Measurement of Speech Disfluency through Magnitude Estimation and Interval Scaling

Abstract

The purpose was to assess whether equal-appearing interval or magnitude-estimation scaling resulted in a data set with a closer correlation to the physical stimuli, made up of speech samples with varying amounts of disfluency. 20 young adults completed two tasks. In Task 1, subjects used a 7-point equal-appearing interval scale to rate the disfluency of 10 speech samples having varying within sentence pause, presented randomly at 65 dB SPL. In Task 2, subjects used magnitude-estimation scaling to rate these stimuli, presented in a randomized order. Analysis showed significantly high correlations for both scaling techniques and the speech stimuli (Spearman rho=.90 and 1.00, respectively). It appears that subjects can use either scaling technique to rate accurately varied speech disfluency.

Comparing Construct Definition in the Angoff and Objective Standard Setting Models

“…This process permits judges to create their own definition rather imposing a definition on them. Evidence exists in the field of psychometrics, speech and language pathology, and social sciences that individuals are able to use self-defined scales to make reliable judgments (Beltyukova, Stone, & Fox, 2008; Eadie & Doyle, 2002; McColl & Fucci, 1999, 2006; Zraick & Liss, 2000). Self-defined scales have also been found to yield precise outcomes (Beltyukova, Stone, & Ellis, 2008; Beltyukova, Stone, & Fox, 2008; Lodge, 1981; Southwood & Flege, 1999), sometimes more precise compared with when a predefined scale was imposed on an individual (Beltyukova, Stone & Fox, 2008; Southwood & Flege, 1999).…”

A Concurrent Mixed Methods Approach to Examining the Quantitative and Qualitative Meaningfulness of Absolute Magnitude Estimation Scales in Survey Research

“…Instructions provided by researchers for making magnitude estimations vary greatly across the literature, making it impossible to compare findings across studies or inferences about the validity of MES as a scaling method. Most researchers follow Stevens’s original instructions, allowing respondents to use whole numbers, fractions, and decimals (Algom & Marks, 1984; Beltyukova, Stone, & Ellis, 2008; Beltyukova, Stone, & Fox, 2008; McColl & Fucci, 2006; Walker, 2002). Additionally, there is some consistency that lower ratings indicate less and higher ratings indicate more.…”

“…Variations of instruction include permitting respondents to use zero (Algom & Marks, 1984; Hamblin, 1974; Tesler et al, 1991) or not (Bard, Robertson, & Sorace, 1996; Bass, Cascio, & O’Connor, 1974; McColl & Fucci, 2006; Walker, 2002). Instructing respondents to be spontaneous (Eisenberg & Barry, 1988; Fucci et al, 1996), to rate stimuli independently of other stimuli (Eisenberg & Barry, 1988; Fucci, Domyan, Ellis, & Harris, 1994, Fucci et al, 1996; Horowitz, 1981), and not to use numbers twice (McColl & Fucci, 1999, 2006; Schriesheim & Gardiner, 1992). Fucci et al (1994) went so far as to instruct respondents “not to apply any general rules previously learned about numbers” (p. 349).…”

“…In absolute MES, respondents are instructed to provide a number representing their subjective judgment of the magnitude of the stimulus, not in proportion to other stimuli (Hellman & Zwislocki, 1963;Zwislocki & Goodman, 1980). Respondents are instructed to assign any number they want to each stimulus (see, e.g., Fucci, Petrosino, Banks, Zaums, & Wilcox, 1996;McColl & Fucci, 2006;Zwislocki & Goodman, 1980). The idea behind absolute MES is to provide respondents the freedom to choose numbers that represent how they feel and in turn make meaningful discriminations.…”