Introduction
Research conducted in Western contexts has found that students from culturally diverse and non-English-speaking backgrounds perform poorly in university degrees for health professionals compared with students from the dominant culture who speak English [
1]. Such students may be domestic students from culturally and linguistically diverse (CALD) backgrounds or international students who are not residents of their country of study. In 2014, there were 4.5 million international students globally, and 53 % of these were from Asia [
2,
3]. In Australia, where this study was conducted, international students comprise 20 % of university students, but limited data exist about the participation of domestic CALD students in higher education [
3]. Studies of health professional students have established that being CALD [
1] and speaking English as an additional language [
4,
5] are predictors of poor academic performance, but the relationship for international students is unclear. Whilst previous studies have focussed on academic outcomes, little is known about the performance of these students in their professional placements.
Health professional education programmes require students to undertake placements situated within a range of health and community settings. These provide opportunities to work directly, and in an authentic way, with clients and situations related to the placement setting [
6]. Placements facilitate opportunities for students to develop and practice professional knowledge, attitudes and skills, embedded within the sociocultural practices of their profession [
7]. In many health professions, students are supervised in their placements by qualified practitioners who provide direct educational support and workplace based competency assessment during the placement [
8,
9]. Findings from qualitative studies suggest that in Western, English-speaking contexts CALD students, international students or those speaking English as an additional language may experience learning challenges in placements [
10‐
13].
To succeed in these placements, international students [
10,
14] and domestic CALD students [
13,
15] must adjust to Western settings; manage culturally unfamiliar approaches to learning, assessment and supervision; and conduct their practice in English. Acculturation describes the process of cultural and psychological change that occurs following contact between individuals or groups from dissimilar cultures. This process of adjustment may influence learning outcomes for students in placements [
16]. The extent of an individual’s acculturation is influenced by several factors including their ethnic and cultural similarity to the receiving community, proficiency in the majority language, social supports, adaptation strategies, and the attitudes of the host culture [
16‐
18]. Students who must undertake multiple or extensive adjustments may experience acculturative stress that affects their ability to cope with these changes [
17] and impacts their academic performance [
18]. In placements, all students must adjust for the context and practice of each setting. However, students who are also undertaking extensive acculturative adjustments may need to complete additional learning [
10,
18] and this may influence their placement learning outcomes.
Previous studies have used acculturation to explain educational, social, cultural and language adjustments made by international students, and how stress related to acculturation affects adaptation to learning at university, but have not investigated the critical learning environment in placements [
18]. Acculturation may also affect domestic CALD students as these changes are known to continue for several generations following migration [
17]. This suggests that acculturative factors may also impact educational outcomes for these students [
18]. Studies identifying that international and domestic students who are CALD experience difficulties learning on placement and meeting performance expectations are small and preliminary in nature [
11,
13,
15]. Students’ placement success does relate to their English as this can impact communication with patients and health teams [
12,
14]. Speaking English as an additional language is an acculturative factor that has been identified to predict academic performance for health professional students [
4,
5,
18]. However, speaking a language other than English at home (LOTEH) includes native English speakers, and is also a contextual variable that affects acculturation [
17], but is rarely included in educational studies. Speaking English as an additional language or a LOTEH may be permutations of English language status that predict placement outcome. Factors related to students’ cultural background or international student status may also hinder placement performance due to the process of acculturation, but little empirical data exist to support this [
11,
15].
Few studies have examined the predictive nature of CALD, being an international student, speaking English as an additional language and speaking a LOTEH together, and no previous research has investigated these for professional placement outcomes [
1,
4,
19]. Investigating outcomes in placements is critical as they are key for the preparation of health professionals who can function competently in the workplace [
6]. The placement context may pose additional acculturative challenges for students who are culturally and linguistically diverse [
14]. This study aims to determine whether factors known to influence acculturation [
17], including students’ background or English status, predict their outcomes in professional placements.
We hypothesized the following:
-
English status categories (‘English as an additional language’ and ‘LOTEH’) and
background status categories (‘international students’ and ‘CALD’) are separately
predictive of poor placement outcome compared with English as first language, domestic or non-CALD students.
-
Interactions between background and English status variables will additionally predict ‘at risk’ placement outcome.
Results
The proportion of students for each predictor group are displayed in Table
1. The majority of international students were from countries in South-Eastern Asia (Singapore = 43.4 %; Hong Kong (SAR of China) = 15.1 %; Malaysia = 10.4 %); whereas ‘perceived CALD’ students were predominantly from Australia (43.0 %), followed by Singapore (16.7 %) and Hong Kong (10.4 %).
‘Perceived CALD’, ‘defined CALD’, ‘international student’, ‘English as an additional language’ and ‘LOTEH’ predictors had proportionately more ‘at risk’ placement outcomes than their alternative categories, as demonstrated in Table
2. Interactions between background and English status predictors had proportionately more ‘at risk’ placements than those from single predictors.
The unconditional model revealed that the odds of any student having an ‘at risk’ placement outcome was small but significant (OR = 0.07;
p ≤ 0.001), as shown in Table
3. All predictors tested in the simple multilevel model were significant. However, ‘international student’ and ‘English as an additional language’ were the only significant student predictors of ‘at risk’ placement outcome in the final multiple predictor model. In this model, international students were 1.83 times more likely to have an ‘at risk’ placement than domestic students (
p < 0.05); and English as an additional language students were 2.38 times more likely than English as a first language students (
p < 0.001). Interactions between background and English status predictors were tested in the multiple multilevel analysis, but ‘perceived CALD and English as an additional language’, ‘Perceived CALD and LOTEH’ and ‘International student and LOTEH’ interactions were eliminated due to estimation problems and high standard error. The remaining interaction between ‘international student and English as an additional language’ was not significant and did not enhance the model. University and programme variables used as controls were not significant, except for ‘Year of placement’ where students who completed a placement in 2012 were significantly more likely to have an ‘at risk’ placement in each model iteration, and 1.63 times in the final model (
p < 0.01).
Atkinson’s
R was calculated to determine the importance of the ‘international student’ and ‘English as an additional language’ predictors to the placement outcome model. Although the ‘international student’ and ‘English as an additional language’ variables were significant predictors of ‘at risk’ placement, their
R values of 0.02 and 0.03 respectively indicate that these cultural and linguistic characteristics have a small contribution to placement outcome overall [
24]. The
R value for 2012 placement year was 0.02.
The median odds ratio was calculated to determine the extent that the individual probability of an ‘at risk’ placement outcome was influenced by heterogeneity among students [
27,
28]. The median odds ratio value of 2.86 in the final model indicated that the residual heterogeneity among students was of greater relevance for placement outcome than ‘international student’ or ‘English as an additional language’ status.
CALD and international student groups are rarely delineated in previous studies, and differences between these variables in the multiple multilevel analysis prompted investigation of the proportion of ‘perceived CALD’ and international students who also spoke English as an additional language. Post-hoc analysis conducted using the ‘
N-1’chi-squared test found no significant difference in the proportion of ‘English as an additional language’ students in the ‘international student’ and ‘perceived CALD’ groups (χ
2 = 0.150;
p = 0.7) [
29].
Discussion
In this study, acculturation theory was used as a lens to identify predictors of placement outcome for health professional students studying in a Western, English-speaking country. Specifically, background status variables of ‘perceived CALD’ or ‘international student’ and English status variables of ‘English as an additional language’ or ‘LOTEH’ were modelled to determine if they predicted poor placement outcome. We also investigated whether interactions between background and English status variables additionally predicted ‘at risk’ placement outcome. Interactions between ‘international student and English as an additional language’ were not significant in the multiple predictor model, but estimation of other interactions was limited.
In the simple multilevel logistic analysis, the single variables each significantly predicted ‘at risk’ placements. However, only ‘English as an additional language’, ‘international student’ and ‘placements in 2012’ remained significant predictors in the final model which utilized a more powerful multiple multilevel logistic analysis [
23,
26]. Longitudinal data are required to determine if the variation in ‘at risk’ placement outcome in 2012 is more than anomalous. Whilst these variables were significant predictors of ‘at risk’ placement, their importance for the overall placement outcome was small, and the median odds ratio indicated that heterogeneity among individual students is also a relevant factor for placement outcome [
27]. Placement outcome is likely to be predicted by many untested factors related to student academic performance, placement setting, supervision and assessment. This study did not intend to model a broad range of variables, but rather to determine whether the acculturative factors identified predicted placement outcome. The contribution of the ‘international student’ and ‘English as an additional language’ predictors in the multiple multilevel analysis indicates that these should be included in future studies. Individual acculturative factors such as personal attributes and values, sociodemographic status and coping strategies may also contribute to placement outcome [
17,
18], and should also be considered for future studies.
Students who speak English as an additional language were more likely to have an ‘at risk’ placement outcome but students speaking a LOTEH were not. This finding extends previous research linking speaking English as an additional language with poor academic performance [
4,
5] into the professional placement context and lends support to qualitative findings identifying English proficiency as influencing students’ placement success [
10,
13,
14]. Acculturation theory predicts that challenges associated with using English as an additional language result in stressors for students which inhibit their adjustment to new educational environments, even when their cultural background is constant [
17]. Adjustments for language are likely to be greater for students who speak English as an additional language than students who speak a LOTEH, as this includes native English speakers. Adjustments may be intensified in placement settings, where the client focussed nature of services reduce opportunities to scaffold students’ professional communication skills [
6]. Whilst speaking English as an additional language has been linked with poor academic performance [
4,
5], it is not a direct measure of English proficiency. Measures of proficiency may better reflect communication challenges in placement settings as speaking English as an additional language does not reflect a students’ application of English [
11,
13,
15]. Further research is needed to examine the relationship between English as an additional language, English proficiency and placement outcome.
Previous studies have identified CALD students as having a greater risk of academic difficulties, but these studies have not clearly differentiated between CALD and international students or addressed placement outcomes [
1]. In this study, being an international student was more predictive of ‘at risk’ placement outcome than belonging to ‘perceived CALD’. As ‘English as an additional language’ also predicted ‘at risk’ placement outcome, differences in the proportion of English as an additional language speakers in the ‘international student’ and ‘perceived CALD’ groups may have assisted to explain their difference in placement outcome. However, they were not significantly different. These groups were distinguished by domestic students born in Australia, who comprised almost half of the ‘perceived CALD’ group. The identification of these Australian students as CALD may relate to the cultural background of their family. Domestic CALD students were also included in study cohorts of previous research that identified CALD as a factor predicting academic failure [
1,
19]. The factors that influence the extent of a student’s acculturation to the receiving culture are likely to impact international students differently to domestic CALD students. These include their prior education experiences, cultural similarity with and fluency in the language of the host country [
17]. This may account for the finding that being an international student predicted ‘at risk’ placement performance, but being ‘perceived CALD’ did not. The time limited and specific intentions of the international students’ sojourn [
18] or the greater acculturation of domestic CALD students to the home culture [
17] may also influence outcomes.
Domestic CALD students have been noted to experience challenges in professional placements [
13]. However, international students may need to undertake greater acculturative adjustment to operate successfully within these environments compared with those of their home culture. These factors may be independent of cultural or language background, as students from Western, English-speaking backgrounds also undertake adjustment to accommodate their learning in foreign professional placement settings [
10]. International students may therefore find adjustment to the language, learning, cultural and/or organizational expectations of professional placements more challenging than their domestic peers, regardless of their background [
10,
11,
14]. This may have a greater contribution to international students’ placement outcomes than previously understood. Acculturative factors may therefore assist to explain differences in placement outcomes between the ‘international student’ and ‘perceived CALD’ groups; however, further research is needed to confirm this.
Frequency findings for interactions between ‘international student and English as an additional language’ predictors indicated that these students may be more likely to have an ‘at risk’ placement outcome than other student groups. International students who also speak English as an additional language may need to undertake more extensive acculturative adjustments than those represented in the other groups [
16,
18]. However, in the multiple multilevel analysis, interactions between ‘international student and English as an additional language’ were not significant. Therefore, students who are international students and speak English as an additional language are not more likely to have an ‘at risk’ placement outcome than students who are either international students or speak English as an additional language. This suggests that the ‘international student’ and ‘English as an additional language’ variables independently predict placement outcome, and both should be included in future studies of placement outcome.
These findings that suggest that international students and students who speak English as an additional language may require support for their learning and performance in placements. However, there is little empirical evidence of what strategies would be effective. Future research should clarify learning needs for these students and explore appropriate placement supports [
10,
13].
Limitations
The identity, background and placement outcome of students who did not consent for this study is unknown, so it is not possible to determine if the results represent speech-language pathology students from the participating universities. However, consent rates did not markedly differ across the three universities, and the proportion of international and domestic students within this study approximates that identified in previous speech-language pathology research [
30].
There are likely to be many predictors of placement outcome including factors related to student, supervisor, placement context and the assessment of competency, but this study did not aim to include these in modelling. Future studies should include a broader range of predictors, including ‘international student’ and ‘English as an additional language’ to determine a model of placement outcome.
Backwards elimination was utilized to remove non-significant predictors from the modelling. While this potentially introduces bias, it is an accepted strategy where research does not aim to identify a causal model [
26]. Multicollinearity limited modelling of students in the ‘defined CALD’ group, which constrained the analysis of CALD as a predictor. Estimation problems and standard error also precluded investigation of interactions between predictors in the multiple multilevel analysis.