different independent measures: phonetic inventory in the initial
and final positions, word shape, phonemic complexity index, and the syllable
structure level. These measures had already been a part of previous research
and are needed because data regarding the temporal stability of these measures
is missing in the literature. The previous studies examined characteristics of
each measure because the participants were only examined once. This study
employs the same independent measures, but obtains multiple assessments instead
of only one, thus examining the changes across time among the measures instead
of their characteristics. Additionally, this method reduces the chance that
changes are simply due to development because the time frame is only one week.
The goal of
this study was to determine the test-retest reliability (and temporal
stability) in one week on the independent measures. These measures were
calculated by employing intelligible words made throughout a speech sample
lasting twenty minutes.
A way to
determine temporal stability that is often used is test-retest reliability. It
is oftentimes accomplished by associating the score achieved on a certain day
with that same person’s score on a different day. This is important in order to
determine the timing effects on the person’s score. This type of reliability is
crucial to this study because the purpose is to examine the temporal stability
among the independent measures in a 1-week timeframe.
participants were between 18-22 months old.
An important result of
this study is that one speech sample that lasts 20 minutes is not enough to
attain reliable and accurate assessments for final and initial phonetic
inventory as well as word shape analysis. Both IPC & SSL were sound to be
consistent throughout the 20-minute sessions, which is important because it can
help professionals in the field with valid alternatives when examining the
phonological skills among children. Another interesting finding in the study is
that no significant differences were found in any of the five measures between
sessions. However, differences were found between measures regarding
test-retest association values.