JCPSLP Vol 20 No 3 November 2018

created and labelled accordingly. Normality testing of interval data influenced the type of inferential test chosen. The Mann-Whitney test (two factors) and Kruskal-Wallis (more than two, including post-hoc testing to reveal specific subgroup differences) were used with the non-parametric data. For the interval and normally distributed data t tests (two factors) and one-way analysis of variance (ANOVA; more than two factors, with post-hoc testing) were run. Significance was established at a p-value of less than .05. Effect sizes were calculated for all inferential tests. Thematic analysis Thematic analysis (Braun & Clarke, 2006) was conducted on responses to open ended questions, with measures taken to enhance rigour. The authors initially analysed and coded 20 questionnaires independently. The derived coding structure was then used by an external colleague to classify responses in five additional questionnaires, with consensus achieved through discussion. The first author then coded all responses using the agreed coding framework. Memos were written during the analytic process to further develop interpretations and understanding. Thematic maps were constructed to visualise the findings and were reviewed multiple times in conjunction with the original data. Results Responses were received from 117 SLPs, with 80 full responses (from 77 females, 3 males) and 37 (gender unspecified) partial responses. Demographics Data included a range of practice demographics (Table 2). Respondents were typically women holding a bachelor’s degree, at the lower end of experience, (mode = 1 year [12 cases]; x = 8 yrs; range 0–40 yrs), practising in a client- focused role with children rather than adults, and working inter-professionally. A statistically significant higher proportion of clinicians from South Australia responded to the survey compared to the wider Australian speech pathology population, consistent with the researchers’ geographical location ( p < .0001). This bias also extended to the sample clinicians’ location, with rural-based clinicians being over represented ( p = .038). Practitioners reported working across multiple areas of practice and in several service areas concurrently, which complicated interpretation. Responses were therefore analysed according to collapsed categories: 1) Area of practice: paediatric focus, adult focus or both and 2) Service delivery: various child settings, mixed settings, community centre and outpatient, acute and rehabilitation. Smaller groups were combined in a final category for inferential analysis. Knowledge of the ICF Practitioners rated their knowledge of the ICF to be: little (N = 18; 16%); basic (N = 30; 27%); good (N = 38; 34%); solid (N = 11; 10%); and comprehensive (N = 15; 13%). Use and perceived utility of the ICF Exploratory factor analysis was used to identify the relationships underlying each of three multicomponent scale questions: components [of the ICF] (Survey section I. 3); adoption [of the ICF] (Survey section I. 4a); and utility [of the ICF] (Survey section II. 2). Two factors were obtained for each section using the Kaiser Meyer Olkin (KMO) statistic, accounting for 56%, 68% and 56% of the variance respectively. One factor was extracted for further adoption (Survey section I. 4b), accounting for 56% of variance.

specifically investigated whether, almost two decades after its introduction, SLPs value and use this tool. In order to inform future education in and support for using the tool, the aims of this exploratory study were to determine the current status of the following in relation to Australian SLPs: • knowledge of and support for the ICF • the extent of explicit ICF use • patterns in ICF uptake • perceived advantages and disadvantages • facilitators and barriers experienced in the application of the ICF. Consideration of these findings should help ascertain whether the ICF has instigated a cultural shift from the traditional medical model to a more holistic way of practising in SLP practice. Methods This exploratory mixed methods, survey-based study received ethics approval from the Human Research Ethics Committee of Flinders University (EC00194; # 7133). Return of the online questionnaire, supported by LimeSurvey (Schmitz, 2015), implied consent to participate. All Australian SLPs were eligible to contribute, and recruitment occurred primarily via a SPA newsletter invitation. Research instrument An existing survey (Stewart et al., 2013) was adapted, with permission, to reduce redundancy, add literature-supported detail regarding potential uses and benefits of the ICF, and improve data retrieval processes. The questionnaire comprised of 15 questions divided across three main sections – two sections addressing questions of use and utility, and one gathering work-related demographic information. An outline of the questions for the first two sections is provided in Table 1. Pilot testing suggested that survey completion took 15–20 minutes. Statistical analysis Data were analysed using IBM SPSS (IBM Corp, 2015). Partial data were used where possible. Descriptive statistics were calculated. Chi-square goodness of fit was used to reveal any bias in the data. Exploratory factor analysis was conducted and, where relevant, composite scores were

Table 1. Survey questions

Section

Questions

Sub- questions

Question types

ICF use

1. Duration of

Multiple choice

implementation 2. Knowledge level 3. Use of components 4. Adoption of the ICF 5. Most influential components

7 24

Likert scale

Free response

ICF utility 6. Advantages and disadvantages 7. Perceptions on utility 8. Barriers and facilitators to use

Free response Likert scale

10

Free response

112

JCPSLP Volume 20, Number 3 2018

Journal of Clinical Practice in Speech-Language Pathology