Levels of Speech Usage: A Self-Report Scale for Describing How People Use Speech

Abstract

People use speech in a variety of ways to fulfill life roles and responsibilities. Documenting speech usage is critical in clinical work to plan relevant intervention goals for individual clients, and in clinical research to better describe participant characteristics. A few voice-use classification scales exist; however, they are limited in scope (e.g., focus almost exclusively on occupation) and in applicability beyond voice-disordered populations. The Levels of Speech Usage is a self-report categorical rating scale intended for use with adults across a wide range of communication disorders and life situations. This article presents data from the initial analysis of this scale in a sample of 200 people with spasmodic dysphonia (SD). Speech usage was significantly associated with age, education level, and work status (full time, part time, no paid work). Speech usage was not significantly associated with gender, SD duration, self-rating of voice, treatment status, presence of other medical conditions, Voice Handicap Index, or a measure of communicative participation. Further research is needed to explore the function of this scale in other populations.

People use speech in different ways in everyday communication situations. Speech usage varies tremendously depending on each person’s roles and responsibilities. For example, people who are employed may give presentations to large groups, speak over noise in a factory, or talk continuously on the phone. Those who manage a household may talk frequently when caring for young children, talk on the phone to conduct household business, or talk loudly to be heard by hard-of-hearing family members. Many people are involved in community organizations where their speech usage might include performances, public speaking, or group discussions. Because speech usage varies from person to person, it is critical for speech-language pathologists (SLPs) to understand the needs of individual clients for planning relevant intervention goals. Such information also facilitates clinical research, particularly when exploring the outcome of interventions designed to address psychosocial aspects of communication disorders.

Speech usage data can be gathered in a variety of ways. Clinicians will often obtain this information through an interview and report it anecdotally in evaluation and treatment records. A variety of in-depth questionnaires are available, particularly in the voice literature, through which clinicians and researchers can gather detailed information about a client’s history including speech and voice usage (Benninger, Jacobson, & Johnson, 1994; Sataloff, 2005). While these methods provide the detailed information important for clinical decisions, they do not necessarily yield an efficient means to summarize speech usage in a report or to code speech usage as a research variable.

Several authors have proposed categorical scales for rating voice demands, and Table 1 contains a summary of three of these scales (Behrman, Sulica, & He, 2004; Koufman & Blalock, 1991; Vilkman, 2000). Other authors have focused on asking participants to estimate the frequency or duration of vocal activities (Gotaas & Starr, 1993; Roy et al., 2004). Current scales to rate voice demands have a number of limitations. First, the rating scales listed in Table 1 refer almost exclusively to “voice.” While voice is certainly a key component of speech, there are subtle differences between using voice terminology exclusively versus broader speech terminology that might limit the usefulness of these voice demand rating scales for people with other speech concerns including dysarthria, aphasia, apraxia, dysfluency, or even accent modification. Second, the existing rating scales focus almost exclusively on occupational voice demands (although Behrman et al. [2004 1 do include “mothers of young children” and social “big talkers” in the high-demand group). While occupational speech usage is a critical concern for many people, these rating scales do not provide a mechanism for reporting speech usage that may be important to people outside of their occupations. This is of particular concern for people who are not in paid employment or for those who value unpaid activities involving speech (e.g., involvement in community organizations). A third limitation, also related to the emphasis on occupational voice demands, is that voice demands are categorized according to job title. The listing of job titles obscures the considerable variation in speech usage that can occur across different people who might fall under the same title. The final limitation of existing rating scales is they are often completed primarily by clinicians or researchers (although Behrman et al. [2004] suggest that clinicians and clients make the rating jointly). Proxy ratings may not accurately reflect speech use patterns. Obtaining clients’ views of their own speech usage can be very informative in terms of the priorities and motivations that clients have for intervention, particularly because clients may value particular speech activities differently than outside observers.

TABLE 1.

Examples of existing voice demand rating scales.

Koufmann & Blalock (1991)		Vilkman (2000)		Behrman et al. (2004)
Level I: Elite Vocal Performer	Actor; singer	High Quality/High Load	Actor; singer	Professional	Singer, actor, radio personnel, vocal performance students
Level II: Professional Voice User	Clergyman, lecturer	High Quality/Moderate Load	Radio and TV personalities	High	Teachers, sales, tech support, mothers of young children, construction or factories with background noise, social “big talkers”
Level III: Nonvocal Professional	Teacher, lawyer	Moderate Quality/High Load	Teacher; telemarketer	Routine	Everyone else
Level IV: Nonvocal Nonprofessional	Laborer, clerk	Moderate Quality/Moderate	Business personnel, physicians, lawyer, nurse
		Load Low Quality/High Load	Factory workers, machinists in background noise

The Levels of Speech Usage is a self-report scale that can be used to efficiently describe and code speech usage for clinical and research purposes. The Levels of Speech Usage is written to be applicable to people across a wide range of communication disorders and life situations. Speech usage is described in terms of the amount, frequency, type, and importance of speaking situations that people might encounter without reference to specific occupations. The purpose of this article is to introduce the Levels of Speech Usage scale by describing the development and initial analysis of the scale. This article presents statistical data about the distribution of a sample of people with spasmodic dysphonia (SD) across the speech usage categories, and associations between speech usage and other participant demographic variables.

METHODS

Development

The Levels of Speech Usage (Table 2) was developed based on literature review and the experience of a multidisciplinary team of clinicians and researchers. The scale was then refined using cognitive interviewing methods (Willis, 2005) with 28 people with a diagnosis of SD. The scale was revised and reevaluated in an iterative manner similar to that described elsewhere (Baylor, 2007; Yorkston et al., 2008).

TABLE 2.

The Levels of Speech Usage categorical rating scale.

How Do You Use Your Speech?
While communication is important to everyone, different people use their speech in different ways. Think of how you have typically used your speech over the past year. Choose the category below that best describes you.
_____	Undemanding:
	Quiet for long periods of time almost every day:
	Almost never
	talk for long periods raise your voice above a conversational level, participate in group discussions, give a speech or other presentation
_____	Intermittent:
	Quiet for long periods of time on many days
	Most talking is typical conversational speech
	Occasionally:
	talk for longer periods raise voice above conversational level participate in group discussions, give a speech or other presentation
_____	Routine:
	Frequent periods of talking on most days
	Most talking is typical conversational speech
	Occasionally:
	talk for longer periods raise voice above conversational level participate in group discussions, give a speech or other presentation
_____	Extensive:
	Speech usage consistently goes beyond everyday conversational speech.
	Regularly:
	talk for long periods talk in a loud voice participate in group discussions, give presentations or performances
	Although the demands 011 your speech are often high, you are able to continue with most work or social activities even if your speech is not perfect.
_____	Extraordinary:
	Very high speech demands
	Regularly:
	talk for long periods of time talk with loud or expressive speech or give presentations or performances.
	The success of your work or personal goals depends almost entirely on the quality of your speech and voice.

Data Collection

Data were collected from 200 participants with SD as part of a larger research project (Baylor, 2007). Community-dwelling adults age 18 years and older who had a diagnosis of SD confirmed by an otolaryngologist were included regardless of type of SD or treatment history. Participants were asked to complete a battery of questionnaires and were offered their choice of completing the questionnaires on-line or on paper.

Data Analysis

The data were analyzed using crosstabs to explore the associations between self-reported speech usage and the following demographic or SD variables: age, gender, education level, employment status, presence of other medical conditions affecting participation, length of time since SD diagnosis, treatment status, and a self-rating of voice quality. Two measures of the psychosocial impact of communication disorders were included: the total score from the Voice Handicap Index (VHI) (Jacobson et al., 1997—higher scores indicate greater handicap), and a measure of communicative participation (Baylor, Yorkston, Eadie, & Amtmann, 2007; Yorkston et al., 2008—higher scores indicate better participation). Statistical tests varied as a function of the type of measure, and SPSS version 12.0 (SPSS, 2003) was used for all analyses.

RESULTS

Summary demographic data are presented in Table 3. Several characteristics of the sample including the predominance of women and the mean age are typical of the SD population (Duffy & Yorkston, 2003; Sulica, 2004). The sample was predominately Caucasian (95%).

TABLE 3.

Demographic and SD characteristics for the SD sample (n = 200).

Demographic Variables	Response Categories	Results
Age	Mean	55.3 years (sd 10.8)
	Range	27–83 years
Gender	Female	78.5%
	Male	22.5%
Education level	High school	9%
	Some college	30%
	Bachelors degree	30.5%
	Graduate degree	30.5%
Work status	Full time paid work	48.0%
	Part time paid work	14.5%
	No paid work	36.0%
	No response	1.5%
Other medical conditions that influence participation	Yes	31.5%
	No	68.5%
Total VHI Score	Mean	83.2 (sd 18.4)
	Range	31–116
Communicative Participation score	Mean	−1.0 logits (sd 2.1)1
	Range	−10.26–4.95 logits
Number of years since diagnosis of SD	Mean	9.4 years (sd 6.9)
	Range	<1–32 years
Treatment status	Currently receiving botox	59.0%
	Currently not receiving botox	40.5%
	No response	0.5%
Self-rating of voice quality	Voice no different than peers without SD	1.0%
	Voice slightly different than peers without SD	13.0%
	Voice somewhat different than peers without SD	22.5%
	Voice very different than peers without SD	62.5%
	No response	1.0%

¹Logits (logs odd units) is a unit of measurement associated with Item Response Theory (Bond & Fox, 2001). Logits are calculated through a logarithmic transformation of raw scores. This transformation creates an interval scale of measurement on which each unit (logit) represents an increase (or decrease) in the odds of a participant selecting a particular response category. In the IRT model, the mean is 0, and scores extend above or below zero.

The distribution of participants across the five speech usage categories represented a normal distribution with the following percentages in each category: Undemanding (15.5%), Intermittent (22.0%), Routine (33.5%), Extensive (23.0%), and Extraordinary (6.0%). The relationships between speech usage and demographic or SD characteristics were explored through measures of associations. Table 4 contains a list of the variables tested for their associations with speech usage levels, the statistical analyses conducted, and the results. Age, education, and work status were the only variables significantly associated with speech usage levels. Figures 1 and 2 present graphs of the relationships between speech usage and age and work status, respectively. Speech usage appears to decline with age. The less demanding speech us age categories had proportionally more people in the “no paid work” category whereas the extensive and extraordinary speech usage categories were dominated by people who worked full time. With regard to education, people with high school education levels were concentrated more in the undemanding speech usage category compared to the other categories. The extraordinary usage category was composed largely of people with bachelors or graduate college degrees.

TABLE 4.

Results of tests of association between speech usage and participant variables.

	Variable	Tests of Association	Results
Demographic Descriptors	Age	Spearman correlation	−0.246*
	Gender	Pearson chi square	1.946
	Education	Kendall-Tau	0.174*
	Work status	Pearson chi square	41.624*
	Other medical	Pearson chi square	5.602
Psychosocial Measures	VHI1	Kendall-Tau	−0.055
	Communicative Participation2	Spearman correlation	0.059
SD Characteristics	Years since diagnosis	Spearman correlation	−0.101
	Receiving botox yes/no	Pearson chi square	5.958
	Self-rating of voice	Kendall-Tau	0.053

^*Significant at p < 0.01

¹Voice Handicap Index (Jacobson et al., 1997)

²Communicative Participation Item Bank (Baylor, 2007; Yorkston et al., in press)

Figure 1.

Mean age (±1 standard deviation) for each speech usage category.

Figure 2.

Distribution of work status within each speech usage category.

An informal examination was conducted of the occupations reported by participants in each of the speech usage categories. Selected examples of job titles in each of the categories are presented in Table 5. Each speech usage category contained participants whose job titles might be expected in that category based on existing voice demand scales in the literature, but there were also many exceptions. Many job titles were represented in multiple speech usage categories.

TABLE 5.

Selected examples of self-reported occupations in each of the speech usage categories. The “expected” occupations are jobs that might be expected in each category based on existing voice demand scales. The “unexpected” occupations are job titles that were not consistent with occupation-based voice demand rating scales.

Speech Usage Category	Expected Occupations	Unexpected Occupations
Undemanding	Lab technician, computer entry	Counselor, supervisor
Intermittent	Engineer, domestic support	Marketing, educator
Routine	Cashier/clerk, financial advisor	Medical billing, lab technician
Extensive	Sales, Counselor, educator	Engineer, emergency medical technician
Extraordinary	Music educator	Tailor, technical writer

DISCUSSION

The Levels of Speech Usage is a scale intended to be used for systematically describing and coding the speech usage of participants in clinical or research settings. The scale provides an alternative to existing voice demand rating scales in the literature in that it is applicable to people across a range of communication disorders and life situations. Clinicians and researchers may want to use the scale to summarize speech usage as part of describing demographic characteristics of clients and participants. Clinicians may find the scale helpful as a starting point for talking with clients about their speech needs and priorities and how these might be addressed in intervention. Clinical researchers may want to use the scale to code speech usage as a variable for statistical analyses to explore the relationships among speech usage and other variables of interest, including the psychosocial impact of communication disorders and intervention.

The results of this study suggest that speech usage may be related to work status (full time, part time, or no paid work), age, and education. These relationships are expected in that people who are older are less likely to be working full time and less likely to be in the more demanding speech usage categories. The self-reported job titles of participants did not necessarily adhere to stereotypical expectations of speech demands. For example, some occupations such as teacher, which is typically rated as a high voice-demand occupation, were represented in multiple speech usage categories. The difference between working full time versus part time or not at all appeared to be a stronger trend across the speech usage categories than the specific occupations. Further investigation of this phenomenon is needed. Speech usage was not significantly associated with psychosocial measures such as the VHI or communicative participation. This may suggest that people at all different levels of speech usage experience important psychosocial consequences of speech disorders, and that psychosocial outcomes are likely dependent on a complex constellation of variables beyond the extent of speech usage alone.

Research is needed to explore the properties of the Levels of Speech Usage scale in other populations and to study the relationships between speech usage and other demographic characteristics.