5.2 Methodological considerations
5.2.2 Reflexivity in research
92 Arguably, researchers do not choose research topics without theoretical or ideological preconceptions or inclinations. Social reality is being modified when researched, while researchers are not impartial observers. Thus, empirical results are not demonstrated from a ‘neutral’ position. Therefore, it is crucial to reflect on the researcher’s own positionality within the research process. This reflects the fact that researchers ‘position themselves’ which means they are inevitably affected by their background (e.g. experience; views; culture; history) which in consequence influences the research they undertake (Yuval-Davis, 1997; Creswell, 2013). “Like the researched or participant, the researcher is a product of his or her society’s social structures and institutions” (Hesse-Biber, 2014, p. 200); hence, no research is ‘value free’ and, to maximise transparency, the researcher should regularly reflect on their social ‘position’ and what effect it may have on the studied topic. Thus, reflexivity is pivotal in any research (Gawlewicz, 2014). This has been applied and reflected upon in this research as the following discussion suggests.
Whilst I am a Polish native speaker and a migrant to the UK, this fact did not make me an unproblematic ‘cultural insider’ to all of the participants. As a Polish migrant woman and a doctoral researcher at the same time I have had to negotiate these roles throughout the course of the research. It could be argued that in this study I was a “familiar stranger” (Agarwal et al., 2009). This is to symbolise the fact that I was somewhat similar (with regard to certain characteristics, e.g. nationality, gender, migration status, sexual orientation, often also age, class, marital status), thus in some ways an ‘insider’ and ‘familiar’ to the research participants. At the same time I was a ‘stranger’ in the sense that I took the role of a researcher gathering information on fellow migrant women who could have had very different experiences (e.g. in Poland/the UK or in migration experience). I recognised the research participants as experts on their own lives who took part in the interview as a “co-construction” of meaning (Hesse-Biber, 2014, p. 199). I, as both an expert and an insider, aimed to compare the accounts of the ‘inside’ and the ‘outside’ (i.e. contrast the participants’ views with what is known from academic literature). It can be asserted that particularly because I am part of the Polish migrant community in the north of England, I had little trouble finding other Polish migrant women who were willing to participate in the research, despite the fact that this community is often referred to as ‘hard to reach’ (Vickers, Craig & Atkin, 2013). However, it should be noted that in almost any interview situation the researcher will be both an insider and an outsider at the same time (Hesse-Biber, 2014). While the former is built on commonalities, the latter makes any differences explicit and in any interview situation the two sides are likely to be similar in some respects and different in others. This is due to various ‘axes of difference’ intersecting one
93 another (cf. ‘intersectionality’, Crenshaw, 1989; see section 3.3.4). It is likely that the researcher will always differ in some respects to the research participants whilst perhaps s/he will also have some commonalities (Ahmed, 2012). Arguably, an interview constitutes an opportunity to explore these and continuously renegotiate them throughout the research process.
From the onset, I was open about the roles I played and my background. At the beginning of each interview, I introduced myself giving the research participants some background to my migration history (i.e. sharing my own biography). Arguably, this exercise can contribute to minimising distance, reducing the hierarchy and extending rapport between the research partners. It also emphasises certain shared characteristics (e.g. the migrant status). The interviewees were given the opportunity to choose their own pseudonyms, which can potentially increase their feelings of research ownership. Reciprocity is vital in qualitative research, particularly that of a feminist/gendered alignment (cf. Hesse-Biber, 2014). I feel strongly about the process of giving back; thus I offered to distribute any relevant research outputs to the research participants and/or the interview transcripts to those interested. Finally, with regards to the relationship between the researcher and the research participants it is important to decide whether to do research on, for or with people. In this project my aim was to do all three. However, because of the methodological choices explained above, I wished to be a reflective partner, a conscientizer with the aim to better the participants’ situation in some way (Blaikie, 2007). This, I believe, was achieved by giving them a voice and opportunity to be heard, thus, providing more balance to the existing literature on gendered lives of Polish migrant women.
I did my best to ensure the power relations of the interview interaction (Mason, 2002) were maintained at a level convenient for both parties (i.e. the interviewer and the respondent), so that no one party was able to exercise power over the other, as is expected when adhering to feminist principles. I accept the view that I am ‘fallible’ and wished to combine my knowledge and experience with that of the research participants (Hoggart, 2012). When asked personal questions that relate to my migratory experiences, I provided an honest and detailed answer. It is difficult to totally eliminate the interviewer effect (if at all possible) (cf. Denscombe, 2010); however, it is believed it was minimal given that I shared sex, ethnic origin, nationality, native language and often age range with the respondents but this cannot be known for sure. I endeavoured to be sensitive to the cultural, religious and socio-economic background of the research participants. At all times, I ensured that I was humble, moderately dressed and used everyday language. This is not to undermine the research participants but rather to acknowledge
94 that I did not wish to be seen as a person of authority exerting her dominance. I simply wanted to be seen as ‘one of them’, a Polish woman with emphasis on commonalities rather than differences. It is noteworthy that many of the interviewees must have felt at ease in my company as many used phrases such as: ‘we’,‘us’, ‘you know how it is, you’ve been through this’, etc. This perhaps suggests they felt that I was ‘one of them’ and in many cases that was true for me too. This is recognized as an expression of mutual belonging to both home and host societies. This suggests that the interviewees not only assumed a shared sense of belonging to the Polish nation but also similar migration experience (cf. Temple 2011; Gawlewicz, 2014). However, this assumed commonality of experiences cannot be taken for granted but should be reflected on and questioned (i.e. probed further). Otherwise, the research is in jeopardy of lacking in transparency as certain information can be lost due to falsely assumed commonality of experiences. In this research, there were many moments of assumed similarity and difference between myself and the respondents. On many occasions the shared origin, spoken language and experience of migration seemed to stimulate discussions and foster mutual rapport. Arguably, the fact that I emphasised my migration ‘journey’ at the beginning of each interview encouraged the interviewees to ‘open up’. At the same time, this assumed sameness sometimes appeared to act as a constraint. This was the case when the women expected me to immediately recognise certain situations or cultural contexts due to sharing the aforementioned characteristics.
It is worth noting that my researching Polish migrant women as ‘one of them’ proved to be challenging at times. On several occasions it became clear to me that I was emotionally involved in the research. Thus, I cared about the end results not only because I wished to carry out rigorous research as part of my degree and career development but because I was concerned about how Polish women are portrayed since I am one. This meant that I had to constantly reflect on my position as an insider (a Polish woman) and an outsider (a researcher) and make sure that my background did not taint the research findings. Although, this may seem like a contradiction to what I argued before with regards to researchers’ background being part of the evidence, I believe this ought to be considered with caution so as not to spoil the empirical contributions.
5.3 Qualitative methodology
A qualitative methodology was adopted in this research. It was found more appropriate than quantitative methodology as a depth of knowledge (‘quality’) rather than its breadth (‘quantity’)
95 was sought. Because gaining in-depth knowledge and understanding and an exploration of certain issues was the purpose of this study, quantitative methodology was deemed unsuitable. Quantitative methodology would not capture sufficient detail or go beyond mere description regarding specific experiences and attitudes, beliefs and values in relation to gender roles (cf. Mason, 2002). I felt that in order to address the set research questions (see next section 5.4), a qualitative methodology was suitable. This is because the ‘why’ and ‘how’ and not only ‘what’ were investigated. This qualitative study was conducted because the research problem, outlined in chapter one, needed to be explored rather than described and a detailed understanding of the issue was sought (Bazeley, 2013).
Additionally, as already noted, qualitative methods are favoured by feminist researchers (Stanley & Wise, 1990; Maynard, 1994). Moreover, I wished to “empower individuals to share their stories, [and] hear their voices” (Creswell, 2013, p. 48). Thus, a qualitative research strategy seemed to be a better fit for the chosen research problem (Silverman, 2011). Further, a qualitative methodology better matches my ontological and epistemological position and better reflects my views on positionality in research as discussed above.
5.4 Research questions
The research aims to explore and unpack the following research questions:
1. How does gender influence Polish migrant women’s lives?
2. How does migration shape Polish migrant women’s gender role expectations?
3. How do Polish migrant women manage their roles over time and space (post migration/return migration)?
4. How, if at all, do gender roles differ between the UK and Poland?
5. How do Polish migrant women experience and negotiate their gender roles in respect of work and care (‘formal’ paid work vs. ‘informal’ familial care work)?
Research objectives:
1. Examine women’s motivations for migration and return migration and motives behind their stay in the destination country.
2. Consider the extent to which migration may be a catalyst for change in Polish women’s (often traditional) perception of gender roles.
96 3. Explore how Polish migrant women negotiate their formal rights and informal
responsibilities in respect of paid work and informal familial care work.
4. Compare and contrast the differences, if any, vis-à-visgender role expectations between Polish migrant women living in the UK and those who have returned to Poland.
5. Add new empirical knowledge and contribute to theory development with regard to international migration and gender, and more specifically to Polish women migrants and gender role expectations.
5.5 Sampling strategy
Strategic purposive sampling was adopted (Mason, 2002). This sampling strategy is particularly appropriate when aiming to approach those individuals who fit the sampling criteria and who are seen as those who could potentially make a considerable contribution and have something to offer when it comes to the area under research. This sampling strategy was also seen as relevant to this study because the sampling criteria were very specific as is explained below.
The aim [of strategic purposive sampling] is to produce a relevant range of contexts or phenomena, which will enable you to make strategic and possibly cross-contextual comparisons, and hence build a well-founded argument. … in this version, then, the sample is designed to encapsulate a relevant range in relation to the wider universe, but not to represent it directly (Mason, 2002, p. 123-124).
Snowball or chain sampling was also employed to enable inclusion of various individuals to take part in the research after first making sure they were suitable and could enhance the results (Ritchie, Lewis & Elam, 2003, p. 94). I sampled across the North West of England. This part of England was chosen as the area with the second highest population density in England (ONS, 2011a). At first, the county of Greater Manchester was chosen due to its vibrant international community and its being the destination for many migrants. However, I felt that Greater Manchester may be too narrow and, as a result, limit the scope of the research. Hence, the North West area was chosen due to the established links with the migrant community. I already had excellent connections with Accession 8 migrants and organisations through my previous work in the area. Through my work at Europiaxxviii, I was very much up-to-date in respect to this community. The fieldwork in the North West was followed by a period of fieldwork in Poland, where I used my links to get in touch with returnees. Snowballing was also adopted to identify suitable individuals who fit the selection criteria to be included in the sample (Ritchie, Lewis
97 & Elam, 2003). I asked each of the UK based respondents whether they knew of anyone who had returned to Poland, and if so, whether I could be put in touch with them.
As noted above, the fieldwork was conducted in the UK and in Poland. It had been anticipated that the UK interviews would be conducted first, and then the interviews in Poland would follow. This was to enable snowballing and identifying individuals who fit the sample selection criteria and can be classed as return migrants. However, the use of snowball sampling when interviewing the UK based women to locate individuals living in Poland resulted in a wide spread of women in different parts of Poland. Given the research project timescale and resources, it would have been extremely difficult for me to travel to all those locations. Therefore, I travelled to some of those interview locations, depending on how far they were from my place of birth (Legnica – a city in Lower Silesia, south-west of Poland). I, when conducting fieldwork in Poland, was based in Legnica where I had access to free accommodation provided by my family. Those respondents who lived in the parts of Poland to which travelling would have been extremely difficult, were invited to participate in an internet interview, provided that they had the appropriate equipment and willingness to do so (see section 5.7.2.1). Snowballing proved to be useful when trying to find participants in Poland as the sample selection criteria made it relatively difficult to find potential respondents. What added to the difficulty is the fact that I permanently live in the UK. This is why internet interviews were conducted (Denscombe, 2010).
It is noteworthy that popular social networking sites such as Facebookxxix were very useful in regard to locating potential respondents. It seemed that people were happier to use those than their regular email accounts. It was also clear that those sites were not only more popular, but people accessed them more regularly. More importantly, social networking sites are more likely to be associated with friends and socialising rather than work and so the response rate to a casual invitation to take part in a research project may be higher and viewed more positively than when it was received by email. Therefore, Facebook was treated as a tool when trying to locate and recruit potential respondents. This is not to say that random individuals were approached. On the contrary, on several occasions I, after making sure I had a provisional ‘yes’ from a friend of a potential respondent, sent a private message with brief information on the research project and asked for this person’s personal email address.
98 The sample consisted of two groups of women who identify themselves to be of Polish origin. For the purpose of this research project, Polish origin is understood as being born in Poland and having Polish as the native language.
Group 1 – migrants: 16 Polish migrant women who migrated to the UK post 2004 and who continue to live in the UK (at the time of the interview, of course) (see Table 1 migrants – UK based).
Group 2 – return migrants: 16 Polish women who migrated to the UK post 2004, lived in the UK for at least 6 continuous months but who have subsequently relocated to Poland (see Table 2 return migrants – Poland based).
The number of interviews was decided during the course of the project and was determined by the point at which new data ceased to emerge from the interviews (i.e. sampling until reaching data saturation point; Ritchie, Lewis & Elam, 2003, p. 80). It is also widely acknowledged that when conducting qualitative research, the sample size is considerably smaller than when carrying out quantitative studies (Baker & Edwards, 2012).
Given the research aim and objectives (see chapter one, section 1.7), only women were invited to take part. The sample included individuals from different age ranges but of working age; who have different socio-economic characteristics; who have been living in the UK for different lengths of time (not less than 6 months); and who are of different relationship status (i.e. married/single/in relationship, with/without children). Another important criterion was that the research participants lived in the UK for a minimum of 6 months. For the purpose of this research project, 6 monthsxxx is seen as the minimal time needed for migrants to settle in their new environment and develop links with the established communities. Further, given the characteristics of the studied migrant community and the circularity of their moves, I felt that the standard 12 months would limit the scope of the sample.
In order to achieve as diverse a sample as possible, its composition was regularly reviewed and monitored (see Table 1 and 2 for further demographic details of the participants). I made sure to invite as wide a range of individuals as possible. This should enable individuals of different demographic characteristics to have their say and as a result capture a wide variety of views and constitute a balanced sample. This was done by approaching a cross-section of individuals with different characteristics (e.g. age, occupation, current work, length of time spent in the
99 UK, etc.). Although snowballing may influence the sample composition and capture similar views due to the involvement of the same range of participants (Olsen, 2012; Creswell, 2013), I regularly reflected on the sample composition to avoid this.
With regard to the sample composition, it included relatively young individuals, with the youngest participant at the age of 20 and the oldest 57, most of whom were in their 20s and 30s. The majority of the migrants (Table 1) possessed secondary level of education, whilst the return migrants (Table 2) were relatively better educated. Overall, the majority of the research participants came from cities; most of the migrants originated in smaller towns and rural areas while most of the return migrants were from larger urban areas. Many of the migrants were single with children, whilst the majority of the return migrants were in a relationship, with an equal split between those with and without children. All of the return migrants who were in a relationship had a Polish partner, whilst a couple of the migrants had British partners. The