Migrant integration: research on foreign language interpretation services - Section 3: methodology

The main aim of this research is to review and seek possible areas for improvement regarding the availability and adequacy of foreign language interpretation services for forced migrants in Wales. In order to respond to the key research themes, service provision was reviewed using a systems approach and grounded theory. 

The qualitative approach focused on the knowledge and perceptions of participants working in and receiving the service. This meant engaging with participants to reflect on their experiences within the system moving onto their assessment of how the system can be improved.

The qualitative approach focusing on the knowledge and perceptions of participants was considered most suited to addressing the key research themes. This is described as ‘perception-based’ methodology. 

A key strength of a perception-based study with stakeholders is that much of this knowledge and experience can provide useful information on a range of more practical delivery issues and be fed back to improve the system and communications. A recognised limitation of this methodology is that at times participants’ views may be swayed by their subjective interests and/or may be inaccurate, and/or may involve misinterpretation of the questions.

This type of perception-based study fits into a grounded theory approach (Holton, J.A. and Glaser, B.G., Eds. 2012). The research is founded in perceptions and information provided by participants through the project interviews. This means that the themes are allowed to emerge from the data, rather than researchers applying or testing a pre-existing hypothesis. This is suited to drawing out complex, interlocking issues such as those presented in the interpreter support system. Grounded theory also allows for the presentation of different angles on the same phenomenon and the effects of perceptions on behaviours within a system. This was important as the stakeholder groups were positioned at different locations in the overall system and thus had different perspectives, interests, and knowledge, including participants’ own ideas about how improvements could be made.

A qualitative intrinsic case study was conducted. What makes this study intrinsic (Stake, R. (1995). The Art of Case Study Research, Sage Publications, Thousand Oaks, California) is that it provides a better and deeper understanding of the experiences of the participants accessing foreign language interpretation services in Wales. A qualitative intrinsic case study is the appropriate approach, as it ensures that various lenses are used to generate reliable empirical evidence, focusing on very specific groups of people (Baxter, P. and Jack, S. (2008). Qualitative Case Study Methodology: Study Design and Implementation for Novice Researchers. The Qualitative Report, 13, (4), 544-556). Intrinsic case studies do not exemplify the evidence, nor do they seek to generate theory, but to gain insight into unique phenomena and groups (Eisenhardt, K. M., Graebner, M. E. (2007). Theory building from cases: Opportunities and challenges. The Academy of Management Journal, 50, (1), 25-32).

To ensure the study’s reliability, (Yin, R. K. (2003). Case study research: Design and methods (3rd ed.). Thousand Oaks, CA: Sage.) protocol was used, which posits that there was a need to be prepared for unexpected findings during data collection. Anticipating unexpected findings allowed them to be dealt with constructively during data analysis, avoiding gaps. According to the same protocol, researcher bias was minimised by performing reliability checks throughout the data collection. Data was collected over the course of two months, using different methods, to avoid misinterpretation and strengthen the study’s reliability (Denzin, N. K. (1970). The Research Act in Sociology. Chicago: Aldine.)

This qualitative case study’s external validity is challenging, as the participants are very specific and “seldom heard” and the policy environment is ever-changing, thus, the findings are not applicable to other situations or populations. Stake (1995) suggests that even though qualitative intrinsic case studies are unique, sufficient contextual and background information is provided to ensure that the issues at hand are well understood. Triangulation was engaged by cross-referencing information that the participants shared with, for example, evidence in UK policy documents that was identified through information gathering. By scrutinising their viewpoints and experiences and verifying them against trustworthy sources, it paints an enriched picture of their experiences (Van Maanen, 1983).

The sampling framework was determined by the sampling population and sampling size. One of the biggest barriers was the lack of organisations that engage exclusively with forced migrants.

In comparison to the rest of the UK, Wales has a significantly smaller number of third sector organisations, service providers, and stakeholders that engage exclusively with forced migrants and who are comfortable discussing their processes and policies.

Also, unlike England, Wales does not consistently collect data on the numbers of forced migrants. This is a conscious decision from the Welsh Government to safeguard such marginalised population. This decision contributed to recruitment issues.

In order to enhance the report’s generalisability and paint an accurate picture for Wales, an All-Wales approach was taken, and the sampling technique was customised based on each population’s criteria. Thus, resulting in 2 distinct sample groups: the forced migrants and various other stakeholders.

All of the interviews with forced migrants were face to face, while the interviews with other cohorts were a mixture of face to face, telephone and video conferencing.

To achieve generalisability, forced migrants from across Wales were recruited. Thirteen forced migrants were interviewed with a balanced ratio across the Equality and Diversity characteristics (gender, race, sexual orientation, age, disability, faith, legal status in the UK). The participants came from the Middle East, North Africa, Indian sub-continent and Sub-Saharan Africa and they belonged to 8 different ethnic minority groups. They were all recruited through a number of organisations, using a combination of extreme case and snowball sampling methods. All participants could speak, understand, and write English to various degrees. Participants were aged between 21 and 50+ years old. The majority were aged 25 to 35. Half of the participants were asylum seekers and half were refugees. They all identified as Black, Asian and Minority Ethnic; a number were LGBTQ+. A third of the participants were disabled, and more than half identified as Muslim. The gender balance was 69% male and 31% female thus almost mirroring the general gender balance of forced migrants. 

The participants were very positive in regard to the interview process. They all said that it was the very first time someone discussed with them their experiences with foreign language interpretation services. They thought that hearing their views was a very positive move towards creating a more inclusive service and they welcomed the opportunity to help to address the various issues.

Ten third sector organisations and people in community groups were approached to contribute to this study with 8 agreeing to take part. Those with a track record of direct engagement with minority ethnic communities and forced migrant communities were approached. The aim was to tap into their knowledge of foreign language interpretation services, the challenges they have faced over the years and the access barriers. Through engagement work, 1 to 1 discussions took place with 8 organisations based across Wales. For this part of the study, purposive sampling was used, as there are only a certain number of such organisations in Wales (Guarte, J.M. and Barrios, E.B., 2006. Estimation under purposive sampling. Communications in Statistics-Simulation and Computation, 35(2), pp.277-284.). The feedback on the interviews was positive, as third sector and support bodies representatives mentioned that it was the very first time someone discussed their challenges with foreign language interpretation services. They thought it was a positive move towards more inclusive and cost-effective services.

Of the 6 contacted providers, 2 fully and 1 partially took part, while the remaining 3 did not engage with this exercise. From the discussions held with the 3 who declined to take part, commercial interests and a belief that this study was not relevant to them appeared to be the main motivation. 

Informal interpreters were very keen to take part with all 5 readily agreeing. Their keenness seemed to come from a position of seeing the value of this opportunity to explain to the Welsh Government what problems they are having as an informal interpreter.

For some of this cohort, their ability to participate within this research has been heavily restricted by the current Ukrainian refugee crisis. A total of 11 people were interviewed, grouped as follows:

• National Health Service: 3
• Local authorities: 5
• Department for Work and Pension: 2
• Crown Prosecution Service: 1

Whilst the cohort figures were less than anticipated, the researchers are confident that the issues and themes emerging from the data is an accurate snapshot which gives a broad understanding of the issues experienced by commissioning agents in Wales.

Grounded theory was used for data analysis to explain social processes around foreign language interpretation and using tools, such as intensive one-to-one interviews and memos, all key ingredients for data collection and analysis (Robrecht, LC. (1995). Grounded theory: evolving methods. Qualitative Health Res, 19, (5)169–77). Grounded theory allows for in depth investigation into experiences that are unique or complex. As such, qualitative interviews using a reflexive model are seen as collaboration between participant and researcher (Ryan, L., Golden, A., (2006), Tick the box please: A reflexive approach to doing qualitative social research, Sociology, 40, (6), 1191-1200).

For the 1 to1 interview questions, based on key themes, using open-ended questions to encourage researchers to keep interviews informal and conversational. This allowed participants to discuss their experiences (personal/organizational) in depth. Semi-structured interviews were separated into three distinct sections that are represented by a different set of questions: initial open-ended questions, intermediate questions and ending questions. A fourth category was added containing probing questions, which were used as a guideline to eliminate intrusiveness. Each of the 3 sections explores a different aspect of the participants’ experiences. The initial open-ended questions serve a double purpose, allowing the participant to give their account. The intermediate questions allow for exploration of themes and ideas but also address any new concepts that the participant may have brought up in the first section. The ending questions were more follow-up questions for clarification purposes to ensure that their point was understood. Closing questions were used to make sure the interview exit is respectful to the participant.

Coding strategy

Converting data into codes is one of grounded theory’s most significant analytic approaches. Codes were created through data extrication to start making analytical sense. Grounded theory coding fragments the data and, thus, connections between codes emerge that lead to theory construction (Charmaz, K. (2012). The Power and Potential of Grounded Theory. Medical Sociology Online, 6, (6), 2-15). Three grounded theory coding phases were adopted, starting with initial line-by-line coding, which is recommended when examining empirical problems in interviews, offering the opportunity to take a closer look at the participant narrative (Glaser, B. (1978). Theoretical sensitivity Advances in the methodology of grounded theory. Sociology Press, Mill Valley). Focused coding then provided a clear theoretical direction, with the purpose of making sense of the large data as it directs the analysis down more conceptual paths (Charmaz, K. (2002). Grounded theory analysis, Handbook of interview research, 675–694. Thousand Oaks, CA: Sage. Charmaz, K. (2003). Grounded theory, Qualitative psychology: A practical guide to research methods, pp. 81–110. London: Sage). Finally, theoretical coding as means of onceptualization, as it is a process through which the relationships between codes are explored and get integrated into theory and abstract analysis (Glaser, B. (2005). The grounded theory perspective 111: Theoretical coding. Mill Valley, CA: Sociology Press). Grounded theory coding ensures that the collected data is approached, understood and analysed with care, eliminating any nuances. Each phase of coding adds a different layer of data interaction that feeds into the next, thus, creating an analytic and theoretical chain (Star, S. L. (1989). Regions of the mind: Brain research and the quest for scientific certainty. Stanford, CA: Stanford University Press.).

There were several cohorts who could not or would not take part. However, the researchers are confident that this report does offer an accurate snapshot of how the service operates in Wales. The researchers would like to thank the Welsh Government project team for always being supportive and informative.