Stimuli

7.1.6.1 Kendall et al. (2005) Nonword Stimuli

The purpose of Kendall et al.’s original study was to evaluate the effects of interphonemic transitional gestural frequency on phonetic encoding. Three groups of stimuli were created controlling for the parameters listed in Table 8, which included varying the interphonemic transitional gestural frequency across three categories: high, moderate, and low gestural frequency of occurrence (see Table 10). Participants had faster reaction times when articulating nonwords with high gestural frequencies compared to stimuli with moderate or low gestural frequencies. No statistical difference in reaction times was noted between the moderate and low gestural frequency stimuli. Replication of these reaction time effects in a younger cohort will support collapsing the moderate and low frequency stimuli into a single stimuli set for the main experiment.

Table 10: Kendall et al. (2005) nonword stimuli

7.1.6.2 Kendall et al. (2005) Real Word Stimuli

Kendall et al. utilized real words in their original study as a control in case non-significant differences were noted between nonword frequency groups. Differences in reaction time between high- and low-frequency words were a predictable and replicated finding, and replication of frequency effects in real words would indicate negative findings in the nonword stimuli were not due to experimental error or unusual participant performance (Kendall et al., 2005). Real word stimuli, noted in Table 11, were extracted from Balota and Chumbley’s (1985) lexical access study, and provided a predictable set of stimuli to evaluate lexical access frequency in real words.

Table 11: Kendall et al. (2005) real word stimuli High-Frequency Words Low-Frequency Words

office nostril mountain trinket student trestle baseball lobster chicken gazelle captain freckle kitchen pebble village pelvis teacher weasel college rudder machine sequin valley gasket cousin beaker yellow nylon market banjo forest tunic cotton navel garden baron coffee ladle window silo 7.1.6.3 Different-Phonemes-Different-Motor-Class (DPDC) Stimuli

The DPDC stimuli were untrained stimuli created by the author for the main experiment (see Table 12). The purpose of piloting the DPDC stimuli was to provide a baseline of the reaction times associated with articulating these nonwords compared to the Kendall et al. (2005) stimuli. Descriptions about the construction of these stimuli are found in Section 7.2.7.6.

Table 12: DPDC Stimuli

7.1.6.4 Stimuli Preparation

As noted above, the Kendall et al. (2005) stimuli were not equal in mean length duration, which potentially biased participants’ reaction times on the high frequency stimuli. To decrease the potential effects of final syllable lengthening associated with the original stimuli, the Kendall et al. stimuli were re-recorded to control for mean length duration. Extensive perceptual reliability checks of the re-recorded stimuli were conducted to ensure perceptual accuracy of the nine phonemes in each nonword. Recording and perceptual ratings of phonetic accuracy will be described separately below.

Stimuli Recording

All stimuli were recorded by an individual knowledgeable in phonetic transcription and trained to maintain a slowed, natural rate of speech. All recordings were completed in a sound booth using a USB Logitech desktop microphone (Model 980186-0403) and recorded through Adobe Audition digital recording software (version 3.0) on a Macbook Pro utilizing Windows 7 operating system. Syllable stress was recorded on the first syllable for each stimulus, which is identical to Kendall et al.’s (2005) original methods.

Stimuli Perceptual Reliability

Phonetic discrimination of all nonword stimuli was conducted by three individuals trained in phonetic transcription to ensure phonetic accuracy of the recordings. Each transcriber had completed a course in descriptive phonetics and routinely transcribed in her research or clinical duties. Transcribers listened to stimuli in a quiet place while using headphones, and played the stimuli through a standard operating system player (e.g., Windows Media Player or iTunes). Several rounds of perceptual phonetic accuracy ratings were conducted in which each transcriber manually transcribed each nonword using the International Phonetic Alphabet (“IPA Chart,” 2005). Two out of three transcribers needed to accurately transcribe each nonword (total of nine phonemes) for the stimulus to be accepted for stimulus validation. Stimuli not transcribed accurately by two of the three transcribers were re-recorded and further perceptual assessments were conducted.

A forced-choice questionnaire was implemented following several rounds of manual transcription and re-recording of the nonword stimuli. Transcribers listened to three nonwords and selected the correct written IPA transcription from two foils with difficult to perceive consonant contrasts. These additional ratings were deemed sensitive in detecting phonetic accuracy of all nine phonemes, especially as the foil contrasts were based on difficult to detect phonetic contrasts noted during earlier rounds of phonetic transcription attempts (M. Dickey, personal communications, September 13, 2013).

Once the nonword stimuli set was determined to be phonetically accurate, a one-way analysis of variance (ANOVA) on the mean duration as a function of frequency group (high, moderate, low) was performed. No significant differences between duration were observed

between frequency group, F (2, 27) = 2.256, p = .124; means and standard deviations for each group are listed in Table 13.

Table 13: Means and SDs for re-recorded Kendall et al. (2005) stimuli Mean (ms) SD

High Frequency 1254 .066

Moderate Frequency 1267 .083

Low Frequency 1320 .070