Stuttering attitudes of students: Professional, intracultural, and international comparisons.

Journal of Fluency Disorders 39 (2014) 34–50

Contents lists available at ScienceDirect

Journal of Fluency Disorders

Stuttering attitudes of students: Professional, intracultural, and international comparisons Kenneth O. St. Louis a,∗ , Aneta M. Przepiorka b , Ann Beste-Guldborg c , Mandy J. Williams d , Agata Blachnio b , Jacqueline Guendouzi e , Isabella K. Reichel f , Mercedes B. Ware g a

West Virginia University, Morgantown, WV, USA The John Paul II Catholic University of Lublin, Lublin, Poland c Minot State University, Minot, ND, USA d University of South Dakota, Vermillion, SD, USA e Southeastern Louisiana University, Hammond, LA, USA f Touro College, New York, NY, USA g Ohio University, Athens, OH, USA b

a r t i c l e

i n f o

Article history: Received 2 March 2013 Received in revised form 29 August 2013 Accepted 2 October 2013 Available online 24 October 2013

Keywords: Stuttering Public attitudes Students Poland Native American

a b s t r a c t Purpose: The study sought to identify major-specific, training, and cultural factors affecting attitudes toward stuttering of speech-language pathology (SLP) students. Method: Eight convenience samples of 50 students each from universities in the USA and Poland filled out the Public Opinion Survey of Human Attributes-Stuttering (POSHA–S) in English or Polish, respectively. USA samples included undergraduate and graduate students in SLP majors or non-SLP majors as well as a sample of non-SLP students who were Native Americans. Polish samples included SLP (logopedics), psychology, and mixed majors. Results: SLP students held more positive attitudes than non-SLP students in both countries. Graduate students held more positive attitudes than undergraduate students in the USA, and this effect was stronger for SLP than for non-SLP students. Native American students’ stuttering attitudes were similar to other American non-SLP students’ attitudes. Polish student attitudes were less positive overall than those of their American student counterparts. Conclusion: SLP students’ attitudes toward stuttering are affected by a “halo effect” of being in that major, by specific training in fluency disorders, and by various cultural factors, yet to be clearly understood. Educational objectives: The reader will be able to: (a) describe major factors affecting SLP students’ attitudes toward stuttering; (b) describe similarities and differences in attitudes toward stuttering of students from the USA and Poland; (c) describe similarities and differences in attitudes toward stuttering of Native American students from the USA and non-Native American students. © 2013 Elsevier Inc. All rights reserved.

∗ Corresponding author at: Department of Communication Sciences & Disorders, 805 Allen Hall, PO Box 6122, West Virginia University, Morgantown, WV 26506-6122, USA. Tel.: +1 304 293 2946. E-mail address: [email protected] (K.O. St. Louis). 0094-730X/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jfludis.2013.10.001

K.O. St. Louis et al. / Journal of Fluency Disorders 39 (2014) 34–50

35

1. Introduction 1.1. Stuttering stereotype among SLPs A number of reports have appeared over the years clearly documenting that, in addition to the general public, many speech-language pathologists (SLPs) hold the so-called “stuttering stereotype,” i.e., that people who stutter are frustrated, anxious, shy, nervous, weak, involved or affected psychologically, and so on. Much of this work emanated from Dean Williams’ and his students’ research with semantic differential scales (Woods & Williams, 1971, 1976; Yairi & Williams, 1970). Other studies of attitudes using open-ended descriptions of hypothetical “stutterers” showed that both practicing SLPs and SLP students generated vastly more negative personality descriptions of those who stutter than positive or neutral descriptions (e.g., Lass, Ruscello, Pannbacker, Schmitt, & Everly-Myers, 1989; Ruscello, Lass, French, & Channel, 1989–1990). One carefully done mixed-method study that explored attitudes of university students who were not SLP majors concluded that many respondents were aware that it was the stuttering—not some underlying personality difference—that might be responsible for the “stuttering stereotype” (Hughes, Gabel, Irani, & Schlagheck, 2010). 1.2. Training, preferences, and bias among SLPs Other commonly reported findings are that SLPs or SLP students lack training in, are uncertain about, prefer not to treat, and/or are biased toward individuals who stutter (e.g., Brisk, Healey, & Hux, 1997; Cooper, 1975; Ragsdale & Ashby, 1982; St. Louis & Durrenberger, 1993; Tellis, Bressler, & Emerick, 2008; Yaruss & Quesal, 2002). These studies used varied survey instruments. One example is Cooper’s (1975) Clinician Attitudes Toward Stuttering (CATS) scale developed in the 1970s to explore SLPs’ beliefs and approaches to understanding and treating stuttering. Using the CATS with SLP students across the USA, St. Louis and Lass (1981) found that many of the students reported similar deficiencies in training and biases in attitudes reported earlier by experienced SLP clinicians (Cooper, 1975). Moreover, students’ attitudes in this cross-sectional study became more pessimistic as they progressed through their undergraduate and subsequent graduate programs (St. Louis & Lass, 1981). In contrast to these results, a few studies demonstrated some improvement in practicing SLPs’ attitudes over time (e.g., Cooper & Cooper, 1985, 1996). Similarly, using a semantic differential scale that was quite different from the ones used in early research, a recent study showed that American SLPs with considerable experience actually reported more positive reactions to a hypothetical adult who stuttered verses an adult who did not (Swartz, Gabel, & Irani, 2009). Other investigations have sought to determine the extent to which SLPs lack appropriate training or positive regard for people who stutter in different cultures such as the UK (Cooper & Rustin, 1985; Crichton-Smith, Wright, & Stackhouse, 2003) and Turkey (Mavis¸, St. Louis, Özdemir, & To˘gram, 2013). These studies, using the CATS, have documented many of the same aforementioned problems observed in the USA, e.g., feeling less comfortable working with people who stutter than with clients who have articulation disorders (Mavis¸ et al., 2013). Nevertheless, the results have been far from uniform internationally, likely due in part to cultural differences and different experiences in stuttering treatment. 1.3. Issues in attitude measurement An obvious weakness in all of this research is that different measures have been utilized to measure attitudes. The openended descriptions and semantic differential scales focus primarily on perceived personality differences between people who stutter or not, but little on what they actually can or cannot do. Moreover, scales that identify specific diagnostic and therapy strategies are based on specific coursework and/or clinical experience. The CATS has been modified to fit either new or emerging clinical strategies, as have some semantic differential scales (e.g. Crichton-Smith et al., 2003; Mavis¸ et al., 2013), further compromising comparisons among investigations. This lack of uniformity in measures of attitudes constitutes a threat to the validity of estimates of stuttering attitudes. Stemming from an attempt to provide a standard, interculturally appropriate measure of public attitudes toward stuttering anywhere in the world, St. Louis and his colleagues have developed the Public Opinion Survey of Human Attributes–Stuttering (POSHA–S) (St. Louis, 2005, 2011a, 2011b). The POSHA–S (described below) has been shown to provide valid and reliable estimates of attitudes both in its English version and in numerous translations to other languages such as Turkish, Arabic, Chinese, Polish, Russian, Bulgarian, Brazilian Portuguese, and French (Abdalla & St. Louis, 2012; Ip, St. Louis, Myers, & An Xue, 2012; Przepiorka, Blachnio, St. Louis, & Wozniak, 2013; St. Louis, 2012c; St. Louis & Roberts, 2010; St. Louis, Andrade, Georgieva, & Troudt, 2005; St. Louis et al., 2011). 1.4. Potential factors affecting SLP attitudes Using the POSHA–S, the principle objective of the current study was to shed more light on factors that affect the attitudes of SLPs toward stuttering. College students were the targeted population because we sought to measure attitudes before the vast majority of respondents had long-term clinical experiences that might affect their attitudes. Our intention was to isolate and measure three overlapping factors that could influence attitudes of SLP students toward stuttering. The first was a conscious or unconscious predisposition of SLP students to regard stuttering in a more positive light than those with other majors. For lack of a better term, we refer to it here as a “halo effect.” The second factor was specific information and/or

36


clinical experience with stuttering certain to be a part of most SLP courses of study. The third factor we chose to include was the influence of culture (Tellis & Tellis, 2003). Obviously, an indefinite number of cultural differences exist that could affect stuttering attitudes, but this study chose to focus on two of them: (a) American English-speaking students with diverse backgrounds versus American English-speaking students who regarded themselves as Native American “Plains Indians” and (b) American English-speaking students versus Polish-speaking Polish students. In spite of considerable early speculation on whether or not stuttering occurred among Native American tribes (e.g., Johnson, 1944; Lemert, 1953; Snidecor, 1947; Stewart, 1960), to our knowledge, systematic study of public attitudes of Native Americans toward stuttering has not been reported except for two recent papers which suggested that Native American attitudes were better than average (Beste-Guldborg & Campanale, 2012; Beste-Guldborg, St. Louis, & Campanale, 2012). These intriguing preliminary findings suggested that cultural differences, related to tribal traditions, or other influences might have played a role in their stuttering attitudes. We selected student populations from Poland as an international comparison because the field of logopedics (the term for speech-language pathology) was a thriving field there with more than 5000 trained logopedists from more than ten training programs around the country (Georgieva, 2010). Poland is a “western” country but with a relatively recent history of being part of the former Soviet Union (Przepiorka et al., 2013). 1.5. Purposes The specific research questions on attitudes toward stuttering addressed were: (a) To what extent are attitudes different between American undergraduate versus graduate SLP students? (b) To what extent are attitudes different between American undergraduate and graduate students in SLP versus other majors? (c) To what extent are attitudes different between Polish undergraduate students in SLP versus non-SLP majors? (d) To what extent are attitudes different between American and Polish SLP versus non-SLP students? (e) To what extent are attitudes different between American students who are Native Americans versus those who are not Native American and who represent a wide range of backgrounds? 2. Method 2.1. Measure of attitudes toward stuttering The Public Opinion Survey of Human Attributes-Stuttering (POSHA–S) was used as the measure of attitudes toward stuttering. As noted, the POSHA–S was developed as a measure of public attitudes that could be translated and used anywhere in the world. At the time of this investigation (circa January 2013), it had been translated to 15 languages and used in 23 countries (St. Louis, 2011a). Sociometric properties such as item analysis and selection, reliability, construct and concurrent validity, internal consistency, user-friendliness, translatability, survey modality (paper versus online), sampling methods (convenience versus probability), and other properties have been reported (St. Louis, 2005, 2012b, 2012c; St. Louis et al., 2011; St. Louis, Lubker, Yaruss, Adkins, & Pill, 2008; St. Louis, Lubker, Yaruss, & Aliveto, 2009; St. Louis, Reichel, Yaruss, & Lubker, 2009; St. Louis & Roberts, 2010). The value of comparisons of individual samples with comparison with a growing database of POSHA–S results has also been documented (St. Louis, 2011b, 2012c), as has the value of a standard protocol for summarizing results in tables and graphs (Ip et al., 2012; Özdemir, St. Louis, & Topbas¸, 2011a, 2011b; St. Louis, 2011b, 2012b). The POSHA–S has a demographic section, a general section comparing stuttering to four other human attributes (intelligent, left handed, obese, and mentally ill), and a detailed section on stuttering. Mean attitude and some scaled demographic ratings for samples are converted to a scale from −100 to +100. Moreover, ratings for various attitude items are inverted so that higher scores uniformly represent more positive or more informed attitudes and lower scores, less positive or less informed attitudes. The Polish version of the POSHA–S was developed according to recommended procedures for translations (Przepiorka et al., 2013; St. Louis & Roberts, 2010). The second author and another Polish SLP academic translated the English version to Polish, and a professional translator who knew English very well back-translated the Polish version to English. The two English versions were checked for the same meanings, although word-for-word similarity was not required. Additionally, readability of the Polish version was verified by 10 students. Accordingly, the Polish version was deemed to be an accurate translation of the original. 2.2. Student respondents Eight groups of students were sampled for this investigation, five in the USA and three in Poland. Undergraduate and graduate students studying SLP (USA SLP UG and USA SLP GR) attended one of four universities in the USA (West Virginia University [WVU], the University of South Dakota [USD], Southeastern Louisiana University [SELU], and Touro College in New York City [TC]). Comparison USA undergraduate and graduate students majoring in other fields attended one of these four universities or the University of Minnesota-Duluth (UM-D) (USA Non-SLP UG and USA Non-SLP GR). Another group of


37

students who were self-identified Native Americans or more specifically, Plains Indians representing more than 12 tribes (USA Non-SLP NAT), were included to provide an intracultural comparison with the other American students. This component was carried out at Minot State University (MSU). Whereas these respondents did not identify the universities they attended, they reported residing in Montana, Minnesota, North Dakota, Wisconsin, and Washington. Most—but not all—of them most likely attended Tribal Colleges to which links to the POSHA–S were sent. In Poland, the three groups of students attended different universities. The SLP students were mostly 3rd-year students majoring in logopedics (POL SLP UG) at Maria CurieSklodowska University in Lublin, the psychology students were similarly mainly 3rd-year majoring in psychology (POL PSY UG) at The John Paul II Catholic University of Lublin, and the control group had a variety of other majors that were unspecified (POL CON UG) recruited from the University of Wroclaw, the University of Warsaw, Adam Mickiewicz University in Poznan, the University of Rzeszow, and the Jan Kochanowski University in Kielce. 2.3. Recruitment and questionnaire administration Recruitment of undergraduate and graduate students at three universities (WVU, USD, and SELU) followed a similar online procedure. Coauthors invited students by email or contacted faculty colleagues at their universities who in turn invited their students to follow a link to an online version of the POSHA–S at WVU. Students were also encouraged to recruit other students to participate. Up to two reminders were sent to students who were contacted directly. Respondents listed their majors and their class ranks (i.e., freshman, sophomore, junior, senior, or graduate student). Students at TC were SLP graduate (master’s) students in a class on fluency disorders who filled out the paper version of the POSHA–S on the first day of class. Combined SLP students who filled out the POSHA–S included 15 freshmen, 16 sophomores, 16 juniors, 32 seniors, and 76 graduate students. Non-SLP respondents included 6 freshmen, 15 sophomores, 9 juniors, 20 seniors, and 26 graduate students. The Native American sample was recruited through email lists obtained by a coauthor that contained tribes designated as Plains Indians in the upper Midwest/Western regions of the USA. These respondents followed a link to an online version of the POSHA–S at MSU. From invitations to 1619 individuals representing 42 tribes, a total of 577 Native Americans responded, of which 161 identified themselves as students. As noted above, they did not report their major, but it is unlikely that more than a very few would be SLP majors. The Polish samples were recruited by a research assistant. A total of 50 SLP (logopedics) students, 84 psychology students, and 50 students with other majors filled out questionnaires. Return rates for the respondents varied as a function of the samples. At WVU, 200 undergraduate SLP students and 22 graduate students were invited to participate, and 48 and 13, respectively, filled out the online POSHA–S, resulting in return rates of 24% and 59%. Also, 112 Non-SLP undergraduates were invited and 17 responded, for a 15% return rate. USD online return rates were as follows, based on the best estimates of the number of links to the online survey sent to students: SLP undergraduates—28% (18/65), SLP graduates—55% (30/55), Non-SLP undergraduates—13% (33/250), and NonSLP graduates—26% (26/100). At SELU, estimates for online return rates for undergraduate SLP students was 13% (13/100) and for graduate students SLP was 30% (12/40). The TC class administration yielded a 100% return rate. Return rates for the Polish samples were 93% (50/54) for logopedics students, 98% (84/86) for psychology students, and 100% (58/58) for students with other majors. The return rate for the Native Americans was 36%. Each sample was reduced at random to 50 respondents, or in the case of the Non-SLP GR sample, increased. Since only 26 non-SLP graduate students filled out the POSHA–S online (all from USD and SELU), an additional 24 graduate students were added from the POSHA–S database from two studies, 20 from an investigation of students and nonstudents carried out at UM-D and four from a study comparing online versus paper POSHA–S surveys (St. Louis, 2011a, 2012b). 2.4. POSHA–S scoring and data conversion To fully explain the rating and scoring system of the POSHA–S, it is important to emphasize the following about its rationale and development. The instrument was designed to be usable by a wide variety of individuals outside the scientific community, including stakeholders in the self-help movement, teachers, clinical SLPs, or anyone interested in comparing or changing attitudes toward stuttering. Experimental versions used a quasi-continuous scale (St. Louis et al., 2008) and, later, a 1–9 scale (St. Louis, Lubker, et al., 2009). The quasi-continuous scale was rejected because it was neither user- nor experimenter-friendly, and the 1–9 scale was replaced because it required too much respondent time. The final version of the POSHA–S adopted 1–5 scales for the general section and a “yes”/“no”/“not sure” response mode for the stuttering section. These shortened response time for the POSHA–S to about 10 min, which we considered the maximum allowable time for widespread public use. With “no” = 1, “not sure” = 2, and “yes” = 3 conversions, the data become a 1/2/3 ratio scale, even though the data could be considered categorical or nominal. As with the earlier, experimental data, the results are further converted to a −100 to +100 scale for each sample to provide more manageable comparisons among and within samples. To verify that the possible nominal to ratio scale conversions were justifiable, St. Louis (2012c) compared the final version of the POSHA–S to the second prototype, i.e., 1–5 (general section) and 1–3 (stuttering section) to 1–9 (both sections) responses from the same respondents who filled out either version two weeks apart, with order counterbalanced. Both paper-and-pencil and online POSHA–Ss were used. Either response version of the POSHA–S yielded very similar results on the converted −100 to +100 scale, with none of the 60 comparisons being significantly different (see below).

38


In numerous POSHA–S studies, the data have been analyzed by t-tests for all pair-wise comparisons using the Bonferroni correction of p ≤ .00417 (.05/12). Multiple reasons support using t-tests. First, it is widely assumed that t-tests are highly robust when some underlying assumptions are not obtained (e.g., lack of homoscedasticity or skewness) (e.g., Fagerland, 2012; Stonehouse & Forrester, 1998). Second, t-tests are arguably the most well-known and best understood inferential statistic that could render POSHA–S results understandable and replicable by collaborators or potential users world-wide. Third, the t-test is easy to calculate, and almost all statistical programs permit unequal frequencies, which is typical in samples using the POSHA–S. Fourth, the t-test can be run on small samples. Numerous samples in the POSHA–S database are smaller than 50 respondents (St. Louis, 2011a), and St. Louis (2008) determined that samples as small as 25 predicted the mean of a “population” of 1500 respondents as well as samples of 50, 100, or 200. Using the t-test with the substantial Bonferroni correction factor, i.e., dividing the alpha level by 12, which was the average number of items under the initial prompts (St. Louis et al., 2008), requires a substantial difference in ratings to report a significant difference. The alpha level of p ≤ .00417 provides an excellent balance between not making Type I errors (i.e., reporting significant differences when one does not exist), but also, and equally importantly, not making Type II errors (i.e., not reporting differences when they do exist). That alpha level has also permitted useful comparisons among POSHA–S samples in a growing number of published investigations (e.g., Flynn & St. Louis, 2011; Özdemir et al., 2011a, 2011b; Przepiorka et al., 2013; St. Louis, 2011b; St. Louis, 2012a, 2012b, 2012c; St. Louis, Lubker, et al., 2009; St. Louis, Reichel, et al., 2009). It must be remembered that any consistent difference, no matter how small, will be statistically significant if the sample size is increased sufficiently. Therefore, it is important to temper consideration of statistical power with the magnitude of effect sizes. In recent studies, t-tests have been followed by Cohen’s d effect sizes (Cohen, 1988) to determine whether or not it is the sample size, effect size, or both that should be considered in interpretation of the results.

3. Results 3.1. Respondents This investigation compared attitudes toward stuttering of students who chose majors in speech-language pathology (SLP or logopedics) with students who selected other majors. It was assumed that the majority of undergraduate students in SLP in the USA only had minimal training in fluency disorders, most likely in an introduction to communication disorders class, although some of them had coursework ranging from a few weeks to one semester. Both SLP and non-SLP students were taken from diverse student populations from several universities in urban and rural regions of the USA. Since most professional training occurs at the graduate level, the study further compared undergraduate with graduate students in the SLP versus non-SLP majors. The design permitted comparison of these American students with Polish students primarily from one university. The Polish non-SLP samples were further divided into psychology students and other majors. Additionally, the study compared a sample of Native Americans who indicated that they were college students with the American and Polish students. Table 1 provides demographic comparisons of the eight student samples. It contains descriptive information as well as ratings of a number of self-identification items. In the last column, the table lists the percentages of 28 possible pair-wise contrasts that were statistically significant using t-tests for independent samples and the Bonferroni Correction of p ≤ .05/12 or .00417. Most of the demographic data is self-explanatory, but several noteworthy similarities and differences appeared among the groups. Combined return rates across the various recruitment procedures ranged from 14% to 100%, with predictably lower return rates from email invitations to fill out an online questionnaire compared to filling out a paper version handed to each potential respondent. Graduate students were older than undergraduate students in the USA, except the Native Americans (mostly undergraduate) were closer in age to the graduate students. All eight samples had more females than males, with the SLP groups consisting of nearly all women (92–98%) and the other groups having from 58% to 84% females. Based on a careful study showing virtually no effects of male versus female respondents on the POSHA–S, the differences in male and female respondents in these student samples are likely miniscule and unimportant (St. Louis, 2012a). Very few of the Polish students were married or had children. The percentages increased for being married or parents for graduate students and even more to about 50% for the Native Americans. Relative incomes could range from −100 to +100, with 0 being neutral. A zero indicates a rating of the same as others among one’s friends and family and the same as all those in one’s country. Relative incomes were close to average for the American students, except the Native Americans, who indicated very low income levels (−49 on the −100 to +100 scale). The vast majority of all the respondents in both countries identified themselves as Caucasian except the Native Americans who all listed their race as American Indian or an affiliation with one or more tribal groups. The majority of the groups, except the Native Americans who were not asked about their religion after tribal elders thought it would be too sensitive, regarded themselves as Christians. Thirty-percent of the SLP graduate students were Jewish, most from TC in New York City. Self-identification as stuttering or mentally ill was 0–2%, except for the Polish mixed majors, whose percentages were 20% and 16%, respectively. This difference, addressed by Przepiorka et al. (2013) most likely reflects uncertainty by the Polish majority of the exact limits of these two categories. Interestingly, only about half the percentage of Poles regarded themselves as intelligent compared to the Americans. Similarly, they rated


39

Table 1 Demographic comparisons for eight samples of students, five from the USA and three from Poland. The last column shows the percentage of pair-wise t-tests among the eight groups that were statistically significant. USA SLP GR

USA Non-SLP UG

USA Non-SLP GR

USA Non-SLP NAT

POL SLP UG

POL PSY UG

POL CON UG

% t-tests significanta

Number 50 50 Return rate 22% 55% Age: mean (yr) 21.5 25.7 Total schooling: mean (yr) 14.5 16.2 Sex: males/females (% total) 6%/94% 8%/92% Married (% of total) 4% 38% Parent (% of total) 8% 22% Relative income (−100 to +100) 11 −6 Race (% total) Caucasian/White 92% 98% Black/Hispanic/Asian 2% 0% Native American 0% 2% Other 0% 0% No response 6% 0% Religion (% total) Christian 88% 66% 0% 30% Jewish 2% 2% Other/no response No response 10% 2% E: 100% English (E)/Polish (P) as native E: 98% language (% responding) Know >1 language (% 12% 34% responding) Self-identification (% total) Stuttering 2% 2% Mentally ill 2% 2% 6% 6% Obese 6% 18% Left handed 52% 50% Intelligent Self-rating of health and abilities (−100 to +100) 57 62 Physical health Mental health 66 65 Ability to learn 69 65 70 60 Speaking ability Self-rating of life priorities (−100 to +100) Be safe/secure 83 86 Be free 73 61 Spend time alone 44 49 Attend social events 19 21 Imagine new things 35 19 65 60 Help less fortunate −26 −36 Have exciting experiences 31 55 Practice my religion 64 67 Earn money 87 82 Do job/duty 88 86 Get things done 71 67 Solve big problems 12.0 12.6 Completion time: mean (min)

50 14% 22.3 14.0 34%/66% 8% 10% 5

50 26%b 28.5 16.5 34%/66% 32% 18% 10

50 36% 26.0 14.1 34%/66% 50% 48% −49

50 93% 20.4 14.1 2%/98% 0% 0% −22

50 98% 20.8 14.2 16%/84% 0% 0% −16

50 100% 20.0 15.4 42%/58% 8% 6% −14

– – 50 50 – – – 39

80% 10% 4% 6% 0%

94% 4% 0% 0% 2%

0% 0% 100% 0% 0%

100% 0% 0% 0% 0%

100% 0% 0% 0% 0%

100% 0% 0% 0% 0%

– – – – –

78% 2% 8% 12% E: 100%

74% 0% 20% 6% E: 98%

– – – – E: 88%

100% 0% 0% 0% P: 100%

100% 0% 0% 0% P: 100%

100% 0% 0% 0% P: 100%

– – – – –

32%

18%

13%

84%

72%

64%

–

0% 2% 16% 10% 56%

0% 0% 8% 14% 52%

2% 2% 10% 10% 60%

0% 0% 0% 2% 24%

2% 2% 2% 0% 26%

20% 16% 24% 16% 22%

– – – – –

47 60 76 60

51 62 79 66

37 50 70 71

41 60 46 77

45 44 44 48

28 34 44 57

14 11 54 11

69 59 45 1 43 54 −10 37 61 84 78 78 12.9

68 72 55 10 40 62 −23 2 57 77 78 78 8.6

80 64 14 18 28 45 −36 −11 62 74 82 62 7.0

91 63 24 36 26 54 −13 43 54 64 71 80 11.1

71 64 21 13 34 31 −11 30 42 57 66 78 9.7

72 71 12 28 43 41 2 −14 52 61 62 66 8.7

14 0 46 11 0 14 11 39 7 32 14 0 46

USA SLP UG

a Number of independent t-tests for all possible pair-wise comparisons between samples that were statistically significant using the Bonferroni correction (p ≤ .00417) for relevant items. b Not including the added graduate students from two other POSHA–S samples.

themselves much lower than the Americans on their ability to learn and slightly to moderately lower for physical health and mental health (except the logopedics students). Twelve items on ratings for life priorities were added to the final POSHA–S such that, after many more additional respondents are available in the database, their potential to predict positive or negative stuttering attitudes can be evaluated. Among the samples in Table 1, however, it can be seen in the percentages of significant pair-wise contrasts between groups that ratings for spending time alone differentiated Polish and Native American from the other four groups of American students. The same profile, but to a lesser extent, characterized ratings for doing one’s job or duty. Practicing one’s religion differentiated the American non-SLP graduate students, Native Americans, and Polish mixed majors from the other groups.

40


3.2. Summary of stuttering attitudes Table 2 provides a summary of the means and standard deviations for each of the POSHA–S items scores for attitudes of the eight samples of students. In the second to last column, the table shows the median sample mean value from 162 different sample comparisons using the POSHA–S taken from 6936 respondents from 22 countries and representing 14 different languages. As in Table 1, the last column shows percentages of pair-wise contrasts that were statistically significant using t-tests for independent samples and the Bonferroni Correction of p ≤ .00417. It can be seen that the number of significant contrasts ranged from 0% (0) to 64% (18) of the 28 possible. The six POSHA–S ratings yielding the greatest differences between the eight samples (54–64%) were for: being accommodating/helping (component), Overall Stuttering Score, source of learning from school (item), amount known about stuttering (item), stuttering being caused by a frightening emotional experience (item), and telling a person stuttering to “slow down” or “relax” (item). Eight ratings with the least group differences (all 0%) were for: traits/personality (component), people who stutter are nervous or excitable (item), stuttering is caused by ghosts/demons/spirits (item), people who stutter can make friends (item) or lead normal lives (item), pitying a person stuttering (item), and wanting to stutter (item). The Appendix (supplementary online material) includes all the results of the statistical tests for each of the eight respondent samples across all of the 60 POSHA–S ratings (45 items, 11 components [means of clusters of items], 3 subscores [means of clusters of components], and the Overall Stutterer Score [mean of two stuttering subscores]). Values are included for 365 t-tests of independent variables significant at the Bonferroni corrected alpha level of p ≤ .00417 as well as Cohen’s d effect sizes (Cohen, 1988) for each contrast. Virtually all of the Cohen’s d values reflected “large” or “very large” effect sizes, reflecting the conservative alpha level set for significance. The percentage of pair-wise ratings that differed significantly among the eight samples ranged from 2% between non-SLP Polish psychology versus control students to 37% between these Polish psychology and American SLP undergraduates. To facilitate comparisons among the eight groups, Table 3 shows these percentages in a matrix. The 2–37% range of differences between groups can best be interpreted by comparing them to other POSHA–S studies summarized in St. Louis’s (2012c) report. In that article, comparisons between numerous samples using the POSHA–S ranged from virtually no change as expected, e.g., in test–retest reliability samples, to large differences as expected, e.g., between Canadian versus Cameroonian respondents or before and after a talk by a person who stutters. In the St. Louis (2012c) summary, large predictable pair-wise differences were reflected by percentage of corrected significant differences exceeding 30% of the items. Moderate changes reflected percentages between 10–30%, and small changes between 5 and 10%. Percentages between 0 and 5% of POSHA–S ratings basically reflected no differences. 3.3. Undergraduate versus graduate SLP students in the USA The first purpose of the study was to determine the extent to which American SLP undergraduate majors differed from their graduate major counterparts. SLP graduate students’ Overall Stuttering Scores (OSS) were10 points higher or more positive than SLP undergraduate students’ scores (33 versus 43). Parallel changes of similar magnitudes characterized both Beliefs about people who stutter subscore (54 versus 64) and Self Reactions to people who stutter subscore (11–21), but not for the Obesity/Mental Illness subscore (−34 versus −38). These differences can be compared in the first versus second tracngs at the left in Fig. 1. Significant differences between these two SLP samples were not as great as might have been expected, with seven (12%) of the contrasts reaching significance (Column 1 of the Appendix list the t-test values and Cohen’s d effect sizes for specific POSHA–S variables). Graduate students had significantly more positive attitudes for OSS, cause (component), sources of information from print (item), amount of information about stuttering (item), and identifying themselves as people to help a person who stutters (item). On the other hand, they had less positive ratings for being concerned or worried if a sibling stuttered and impression of a person with mental illness. 3.4. Undergraduate versus graduate non-SLP students in the USA The second purpose of the study addressed differences in attitudes between students in SLP versus other majors in the USA. The third versus fourth tracings from the left in Table 1 indicate that smaller positive increments characterized American undergraduate versus graduate students who were not SLP majors, i.e., 24 versus 31 for OSS, 48 versus 54 for Beliefs, and 1 versus 8 for Self Reactions. The opposite was true for Obesity/Mental Illness (−26 versus −34). In spite of these consistent mean differences, only three (5%) POSHA–S ratings were statistically significantly different between the two samples using the conservative Bonferroni correction (Appendix, Column 19). Graduate students indicated they were less likely to tell a stuttering person to “slow down” or “relax” (item) and the related accommodating/helping component. They also had slightly to somewhat more accurate beliefs about the cause of stuttering (component). 3.5. SLP and non-SLP students in Poland The third purpose of the study focused on international comparisons of SLP and non-SLP students. Compared to the USA samples, logopedics (SLP) students in Poland (all undergraduates) were more positive than either the psychology students or the control students with other majors. Logopedics (SLP) students’ OSS, Belief, Self Reaction, and Obesity/Mental Illness

Table 2 Mean ratings for POSHA–S items, components, subscores, and overall stuttering scores (−100 to +100) followed by standard deviations in parentheses for the mean ratings of eight samples of university students and the median mean from the POSHA–S database. The last column shows the percentage of pair-wise t-tests among the eight groups that were statistically significant.

Overall stuttering score Beliefs about persons who stutter Traits/personality c

Have themselves to blame

Shy or fearfulc Stuttering should be helped by. . . Speech and language therapist Other people who stutter Medical doctorc Stuttering is caused by. . . Genetic inheritance Learning or habitsc A very frightening eventc c

An act of god

A virus or diseasec Ghosts, demons, spiritsc Potential Can make friends Can lead normal lives Can do any job they want Should have jobs requiring good judgment

USA SLP GR

USA Non-SLP UG

USA Non-SLP GR

USA Non-SLP NAT

POL SLP UG

POL PSY UG

POL CON UG

Database mediana

% t-tests significantb

33 (14) 54 (18) 34 (52) 98 (14) 0 (78) 4 (86) 47 (35) 98 (14) 16 (74) 27 (84) 49 (26) 2 (71) 26 (80) 39 (76) 50 (79) 78 (42) 98 (14) 88 (16) 96 (28) 100 (0) 90 (36) 66

43 (17) 64 (24) 48 (58) 88 (44) 34 (87) 21 (92) 63 (35) 100 (0) 40 (71) 50 (77) 67 (30) 37 (81) 57 (74) 63 (73) 53 (82) 90 (42) 100 (0) 81 (25) 100 (0) 96 (28) 57 (76) 69

24 (16) 48 (22) 45 (47) 88 (39) 24 (77) 22 (76) 30 (42) 84 (37) −6 (77) 12 (82) 36 (31) −6 (69) −18 (83) 20 (81) 66 (63) 60 (64) 96 (29) 80 (32) 96 (28) 96 (28) 62 (75) 64

31 (15) 54 (17) 50 (49) 88 (39) 33 (72) 30 (76) 33 (38) 94 (31) 8 (78) −4 (83) 59 (29) 35 (69) 29 (84) 50 (74) 84 (47) 61 (67) 98 (14) 76 (28) 96 (20) 98 (14) 58 (70) 50

31 (17) 53 (23) 39 (59) 84 (47) 10 (93) 24 (89) 45 (38) 94 (24) 16 (93) 24 (87) 42 (32) 26 (83) −16 (89) 24 (87) 86 (35) 46 (76) 86 (35) 85 (22) 94 (31) 98 (14) 88 (44) 62

30 (16) 44 (21) 35 (40) 98 (14) 40 (78) −32 (87) 49 (35) 100 (0) 55 (79) −8 (90) 32 (27) −10 (90) 44 (84) −92 (34) 94 (31) 56 (76) 100 (0) 59 (31) 100 (0) 100 (0) −8 (94) 43

14 (17) 30 (21) 29 (50) 78 (62) 36 (85) −28 (88) 9 (41) 100 (0) −8 (90) −66 (69) 26 (24) 24 (89) 8 (88) −74 (60) 66 (66) 32 (79) 98 (14) 58 (32) 98 (14) 100 (0) 16 (93) 18

19 (18) 38 (20) 30 (42) 72 (58) 28 (81) −11 (85) 17 (36) 83 (54) 24 (85) −55 (77) 40 (33) 59 (73) 25 (89) −39 (84) 70 (59) 45 (79) 80 (51) 64 (36) 91 (28) 87 (34) 35 (82) 41

16

61

34

43

18

0

81

7

7

0

−4

18

26

39

89

11

11

14

−15

46

34

43

14

11

17

18

3

57

63

11

51

14

89

0

61

32

93

0

84

0

50

32

40

14

(52)

(58)

(63)

(74)

(73)

(82)

(75)

(65)


Nervous or excitablec

USA SLP UG

41

42

Table 2 (Continued )

Self reactions to people who stutter Accommodating/helping Try to act like the person was talking normally Person like me c

Fill in the person’s words

Make joke about stutteringc Should try to hide their stutteringc Social distance/sympathy Feel comfortable or relaxed Feel pityc Feel impatient (not want to wait while the person stutters)c Concern about my doctorc Concern about my neighborc c

Concern about my brother or sister Concern about mec Impression of person who stutters Want to have stuttering Knowledge/experience Amount known about stuttering Stutterers known

Personal experience (me, my family, friends)

USA SLP GR

USA Non-SLP UG

USA Non-SLP GR

USA Non-SLP NAT

POL SLP UG

POL PSY UG

POL CON UG

Database mediana

% t-tests significantb

11 (21) 71 (23) 98

21 (22) 83 (19) 86

1 (22) 47 (26) 90

8 (24) 61 (25) 80

9 (24) 66 (26) 90

16 (20) 54 (26) 94

−1 (22) 40 (31) 82

0 (26) 25 (44) 80

−2

25

49

64

77

0

(14) 18 (87) 70 (58) 58

(45) 88 (44) 88 (44) 69

(37) −26 (75) 47 (77) 22

(57) −22 (90) 61 (67) 76

(36) 4 (95) 80 (53) 48

(31) 54 (73) 46 (81) −40

(56) −12 (87) 28 (90) −4

(55) −7 (75) 4 (93) −27

−22

39

51

29

29

54

(70) 100 (0) 82 (48) 28 (35) 40 (76) 0 (83) 71

(62) 94 (24) 74 (60) 10 (35) 22 (82) 8 (84) 60

(86) 86 (45) 65 (60) 26 (38) 28 (76) 10 (88) 55

(52) 98 (14) 74 (56) 26 (40) 26 (85) 18 (90) 49

(81) 88 (44) 84 (51) 13 (38) 46 (84) 26 (96) 86

(83) 92 (28) 78 (62) −4 (29) −42 (76) 6 (91) 80

(88) 78 (58) 65 (66) 1 (35) −18 (85) 10 (91) 38

(86) 67 (67) 33 (83) 0 (41) 9 (85) −4 (95) 42

88

7

69

11

4

32

17

29

22

0

58

7

(58) 84 (47) 90 (42) 37 (93) −18 (92) 21 (38) −75 (35) −40 (43) −7 (40) −81 (29) −32

(70) 55 (77) 96 (28) −20 (95) −66 (72) 17 (39) −81 (33) −27 (43) 32 (49) −79 (24) −33

(82) 64 (72) 98 (14) 64 (75) −22 (95) 9 (41) −70 (34) −32 (46) −33 (45) −75 (16) 12

(79) 69 (66) 92 (40) 50 (84) 8 (94) −3 (36) −78 (34) −28 (38) −40 (45) −81 (15) 38

(40) 30 (93) 48 (86) 14 (95) −44 (88) −9 (33) −81 (36) −53 (43) −51 (41) −88 (20) −21

(53) 48 (81) 86 (50) −60 (78) −88 (44) 15 (36) −83 (30) −25 (43) 32 (43) −90 (10) −18

(88) 36 (92) 100 (0) −18 (98) −72 (67) 1 (43) −70 (34) −32 (43) −21 (50) −76 (18) 0

(87) 37 (85) 72 (66) −22 (92) −48 (84) −7 (35) −77 (38) −18 (49) −20 (53) −50 (51) 14

29

11

67

21

−33

43

−51

36

1

25

−71

0

−38

11

−21

61

−85

14

−9

14

(94)

(95)

(98)

(90)

(99)

(98)

(101)

(95)


Tell the person to “slow down” or “relax”c

USA SLP UG

Knowledge source Television, radio, films Magazines, newspapers, books

School Doctors, nurses, other specialists Obesity/mental illness subscore Overall impression Obese Mentally ill Want/have Obese Mentally ill Amount known about Obese Mentally ill a b c

17 (54) −39 (93) 18 (99) 0 (101) 100 (0) 4 (100) −38 (19) −17 (31) −22 (38) −12 (36) −95 (11) −96 (17) −95 (15) −2 (34) −1 (28) −3 (46)

−39 (46) −8 (97) −49 (84) −48 (85) −20 (97) −70 (65) −26 (26) −8 (41) −19 (51) 4 (47) −85 (28) −85 (32) −86 (31) 16 (38) 29 (44) 3 (49)

−26 (57) −10 (96) −35 (90) −44 (86) 2 (98) −44 (86) −34 (20) −22 (25) −33 (31) −10 (33) −90 (23) −89 (28) −91 (24) 10 (44) 12 (48) 7 (54)

12 (62) −6 (100) −6 (100) 8 (101) 64 (78) 0 (101) −35 (23) −16 (40) −18 (49) −13 (43) −84 (36) −84 (37) −85 (42) −4 (35) −5 (43) −3 (40)

39 (44) 2 (98) 64 (75) 56 (81) 74 (66) 0 (99) −30 (20) −2 (34) 5 (44) −8 (36) −94 (20) −91 (30) −97 (16) 5 (37) 21 (44) −10 (47)

−14 (53) 28 (95) −8 (101) 2 (100) −20 (97) −70 (71) −29 (23) −6 (28) −3 (32) −9 (45) −74 (29) −72 (35) −76 (29) −8 (41) −1 (47) −16 (52)

−7 (54) 11 (95) −11 (89) 21 (91) −7 (95) −50 (79) −43 (26) −25 (37) −27 (46) −24 (39) −86 (30) −85 (35) −88 (33) −17 (49) −3 (52) −31 (54)

−24

36

−12

7

−13

29

−49

32

−25

61

−36

25

−33

4

−15

11

−23

18

−9

18

−83

14

−82

7

−85

11

2

4

11

18

−10

7


Internet

−15 (57) −40 (93) −42 (92) −44 (91) 72 (70) −20 (99) −34 (18) −6 (33) −25 (42) 14 (45) −98 9 −97 (12) −98 (10) 1 (33) 15 (43) −12 (44)

Median of 162 separate sample means from the POSHA–S database containing 6936 respondents (circa January 2013). Number of independent t-tests for all possible pair-wise comparisons between samples that were statistically significant using the Bonferroni correction (p ≤ .00417). Indicates that ratings are reversed so more positive, accurate, or desirable ratings are higher.

43

44


Table 3 Matrix of percentages of 60 POSHA–S comparisons that were statistically significant using pair-wise t-tests using the Bonferroni correction (p ≤ .00417).

USA SLP UG USA SLP GR USA Non-SLP UG USA Non-SLP GR USA Non-SLP NAT Poland Logoped UG Poland Psych UG Poland Control UG

USA SLP UG

USA SLP GR

USA Non-SLP UG

USA Non-SLP GR

USA Non-SLP NAT

Poland Logoped UG

Poland Psych UG

Poland Control UG

– 12% 32% 8% 10% 27% 37% 35%

12% – 8% 18% 20% 25% 36% 25%

32% 8% – 5% 12% 33% 20% 22%

8% 18% 5% – 7% 32% 27% 23%

10% 20% 12% 7% – 25% 28% 27%

27% 25% 33% 32% 25% – 26% 32%

37% 36% 20% 27% 28% 26% – 2%

35% 25% 22% 23% 27% 32% 2% –

Note: Shaded and unshaded percentages are identical, but the duplicate information is included to facilitate easier comparisons for each sample group.

ratings were, respectively, 30, 44, 16, and −30. The three stuttering ratings were 14–17 points higher than for the psychology students whose OSS, Belief, and Self Reactions subscores were 14, 30, and −1. Their Obesity/Mental Illness subscore was about the same as that of the logopedics students at −29. The control (non-SLP and non-psychology) students from Poland had scores of 19 for OSS, 38 for Beliefs, 0 for Self Reactions, and −43 for Obesity/Mental Illness, or 4–16 points less positive than the logopedics students for the stuttering ratings and 13 points less positive for the Obesity/Mental Illness subscore. The right three set of bars in Fig. 1 illustrate these differences graphically, i.e., the sixth versus the seventh and eighth tracings. The percentage of significant differences are consistent with these larger differences for SLP versus non-SLP Polish students compared to those from the USA. The logopedics students were significantly different from the Polish psychology students for 26% of the POSHA–S ratings and for 32% of those for the Polish control students (Appendix, Columns 16 and 17). By contrast, the Polish psychology and control respondents held very similar attitudes, with only one (or less than 2%) of the POSHA–S ratings being significantly different, i.e., concern about a neighbor who stutters (item) (Appendix, Column 28). The logopedics students had significantly more positive OSSs, Self Reactions subscores, social distance/sympathy (component), Belief subscore (between logopedics versus psychology students only), and who should help (component). The logopedics students rated their knowledge of stuttering higher than the psychology or control students but reported knowing fewer who stuttered than the controls. Sources of information for logopedics students were much greater from school and magazines and books than for the psychology or control students, but not for radio or television. A trend to be more concerned if a sibling stuttered (item) occurred for logopedics students (p = .02 or .03) and was significant for logopedics students versus controls for concern if they, themselves stuttered (item) or feeling comfortable around stuttering (item). Even so, logopedics students were much more likely to regard themselves as someone who should help individuals who stutter (item), compared to psychology or control students. They were also less likely to reject an emotionally traumatic event (item) or to accept genetics (item) as a cause of stuttering than either of the other groups, significantly so compared to the controls. Psychology

Fig. 1. Subscores and Overall Stuttering Scores for the eight samples.


45

students were more likely than logopedics students to agree that medical doctors should help those who stutter (item), and the reverse was true for other people who stutter as those who should help (item). 3.6. SLP and non-SLP students in the USA versus Poland The fourth purpose of the study is related to the third, but in this case sought to identify and measure differences between American and Polish students (see the first and second versus the sixth tracings in Fig. 1). Comparing Polish logopedics students and undergraduate or graduate SLP American students, 27% and 25% of the comparisons were significant (Appendix, Columns 2 and 3). They differed on eight of the same POSHA–S comparisons, e.g., being less positive in accommodating/helping (component) and less accurate in identifying stuttering’s cause (component). In the latter case, the Poles were much less likely to reject an emotional cause (item) but more likely to reject an act of God cause (item). They had more positive impressions of obesity (item) than either American SLP sample. The Polish logopedics students rated their knowledge (component) greater and from more sources (component) than USA SLP undergraduates but rated their concern about themselves (item) or a sibling stuttering (item) as greater. Compared to American SLP graduate students, the Polish logopedics students had less positive ratings for OSS, Belief subscore, and filling in a stuttered word (item). Comparing the non-SLP students from the two countries (but excluding Native Americans), 20% and 22%, respectively, of the Polish psychology and control students differed from American non-SLP undergraduates (Fig. 1, third versus the seventh and eighth tracings and Appendix, Columns 24 and 25). Parallel figures for USA graduates were 27% and 23%, respectively (Fig. 1, fourth versus the seventh and eighth tracings Appendix, Columns 26 and 27). In most cases, the attitudes of the American students were more positive or accurate than those for the Poles. All four country-wise comparisons favored the Americans for distance/sympathy (component), emotional traumatic etiology (item), and concern if a sibling stuttered (item). Three out of four were significant for OSS, concern if oneself stuttered (item), the Belief subscore, and accommodating/helping (component). 3.7. Native American versus non-Native American students in the USA The fifth purpose of the study was to add an intracultural variable to the USA samples and compare a sample of Native American students to the other students, especially the American non-SLP undergraduate and graduate students. The OSS, Belief, Self Reactions, and Obesity/Mental Illness mean ratings for the Native American student sample were: 31, 53, 9, and −35. These were very similar to those of the American non-SLP graduate students, whose corresponding mean ratings were: 31, 54, 8, and −34 (Fig. 1, fourth versus fifth tracings). Only 7% of these ratings were significantly different (Appendix, Column 21). Mean ratings for the American non-SLP undergraduate students were 24, 48, 1, −26 (Fig. 1, third versus fifth tracings), where 12% of the POSHA–S ratings differed (Appendix, Column 20). The Native Americans identified stuttering information from more sources (component) and items relating to the Internet, school, and specialists. They also rated more concern if a neighbor stuttered (item) than either American non-SLP student sample. A trend for somewhat greater tolerance for stuttering characterized Native Americans’ ratings of themselves being less likely to be impatient when a person was stuttering (item) (p = .02 and .0042) although, compared to non-SLP graduate students, they identified less personal experience with stuttering (item). Differences between the Native Americans and the other samples in the study were generally similar to comparisons of non-SLP American students with SLP students in the USA. Native American students were also quite different from Polish students, although the percentages of significant differences were somewhat lower than between the non-Native American students and Polish students. 4. Discussion This ambitious study sought to explore and further delineate factors that affect attitudes of future speech-language pathologists. It combined samples of students majoring in SLP with other majors at different levels of training, different institutions, different countries, different languages, and different cultural backgrounds and heritage. Its goal was to identify issues of positive predispositions to stuttering or people who stutter (i.e., a “halo effect”), training, and cultural influences related to typical American, Native American, and Polish upbringing and heritage. Eight 50-respondent samples of university students filled out the POSHA–S, a measure of public attitude toward stuttering designed to be used worldwide, five samples from the USA and three from Poland. 4.1. “Halo effect” of speech-language pathology students The study clearly identified differences between SLP students and non-SLP students, whether in the USA or in Poland. SLP undergraduates from the USA had mean ratings for the Belief, Self Reactions, and OSS that were 6–9 points more positive than for non-SLP undergraduates. For graduate students, the differences were 10–13 points. Polish logopedics students had 6–16 points more positive Belief, Self Reaction, and OSS ratings than Poles with mixed majors and 14–17 points higher than Polish psychology students. These results are consistent with the hypothesis that either deciding to study SLP or being around others who are in the field results in more positive attitudes toward stuttering than when studying other majors,

46


even those that are involved in helping other people. Additionally, a “halo effect” for stuttering appears not to extend to obesity and mental illness in this study as the POSHA–S items relating to those attributes were generally less positive for SLP students than for non-SLP students. The POSHA–S divides stuttering items and components into Beliefs and Self Reactions. Beliefs pertain to items that are external to the respondent, e.g., “People who stutter are nervous and excitable” or “People who stutter can make friends.” Self Reactions pertain to items that are internal to or involve the respondent, e.g., “If I were talking with a person who stutters, I would make a joke about stuttering” or “A person who stutters should be helped by people like me.” We suspected that differences in the Overall Stuttering Scores (OSS) would reflect changes in both areas, but we hypothesized that a “halo effect,” if present, would more strongly be reflected in differences in Self Reactions. For American SLP students, both Beliefs and Self Reactions were more positive by 10 points compared to American non-SLP students. The hypothesis was partly supported in the differences between Polish logopedics versus control students, where Self Reaction differences were greater than Belief differences, but not between the logopedics versus psychology students. 4.2. Training effects for speech-language pathology students Aside from some changes in maturity and motivation to study speech-language pathology, we assumed that the primary difference between American undergraduate and graduate students in SLP would be the amount of relevant training students had received in fluency disorders. This hypothesis, indeed, was supported as had been reported in India (Dhingra & Bhnatnagar, 2006). For most of the Beliefs, Self Reactions, and OSS comparisons, the data from this study revealed that additional training in SLP also resulted in more positive attitudes toward stuttering. American graduate students in SLP manifested 10-point higher scores for all three POSHA–S summary measures than their SLP undergraduate counterparts. A part of this advantage was likely due to older, more mature students at the graduate level. The SLP graduate students were about four years older on average than their undergraduate counterparts. To estimate the effect of age or maturity alone requires comparing the aforementioned 10-point advantage with any corresponding advantage with the non-SLP graduate versus undergraduate students. In this case, age or maturity resulted in a positive change, but not as large, i.e. a 6–7 point advantage, or about two-thirds of the 10-point advantage seen in SLP students. This was true even though the non-SLP graduate students were older than their undergraduate counterparts by three more years compared to the SLP graduate versus undergraduate students. It is likely that if the SLP graduate students had been measured near the end of their training, training effects would have been more pronounced. Instead, in this study, some of the graduate students in SLP had not yet completed even one graduate course in stuttering, e.g., at TC, where they filled out the POSHA–S on the first day of their fluency disorders class. Nevertheless, many or most of the other SLP graduate students had taken such coursework, and no doubt many had some clinical experience with stuttering clients. Not surprisingly, the SLP graduate students reported more knowledge about stuttering than undergraduates with the greatest source of that knowledge coming from schooling. 4.3. Cultural effects on speech-language pathology and other students We selected both international and intracultural comparisons to begin to explore cultural factors that might affect SLP student attitudes. Polish logopedics students were compared with both American SLP student groups and found to have less positive stuttering attitudes generally. The differences were substantial comparing Polish logopedics undergraduates to American graduate students, with Belief, Self Reactions, and OSS mean ratings from 5 to 20 points lower. Comparing the Poles to American undergraduate SLP students, the Americans had 2- to 5-point higher Belief and OSS scores but Self Reactions scores that were 5 points lower. Among non-SLP students in the two countries, the American undergraduate students manifested Belief, Self Reaction, and OSS ratings from 1–18 points more positive, while the differences between the Polish mixed and psychology majors and American non-SLP graduate students were 8–24 points more positive. Comparing the Polish psychology and control students to a previous study of 268 adults from several locations in Poland (Przepiorka et al., 2013), the results were similar. For example, their OSSs of 14 and 19, respectively, correspond closely to the OSS for the adults, whose OSS was 17. Like the non-SLP Polish students, compared to other samples in the POSHA–S database, the Polish adults (nonstudents) had lower than average scores for physical/mental health and ability to learn. Similarly, they had higher than average scores for life priorities of wanting to be free and having exciting but potentially dangerous experiences (although this item received the lowest rating as in other samples). Only 23% of the adults, compared to 22–26% of the non-SLP students, regarded themselves as intelligent. In terms of regarding themselves as mentally ill or stuttering, the adult percentages were 20% and 11%, compared to the control students at 16% and 20%. Overall, these comparisons suggest that educated Polish citizens are less likely to distinguish between “normal” and “abnormal” mental illness and stuttering than Americans. By contrast the psychology students identified themselves as mentally ill or stuttering at 2% for both attributes, which means that the psychology curriculum at least clarified this ambiguity. We advance the possibility that the Polish students (and Polish adults) are more reticent to rate themselves positively in some ways, perhaps even on an anonymous questionnaire, because they wish to remain humble. Clearly, the prevalence of intelligent people in Poland could be only very little different from other populations, but how the citizenry regards itself reflects some potentially important cultural differences. Wojciszke (2011) reported that a socially recognized value of Polish


47

people is modesty. Wojciszke also points out that Poles manifest tendencies to be self-critical and pessimistic about their mood, political or economic situation, or life situation. Cultural differences observed from Native Americans were much less pronounced. With the most appropriate comparisons with American non-SLP undergraduates and graduates, the POSHA–S summary ratings for Native Americans were 5–8 points more positive than the undergraduates and almost identical to those for graduate students with differences, positive or negative, from 0 to 1 point. Johnson (1944) titled his now-well-known article as “The Indians Have No Word for It” (i.e., “stuttering”). Subscribing to the Whorfian hypothesis that words determine cognitive categories (Whorf, 1956), Johnson reasoned that if Indians did not have a word for stuttering (which Johnson believed), then stuttering would not exist. That notion has been soundly rejected (e.g., Zimmermann, Liljeblad, Frank, & Cleeland, 1983); words for stuttering have been reported in virtually all languages, including Native American languages (Van Riper, 1971), but it raises the intriguing question of whether or not Native Americans might have less negative attitudes toward stuttering. This study suggests that their attitudes were somewhat more positive than might otherwise be expected, if they were to be compared with American non-SLP undergraduates. Their attitudes were more similar to non-SLP graduate students or SLP undergraduates. 4.4. Interpretation The results provide solid support for the assumption that SLP students do indeed appear to have more positive beliefs and reactions to stuttering and people who stutter than other students. The evidence presented here suggests this emanates in part from a “halo effect” regarding the chosen field of study as well as specific training and clinical experiences provided in that course of study. Numerous cultural factors also play roles in SLP—and non-SLP—student attitudes as well. In this study, it is clear that, aside from differences in professional views of stuttering between American and Polish SLPs, there are other factors at work. How else could one explain such differences as much lower ratings for feeling comfortable around a person stuttering or being much more concerned or worried if a sibling or oneself stuttered. In the case of the Polish-American differences, we observed greater differences than we expected. We hypothesized that Polish society and traditions would be similar to those from America even though Poland was part of the Soviet Union for more than a generation. By contrast, we expected greater differences between the Native American students and the typical American students than were observed. At least with respect to the constructs measured by the POSHA–S, this suggests that the Native Americans in this study have adopted the majority “western” American attitudes much more than they have held traditional attitudes that might differ (Garrett & Pichette, 2000; Keltner, Crowell, & Taylor, 2005). It remains to be discovered what attitudes toward stuttering are the result of “western” values versus other religious or cultural influences (Ip et al., 2012; Özdemir et al., 2011b). It seems likely, however, that the Internet and other global influences will foster an increasing trend toward similarities of professional and personal attitudes toward stuttering in the future. Comparing the results of this study with previous research, it can be seen that even the American graduate students in SLP or Polish logopedics students have a great deal of room for improvement in their attitudes toward stuttering. The aforementioned “stuttering stereotype” as well as other beliefs and self reactions, though less pronounced than in non-SLP students, are still evident in their ratings. Data from St. Louis and George (2008) collected from self-help leaders and fluency specialists in the USA indicate that the students in this study are far from the “gold standard” for attitudes suggested in that study. 4.5. Limitations and future directions We acknowledge that our data were obtained from samples of convenience and may not be representative of the various populations of students therein. The representativeness of the data might be further compromised by both online and paperand-pencil administrations, although St. Louis (2012b) showed that this difference in administration did not affect mean POSHA–S ratings. It is also possible that the lack of information regarding the Native Americans’ specific universities and majors limits inferences to this population. For example, their mean level of education was lower than that of other American undergraduates even though they were about four years older. The same lack of specific information about universities and majors characterized about half of the non-SLP graduate students. Undergraduate and graduate students in the USA did not come in the same proportions from each of the various universities which likely had minimal effects on the results. The study does not provide a precise estimate of the amount and kind of training received for undergraduate and graduate students in either the SLP or non-SLP groups. Such information would further refine the inferences that might be made. Even so, this study is more comprehensive than any other published studies on SLPs’ or SLP students’ attitudes toward stuttering of which we are aware. Thus, in spite of these limitations (which also characterize much of the previous literature), it provides needed evidence on characteristics of future SLPs who will be most intimately involved in the treatment of stuttering. We encourage further research on stuttering attitudes of SLP students and SLP practitioners using the POSHA–S. Such investigations could provide important comparisons with past and future studies using the standard instrument. We suggest that fruitful avenues for such research would be to investigate the effects on stuttering attitudes of specific training content in stuttering coursework, clinical observation of stuttering treatment, beginning experience in stuttering practicum,

48


extensive experience in stuttering treatment, and positive versus negative outcomes in one’s clients. We also encourage more comparative studies of SLP and SLP student attitudes toward stuttering as a function of different cultural and/or geographic influences. CONTINUING EDUCATION Stuttering attitudes of students: Professional, intracultural, and international comparisons QUESTIONS 1. American college students with the most positive attitudes toward stuttering are: a. SLP undergraduate students b. SLP graduate students c. Non-SLP undergraduate students d. Non-SLP graduate students 2. The so-called “halo effect” in this study refers to: a. Being a Native American b. Being a Polish student c. Being an SLP major d. Having an advanced degree 3. The cultural difference that made the greatest difference in student attitudes toward stuttering in this study was: a. Urban upbringing and living b. Rural upbringing and living c. Native American upbringing and living d. Polish upbringing and living 4. Relative to affecting attitudes toward stuttering in this study, training in fluency disorders appeared to: a. Make no difference either in the USA or in Poland b. Make a difference in the USA but not in Poland c. Make a difference in Poland but not in the USA d. Make a difference in the USA and in Poland 5. Native American students in this study: a. Manifested Overall Stuttering Scores most similar to American Non-SLP graduate students b. Manifested the most negative Overall Stuttering Scores of all the groups measured c. Manifested Overall Stuttering Scores most similar to Polish control students d. Indicated that they had no word in their language for “stuttering” Financial Disclosures Dr. St. Louis as author of the paper “Stuttering Attitudes of Students: Professional, Intracultural, and International Comparisons” acknowledges funding from a grant from the College of Education and Human Services at West Virginia University. None of the other authors for this paper “Stuttering Attitudes of Students: Professional, Intracultural, and International Comparisons” by St. Louis, K.O., Przepiorka, A.M., Beste-Guildborg, A., Williams, M.J., Blachnio, A., Guendouzi, J., Reichel, I.K., & Ware, M.B. reported any relevant financial relationships to be disclosed.

Non-Financial Disclosures None of the authors for this paper “Stuttering Attitudes of Students: Professional, Intracultural, and International Comparisons” by St. Louis, K.O., Przepiorka, A.M., Beste-Guildborg, A., Williams, M.J., Blachnio, A., Guendouzi, J., Reichel, I.K., & Ware, M.B. reported any relevant non-financial relationships to be disclosed.

Acknowledgment Presentation of the data from this study was partly funded by a grant from the College of Education and Human Services at West Virginia University.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jfludis. 2013.10.001.


49

References Abdalla, F., & St. Louis, K. O. (2012a). Arab school teachers’ knowledge, beliefs and reactions regarding stuttering. Journal of Fluency Disorders, 37, 54–69. Beste-Guldborg, A., & Campanale, N. (2012 November). Eastern American Indian perceptions of stuttering. In Poster presented at the Annual Convention of the American Speech-Language-Hearing Association, Atlanta, GA. Beste-Guldborg, A., St. Louis, K. O., & Campanale, N. (2012 November). Knowledge, perceptions, & beliefs of northern plains Indians toward stuttering. In Poster presented at the Annual Convention of the American Speech-Language-Hearing Association, Atlanta, GA. Brisk, D. J., Healey, E. C., & Hux, K. A. (1997). Clinicians’ training and confidence associated with treating school-age children who stutter: A national survey. Language, Speech, and Hearing Services in Schools, 28, 164–176. Cohen, J. (1988). Statistical power for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cooper, E. B. (1975 November). Clinician attitudes toward stuttering: A study of bigotry? In Paper presented at the Annual Convention of the American Speech-Language-Hearing Association, Washington, DC. Cooper, E. B., & Cooper, C. S. (1985). Clinician attitudes towards stuttering: A decade of change (1973–1983). Journal of Fluency Disorders, 10, 19–23. Cooper, E. B., & Cooper, C. S. (1996). Clinician attitudes toward stuttering: Two decades of change. Journal of Fluency Disorders, 21, 119–135. Cooper, E. B., & Rustin, L. (1985). Clinicians attitudes toward stuttering in the United States and Great Britain: A cross-cultural study. Journal of Fluency Disorders, 10, 1–17. Crichton-Smith, I., Wright, J., & Stackhouse, J. (2003). Attitudes of speech & language therapists towards stammering: 1985 and 2000. International Journal of Language & Communication Disorders, 38(3), 213–234. Dhingra, Y., & Bhnatnagar, V. (2006). Attitudes of speech-language pathology students toward persons with disability. Asia Pacific Rehabilitation Journal, 17, 94–99. Fagerland, M. W. (2012). T-tests, nonparametric tests, and large studies—A paradox of statistical practice? BMC Medical Research Methodology, 12, 78–84. Flynn, T. W., & St. Louis, K. O. (2011). Changing adolescent attitudes toward stuttering. Journal of Fluency Disorders, 36, 110–121. Garrett, M. T., & Pichette, E. F. (2000). Red as an apple: Native American acculturation and counseling with or without reservation. Journal of Counseling and Development, 78, 3–13. Georgieva, D. (2010). Education of logopedists or speech-language pathologists in Bulgaria, Greece, Macedonia, Poland and Russia. Folia Phoniatrica et Logopaedica, 62, 217–222. Hughes, S., Gabel, R., Irani, F., & Schlagheck, A. (2010). University students’ explanations for their descriptions of people who stutter: An exploratory mixed model study. Journal of Fluency Disorders, 35, 280–298. Ip, M. L., St. Louis, K. O., Myers, F. L., & An Xue, S. (2012). Stuttering attitudes in Hong Kong and Mainland China. International Journal of Speech-Language Pathology, 14, 543–556. Johnson, W. (1944). The Indians have no word for it: I. Stuttering in children. Quarterly Journal of Speech, 30, 330–337. Keltner, B. R., Crowell, N. A., & Taylor, W. (2005). Attitudes about disabilities in a southeastern American Indian Tribe. American Indian Culture and Research Journal, 29, 57–74. Lass, N. J., Ruscello, D. M., Pannbacker, M., Schmitt, J. F., & Everly-Myers, D. S. (1989). Speech-language pathologists’ perceptions of child and adult female and male stutterers. Journal of Fluency Disorders, 14, 127–134. Lemert, E. M. (1953). Some Indians who stutter. Journal of Speech and Hearing Disorders, 18, 168–174. Mavis¸, I˙ ., St. Louis, K. O., Özdemir, S., & To˘gram, B. (2013). Attitudes of Turkish speech and language therapists toward stuttering. Journal of Fluency Disorders, 38, 157–170. Özdemir, R. S., St. Louis, K. O., & Topbas¸, S. (2011a). Public attitudes toward stuttering in Turkey: Probability versus convenience sampling. Journal of Fluency Disorders, 36, 262–267. Özdemir, R. S., St. Louis, K. O., & Topbas¸, S. (2011b). Stuttering attitudes among Turkish family generations and neighbors from representative samples. Journal of Fluency Disorders, 36, 318–333. Przepiorka, A., Blachnio, A., St. Louis, K. O., & Wozniak, T. (2013). Public attitudes toward stuttering in Poland. International Journal of Language and Communication Disorders, 48, 703–714. Ragsdale, J. D., & Ashby, J. K. (1982). Speech-language pathologists’ connotations of stuttering. Journal of Speech and Hearing Research, 25, 75–80. Ruscello, D. M., Lass, N. J., French, R. S., & Channel, M. D. (1989–1990). Speech-language pathologists’ perceptions of stutterers. National Student Speech Language Hearing Association Journal, 17, 86–89. Snidecor, J. (1947). Why the Indian does not stutter. Quarterly Journal of Speech, 33, 493–495. Stewart, J. L. (1960). The problem of stuttering in certain North American Indian societies. Journal of Speech and Hearing Disorders Monograph Supplement, 6. St. Louis, K. O. (2005). A global project to measure public attitudes of stuttering. The Asha Leader, 10(12–13), 22. St. Louis, K. O. (2008 November). Surveying public attitudes toward stuttering: Considerations of sample size. In Technical paper presented at the Annual Convention of the Speech-Language-Hearing Association, Chicago, IL. St. Louis, K. O. (2011a). International project on attitudes toward human attributes (IPATHA). Morgantown, WV: Populore. http://www.stutteringattitudes.com St. Louis, K. O. (2011b). The Public Opinion Survey of Human Attributes-Stuttering (POSHA–S): Summary framework and empirical comparisons. Journal of Fluency Disorders, 36, 256–261. St. Louis, K. O. (2012a). Male versus female attitudes toward stuttering. Journal of Communication Disorders, 45, 246–253. St. Louis, K. O. (2012b). POSHA–S public attitudes toward stuttering: Online versus paper surveys. Canadian Journal of Speech-Language Pathology and Audiology, 36, 116–122. St. Louis, K. O. (2012c). Research and development for a public attitude instrument for stuttering. Journal of Communication Disorders, 45, 129–146. St. Louis, K. O., Andrade, C. R. F., Georgieva, D., & Troudt, F. O. (2005). Experience and personal report about an international cooperation research—Brazil, Bulgaria and Turkey—Attitudes toward stuttering. Pró-Fono Revista de Atualização Cientifica, 17, 413–416. St. Louis, K. O., & Durrenberger, C. H. (1993). What communication disorders do experienced clinicians prefer to manage? Asha, 35, 23–31. St. Louis, K. O., Filatova, Y., Cos¸kun, M., Topbas¸, S., Özdemir, S., Georgieva, D., et al. (2011). Public attitudes toward cluttering and stuttering in four countries. In E. L. Simon (Ed.), Psychology of stereotypes (pp. 81–113). Hauppauge, NY: Nova Science Publishers. St. Louis, K. O., & George, R. D. (2008 November). Attitudes toward stuttering: In search of a gold standard. In Poster presented at the Annual Convention of the Speech-Language-Hearing Association, Chicago, IL. St. Louis, K. O., & Lass, N. J. (1981). A survey of communicative disorders students’ attitudes toward stuttering. Journal of Fluency Disorders, 6, 49–79. St. Louis, K. O., Lubker, B. B., Yaruss, J. S., Adkins, T. A., & Pill, J. C. (2008). Development of a prototype questionnaire to survey public attitudes toward stuttering: Principles and methodologies in the first prototype. The Internet Journal of Epidemiology, 5(2) http://www.ispub.com/journal/theinternet-journal-of-epidemiology/volume-5-number-2/development-of-a-prototype-questionnaire-to-survey-public-attitudes-toward-stutteringprinciples-and-methodologies-in-the-first-prototype.html St. Louis, K. O., Lubker, B. B., Yaruss, J. S., & Aliveto, E. F. (2009). Development of a prototype questionnaire to survey public attitudes toward stuttering: Reliability of the second prototype. Contemporary Issues in Communication Sciences and Disorders, 36, 101–107. St. Louis, K. O., Reichel, I., Yaruss, J. S., & Lubker, B. B. (2009). Construct and concurrent validity of a prototype questionnaire to survey public attitudes toward stuttering. Journal of Fluency Disorders, 34, 11–28. St. Louis, K. O., & Roberts, P. M. (2010). Measuring attitudes toward stuttering: English-to-French translations in Canada and Cameroon. Journal of Communication Disorders, 43, 361–377. Stonehouse, J. M., & Forrester, G. J. (1998). Robustness and the t and U tests under combined assumption violations. Journal of Applied Statistics, 25, 63–74.

50


Swartz, E., Gabel, R., & Irani, F. (2009). Speech-language pathologists’ attitudes towards people who stutter. Canadian Journal of Speech-Language Pathology and Audiology, 33, 99–105. Tellis, G. M., Bressler, L., & Emerick, K. (2008). An exploration of clinicians views about assessment and treatment of stuttering. Perspectives on Fluency and Fluency Disorders, 18, 16–23. http://dx.doi.org/10.1044/ffd18.1.16 Tellis, G. M., & Tellis, C. M. (2003). Multicultural issues in school settings. Seminars in Speech and Language, 24, 21–26. Van Riper, C. (1971). The nature of stuttering. Englewood Cliffs, NJ: Prentice-Hall. Whorf, B. L. (1956). In J. B. Carroll (Ed.), Language, thought, and reality: Selected writings of Benjamin Lee Whorf. Cambridge, MA: MIT Press. Wojciszke, B. (2011). Psychologia społeczna. Warszawa: Wydawnictwo Naukowe Scholar. Woods, C. L., & Williams, D. E. (1971). Speech clinicians’ conceptions of boys and men who stutter. Journal of Speech and Hearing Disorders, 36, 225–234. Woods, C. L., & Williams, D. E. (1976). Traits attributed to stuttering and normally fluent males. Journal of Speech and Hearing Research, 19, 267–278. Yairi, E., & Williams, D. E. (1970). Speech clinicians’ stereotypes of elementary-school boys who stutter. Journal of Communication Disorders, 3, 161–170. Yaruss, J. S., & Quesal, R. W. (2002). Academic and clinical education in fluency disorders: An update. Journal of Fluency Disorders, 27, 43–63. Zimmermann, G., Liljeblad, S., Frank, A., & Cleeland, C. (1983). The Indians have many terms for it: Stuttering among the Bannock–Shoshoni. Journal of Speech and Hearing Research, 26, 315–318.

Kenneth O. St. Louis, Ph.D., founder of the International Project on Attitudes Toward Human Attributes, has carried out research on attitudes of clinicians and the general public toward stuttering and other potentially stigmatizing conditions. He has published on numerous aspects of stuttering and cluttering for more than 35 years. Aneta M. Przepiorka, Ph.D., is an assistant professor in the Department of Emotion and Motivation Psychology at John Paul II Catholic University of Lublin in Poland. As a speech-language therapist she specializes in stuttering disorders. Her scientific interests also include social psychology and cross-cultural psychology. Ann Beste-Guldborg, Ph.D., CCC-SLP, is an assistant professor of Communication Disorders at Minot State University. Her research interests include stuttering perceptions, cluttering, effective pedagogy, and language as it relates to academic literacy. Mandy J. Williams, Ph.D., CCC-SLP, is a speech pathologist and an assistant professor at the University of South Dakota. Her research interests include fluency disorders, temperament, and psychosocial factors influencing treatment outcomes. Agata Blachnio, Ph.D., is in the Department of Emotion and Motivation Psychology at the John Paul II Catholic University of Lublin. Her main areas of scientific interests are cross-cultural psychology, psychology of emotions, and such human behaviors as dishonesty, disloyalty, and academic cheating. Her hobbies include travel and photography. Jacqueline Guendouzi, Ph.D., is a clinical linguist who received her doctorate from Cardiff University. Currently she is a Professor in the Communication Sciences and Disorders department at Southeastern Louisiana University. Her research interests include psycholinguistics and discourse analysis. Isabella K. Reichel, Ed.D., CCC-SLP/A, has specialized in the treatment of stuttering for over three decades. She is the chair of the Committee of International Representatives of the International Cluttering Association. Her studies on integrating additional curricula into fluency disorders courses were published and presented nationally and internationally. Mercedes B. Ware, B.S., is a speech-language pathology master’s student at Ohio University. She received her baccalaureate degree from West Virginia University.

Professional and personal attitudes of physiotherapy students toward disabled persons.

Medical students' attitudes toward people with dementia: an international investigation.

Public attitudes toward stuttering in Poland.

International comparisons of physicians' salaries.

Measuring interprofessional competencies and attitudes among health professional students creating family planning virtual patient cases.

Effect of "attitudinal content" on formation of professional attitudes in medical students.

Modifying attitudes of Arab school teachers toward stuttering.

Intention to seek professional psychological help among college students in Turkey: influence of help-seeking attitudes.

Professional attitudes of doctors and medical teaching.

Smoking and attitudes towards its cessation among native and international dental students in Lithuania.

Comparisons between the attitudes of medical and dental students toward the clinical importance of gross anatomy and physiology.

International comparisons of acute myocardial infarction.

International comparisons of acute myocardial infarction.

Are paramedic students ready to be professional? An international comparison study.

International comparisons of acute myocardial infarction.

Ranking health between countries in international comparisons.

Health care systems: lessons from international comparisons.

International comparisons: selected mental health data.

Acquired immune deficiency syndrome: international attitudinal comparisons.

Do professional attitudes change during medical education?

Habits and attitudes of public health students.

International comparisons of perinatal and infant mortality rates.

Learning from others: international comparisons of death and dying.

Author attitudes to professional medical writing support.