EAP Gender FEBRUARY 2018 Integrating Biographical Research into Focus Group Discussions Lessons Learned from a Qualitative Study of Constraints Underlying Gender Disparities in Mongolia’s Labor Market A Knowledge Note from the East Asia and Pacific Umbrella Facility for Gender Equality Photo: World Bank While quantitative studies are useful in documenting gender disparities in G ender inequalities in access to economic opportunity, earnings, and pro- ductivity persist across the East Asia and Pacific region, as emphasized in Toward Gender Equality in East Asia and the Pacific: A Companion Report to the economic opportunity, earnings, and World Development Report (World Bank 2012). Labor-market disparities are likely to reduce macroeconomic efficiency; limit the role of jobs as direct instruments productivity, qualitative research can of women’s development and empowerment; and limit the positive spillovers of help to uncover the constraints that jobs on women’s agency, control, and power. underlie those disparities. A particularly Several quantitative studies have analyzed and documented gender disparities in Mongolia, but the constraints underlying them have long remained largely unclear, promising methodological approach providing little guidance for program design. Moreover, scant empirical evidence taken in Mongolia combines elements exists regarding the type of support that people generally expect from labor-mar- ket policies or how effective support is in addressing gender-specific constraints of biographical research with focus in Mongolia and elsewhere. The qualitative tool developed and launched for the group discussions in the context of report on which this note is based, Perceptions of Precariousness-A Qualitative Study of Constraints Underlying Gender Disparities in Mongolia’s Labor Market reconstructive research. Once success- (Schmillen and Weimann-Sandig 2017), can help fill these knowledge gaps. The fully implemented, this combination of tool addresses three related questions: qualitative methods can serve as a posi- • What gender disparities can be identified or confirmed through qualitative research? tive example of the central premises of • What are the reasons underlying the gender disparities? qualitative research: openness, engage- • Are current government policies, including active labor market policies, effec- ment with different spheres of society, tive in addressing gender disparities? If not, what can be done to improve and the impartiality of the researcher. them? The qualitative approach confirms the range of gender dis- representativeness plays no role in a qualitative research parities in Mongolia’s labor market that have been identified design, the sample must exhibit the appropriate contrasts. through quantitative research, painting a picture of wide- It is also necessary to use an empirically appropriate spread precarious employment, particularly in the private sec- method (Glaser and Strauss 1967). The qualitative research tor. Among the reasons believed to underlie gender disparities approach to understanding social environments is usually in the labor market are: (i) norms, such as the prevailing views quite distinct from quantitative research; as a result, the on the roles of men and women with respect to marriage, two approaches also differ in what constitutes an empiri- household and family-care duties, and suitable career choices cally appropriate method. Qualitative research, rather than and jobs; (ii) deficiencies in the political environment; and (iii) relying on established and preformed role models, values, the near-total absence of government support services and and norms, stresses the idea that social reality is created programs. Three contributing causes are considered salient. by the actions and reciprocal interactions of individuals, as • A prevailing societal expectation of women is that they well as by interpretation of these interactions. Social order is devote the majority of their adult lives to supporting therefore not inherent but rather constantly built and rebuilt their husbands and raising their children. by the interpretations of interactive persons. Blumer (1973) called this paradigm symbolic interactionism. • Some women manage to combine fulfilling societal expectations with maintaining successful careers, but Based on the theoretical approach of symbolic interaction- many others perceive the quality and quantity of child- ism, the qualitative tool developed and launched for the care facilities as inadequate to make this possible. Perceptions of Precariousness report combines elements of biographical research and focus group discussions for the • Differential legal treatment of men and women appears purpose of reconstructive research (Bohnsack 2014). One cen- to cement the acceptance of traditional gender roles. tral assumption of biographical research is that the actions and Policies and programs designed to address some of these decisions of individuals can be scientifically understood only underlying constraints are widely regarded as ineffective, at when viewed within the context of their life course and social least in their current form, at addressing gender disparities. context. The researcher must therefore create a survey situa- tion that enables people to explain what is most important to Methodological Approach them-rather than to the interviewer. The researcher must also consider an appropriate evaluation strategy. Whereas quantitative research deals with statistical param- eters and aims to achieve a representative and random Unlike content analysis, which seeks to analyze and sys- sample, qualitative research relies on a very different set tematically interpret what has been said explicitly during an of assumptions. The objective is not to test standardized interview or focus group discussion, reconstructive research hypotheses but rather to reconstruct typical cases by iden- not only identifies what has been said but also why and tifying contrasting and common issues among varied experi- how it was said. Biographic narratives are seen as signifying ences and perceptions. This approach is possible only if there more than a simple stringing together of actions and events. actually are contrasting cases in the sample; therefore, the Instead, a distinction is made between the real chronological composition of an appropriate sample according to observ- order of events and actions in a person’s life and the mean- able attributes is of paramount importance. While statistical ing of those events and actions to the person; this leads to 2 shifting the prioritization of their life experiences. In other words, a distinction is made between lived lifetimes and nar- Implementation The Mongolia Center for Development Studies (MCDS), a rated lifetimes (Rosenthal 2002). Biographical research usu- professional survey research firm, prepared, implemented, ally assumes that the distinction begins with an adolescent’s and documented 22 distinct focus group discussions for this secondary socialization, that is, the period of learning what study, each with a minimum of six and a maximum of nine is appropriate behavior outside the home, which follows a participants. For each discussion, there was a moderator to child’s primary socialization, mainly influenced by immedi- lead it and a secretary to assist the moderator and to record ate family and close friends (Tillmann 1989). comments word for word (electronic recordings were also The use of focus group discussions is among the most promi- made). The role of the moderator is to guide the discus- nent qualitative social research methods, with a long tradition sion while disturbing it as little as possible (Weimann-Sandig dating back to the 1930s. The first discussions were developed 2014).2 Data quality was the top priority before, during, and in the United States by Kurt Lewin and his students, based on after the fieldwork. assumptions borrowed from quantitative research. The survey research firm was provided survey instruments Purely qualitative focus group discussions, like those used for every focus group discussion, including: (i) detailed the- here, originated in Germany. Their development was deeply matic guidelines to lead the discussion; (ii) a one-page form influenced by the work of Friedrich Pollock (1955) and Wer- on which the moderator and secretary could summarize their ner Mangold (1973). Important contributions include Bohn- perspectives of the discussion; and (iii) a one-page question- sack’s (2014) refinement of the methodological approach of naire for participants, mainly to gather basic sociodemo- Mangold (1973) through the use of focus group discussions as graphic information. All material gathered at a focus group a method to determine the collective consensus of a group, discussion was fully transcribed immediately following the the premise being that they offer insights into both individ- session, before being translated into English and systemati- ual opinions and collective knowledge that are uncovered cally coded using MAXQDA software. during the focus group discussion. Moderators and secretaries attended a one-day training In this way, focus group discussions are an ideal approach workshop prior to the focus group discussions with the fol- to analyzing the values and standards of certain groups and lowing key objectives: (i) reach a common understanding of comparing them with a society’s common values and stan- the expectations of a focus group discussion; (ii) provide a dards. A group in social research is defined by common fea- detailed explanation of the roles of the moderator and sec- tures or interests. These common features or interests do not retary; (iii) discuss how to encourage participants to speak necessarily lead all member of the group to pursue the same out openly; and (iv) provide detailed instruction on how to actions, but they help the research to develop a deep under- acquire participants, guide discussions, and prepare tran- standing of why persons in the group are behaving in a par- scripts and other materials. ticular way. In other words, the common features or interests MCDS also prepared the Field Survey Guidance Man- create a sense of community-and not only in closed groups ual, which was used during the training session and the in which members know one another well or meet regularly.1 2. Focus group discussions differ from group interviews, which involve 1. Mangold (1973) and Weimann-Sandig (2014) show that even strangers can bilateral questioning and answering; the discussions involve a moderator identify with each other as long as they are members of the same group who stimulates discussion but avoids influencing participants as much as and group-specific topics are discussed. possible. 3 subsequent fieldwork. Most focus group discussions lasted The following excerpt from a biographical interview with a between two and three hours. Thirteen were conducted Mongolian woman in her mid-30s illustrates how emphasiz- in Ulaanbaatar. Nine took place in aimags (provinces other ing individual narratives and distinguishing between narrated than Ulaanbaatar); of these, seven took place in aimag cen- and lived lifetimes can offer deep insight into a person’s life: ters (provincial capitals) and two in soum centers (county After finishing secondary school, I decided to go to univer- seats). The 13 sessions in Ulaanbaatar were spread out over sity and study English at [a university]. I wanted to become an the city’s four different districts (Sukhbaatar, Bayangol, English teacher. Everything that had to do with the western Songinokhairkhan, and Nalaikh); those in the aimag cen- world and Europe fascinated me because in my childhood, ters and soum centers occurred in three different aimags the Soviet and socialist paradigm had dominated everything. (Khovd, Dornod, and Dundgobi). The districts and aimags I was raised during the transition from a socialist country to a were selected for their varying socioeconomic and regional democratic country, so we had no contacts with other coun- characteristics. For example, the three selected aimags rep- tries except Russia and China. So I chose to learn the English resent Mongolia’s three main regions. language because it was a key: I would be able to communi- Based on a few exploratory biographical interviews and cate with the other world and would have access to infor- some expert interviews conducted by experienced quali- mation about the other world, and that is the reason why tative researchers before the focus group discussions, it I chose English; another reason was that I wanted to travel was determined that the composition of the focus groups abroad. Well, and I have to say that at this time the qual- should vary depending on six attributes: (i) gender, (ii) age, (iii) ity of English teaching at [the university] was not too good educational background, (iv) employment status, (v) marital and also the quality of the teachers differed a lot. Okay, I status, and (vi) locality. Additional restrictions were placed understand that success depends on personal behavior, on some discussions, such as limiting participation to busi- but university was not as enjoyable and interesting as I had ness owners or human resource professionals of small or imagined. So, in fact, I didn´t end up working as an English medium enterprises, the disabled, or individuals who had teacher. From the present perspective, I should have done experienced internal migration. The precise composition of better because as a teacher I would have left Mongolia. But I the 22 focus group discussions, in terms of the participants’ didn´t work as an English teacher. I started my career […] and observable characteristics, is provided in table 1. then I passed the civil service exam [in the mid-2000s] and I was offered a job at [a ministry]. […] I started working for [an office at the ministry] and came into contact with people Illustrative examples from countries like Korea, Japan, and China. I really liked the At the outset, our hypothesis was that the decisions Mon- work there because I got a good impression of Mongolian golian women make over the course of their lifetimes are development. I worked there for three years […]. After three influenced by individual actions as well as by cultural, social, years of work, the elections came, and in accordance with and political circumstances. To explore this idea, elements tradition, the whole staff within the ministries was changed. of biographical research were added to the focus group dis- So I lost my job in [the late 2000s]. Fortunately, I got hired cussions. Ultimately, the combination of focus group discus- [by an NGO]. While I was working there, I organized [a big sions and biographical research was successful because both meeting]. But my NGO salary was not enough so I had to share one thing in common: treating individual narratives change my job. I applied for a job in the mining industry. In as key to understanding the social world (Przyborski 2004). that time, in [the late 2000s] mining was a booming industry 4 Table 1. Composition of focus groups Group Gender Age range Education Employment status Marital status Locality Other 1 Female 20–30 Tertiary Mixed Mixed Sukhbaatar district – 2 Female 20–30 Tertiary Mixed Mixed Khovd aimag – 3 Male 20–30 Tertiary Mixed Mixed Sukhbaatar dirstrict – 4 Male 20–30 Tertiary Mixed Mixed Dornod aimag – 5 Female 30–50 Secondary Inactive Mixed Bayangol district – (minimum) 6 Female 30–50 Secondary Inactive Mixed Dundgobi aimag – (minimum) 7 Female 30–50 Mixed Unemployed Mixed Songino–khairkhan – district 8 Female 30–50 Mixed Unemployed Mixed Nalaikh district Beneficiary of training program or other active labor–market policy 9 Female 30–50 Mixed Unemployed Mixed Dornod aimag – 10 Male 30–50 Mixed Unemployed, Mixed Nalaikh district – spouse employed 11 Male 30–50 Mixed Unemployed, Mixed Khovd aimag – spouse employed 12 Female 30–50 Mixed Employed Single mother Nalaikh district – 13 Female 30–50 Mixed Employed Single mother Dundgobi aimag – 14 Female 30–50 Mixed Inactive Single mother Bayangol district – 15 Female 30–50 Mixed Inactive Single mother Dornod aimag – 16 Female 20–50 Mixed Mixed Mixed Songino–khairkhan Disabled district 17 Female 50–60 Mixed Inactive Mixed Bayangol district – 18 Mixed Mixed Mixed Manager Mixed Sukhbaatar district From small or medium enterprise 19 Mixed Mixed Mixed Manager Mixed Songino–khairkhan From large firm district 20 Female Mixed Mixed Herder Mixed Khovd aimag – (Erdeneburen soum) 21 Female Mixed Mixed Herder Mixed Dundgobi aimag – (Erdenedalai soum) 22 Female Mixed Mixed Mixed Mixed Songino–khairkhan Internal migrant district 5 in Mongolia. So I chose to work for [a private company in any possibilities to work there due to me having two young the mining sector]. I worked there and then after a while children. So I worked at the job for six days. I don’t have any my workload was very high […] and my bosses really set me chance to work constantly, and I can only do part-time jobs. under pressure so I had a breakdown. I had to go to hospital – Participant 7.7 and then they fired me. Most young unemployed women have young children. In the biographical interview, the woman describes her life Because the companies don’t permit them to take leave, their from her present perspective, which is marked by grief. She husbands get annoyed with them and complain that they don’t is critical of her past behavior, such as her performance in pay enough attention to the family. Consequently, they pre- college, and her past decisions, especially not becoming an fer part-time jobs—but there aren’t sufficient part-time jobs. English teacher and leaving Mongolia. She sees her motives – Participant 7.1 for studying English as being related to her social circum- stances: the socialist period meant reduced rights and fewer Lessons Learned possibilities; learning English was synonymous with striving Methodologically, the main lesson learned is how to success- for western standards and the western way of life. fully combine elements of biographical research and focus The biographical interviews and focus group discussions group discussions to answer the research questions posed demonstrate that the entire career of a Mongolian woman at the outset of the investigation. Criticism of approaches is highly influenced by biographical disruptions—stages of that integrate biographical elements into qualitative stud- life when an individual takes time to pause and reflect on ies is often focused on the argument that stimulating self- her life and decide if she wants to continue on her path or generated biographical narratives without unduly influencing make an immediate change (Schoyerer and Weimann-Sandig or disturbing the discussants in a concrete interview situ- 2015). Decisions often depend on social networks and the ation or discussion is very difficult because, in many cases, availability of political and economic resources, according to biographical research is relatively foreign to interviewers and our focus group discussion participants. For example, many interviewees, and it is never clear, a priori, clear if an inter- Mongolian women drop out of the labor market owing to viewee is willing to tell their entire life story. According to social practices and mindsets and because of their employ- this view, interviewees associate qualitative research with a er’s attitude toward working mothers. Following are exam- kind of question-and-answer game and are insecure about ples of two unemployed women, aged between 30 and 50, the type of narrative expected of them. discussing the balance between work and family life: Given these concerns, extraordinary diligence was devoted I started work in the dumpling making industry while my to our sampling of discussants and interviewees for field- children went to kindergarten. […] I thought I could work work in Mongolia. Not every person is likely to narrate freely, there because I had chances to take children to kindergarten because doing so requires communication and social skills at 8:00 a.m. and would be able to pick up them at 5:30 p.m. (Spöhring 1989). Our fieldwork reveals that a combination during working hours. However, I worked with a team, and I of open questions and narrative stimuli substantially affects couldn’t leave the team to do the work in order to pick up the willingness of a focus group participant to offer deep children. Other members of the team would be angry and say insights from their individual biographies. We therefore “Are you the only one who has children?” Thus, I didn’t have began every focus group discussion with an ice-breaking 6 question worded in a very open way that did not directly relate to the topic at hand from the participants’ point of Conclusions Our methodological approach aimed to show the advan- view. An example: tages of combining elements of biographical research with Each one of us has an idea for our life. Can you tell us what focus group discussions in the context of reconstructive your ideal life plans were like when you were a teenager research. Reconstructing the social world through biographi- or young adult-something about your dreams, plans, and cal research means reconstructing people’s lives, because wishes in terms of your professional and private affairs, as people’s actions and interactions are what create the social well as what influenced them? world in the first place. Reconstruction also fosters a gen- eral understanding of social norms and values. Combining These ice-breaking exercises liberated participants from biographical research and focus group discussions was pos- chronological narratives, allowing them to concentrate on sible because both are based on the same theoretical foun- their own prioritization of events. A connection was made dations (symbolic interactionism) and the same narrative between life plans and the important biographical phase of methodology. being a teenager or young adult because of the relevance of that phase for every focus group discussion participant. The synergy thus obtained enabled us, in the case of Mongo- Most of the women who had grown up in the socialist lia, to gain knowledge on the constraints underlying gender period recalled their lives mostly in relation to those politi- disparities in access to economic opportunity, earnings, and cal and social circumstances, but many younger women nar- productivity and to launch a qualitative tool that can con- rated the stories of their lives in relation to a political and tribute to filling similar knowledge gaps in other contexts. economic system under transition. Eliciting narratives through qualitative research is a complex Another crucial lesson learned is that the qualitative approach task that requires an experienced research team and a gen- was successful because it was a complement and not a sub- eral openness toward appropriate research design and pro- stitute for quantitative research on Mongolia’s labor market. cess. Unlike structured interviews, this approach does not For the most part, this quantitative research has relied on rely on detailed survey instruments for individual interview- labor-force survey data to document gender disparities in ees. Instead, only thematic guidelines are available, which the labor market, including educational attainment, field of need to be as open and flexible as possible to capture the study, labor-force participation, unemployment, average participants’ narratives. It is therefore necessary that the earnings, and occupation- and sector-based segregation. In interviewers have sufficient expertise on the topic under other countries without a recent or high-quality labor-force study and that the research designers have sufficient exper- survey, some of the relevant quantitative information could tise in selecting participants for focus group discussions and potentially be collected from a measurement of living stan- moderators who will be able to lead focus group discussions dards or other household survey. without imposing preconceived notions. Once success- fully implemented, this combination of qualitative methods can serve as a positive example of the central premises of qualitative research: openness, engagement with different spheres of society (legal, economic, familial, etc.), and impar- tiality of the researcher (Lamnek 2016). 7 References Schmillen, Achim, and Nina Weimann-Sandig. 2017. Perceptions of Pre- cariousness-A Qualitative Study of Constraints Underlying Gender Blumer, Herbert. 1973. “A note on symbolic interactionism.” American Disparities in Mongolia’s Labor Market. World Bank, Washington, DC. Sociological Review 38(6): 797-798. Schoyerer, Gabriel, and Nina Weimann-Sandig, 2015. Tagespflegepersonen Bohnsack, Ralf. 2014. Rekonstruktive Sozialforschung. Einführung in quali- in tätigkeitsbegleitender ErzieherInnenausbildung: Berufsmotivation, tative Methoden (9th edition). Opladen, Germany: Verlag Barbara Alltagsmanagement und öffentliche Förderung. Munich. Budrich. Spöhring, Walter. 1989. Qualitative Sozialforschung. Stuttgart: Teubner. Glaser, Barney G., and Anselm L. Strauss. 1967. The Discovery of Grounded Tillmann, K. 1989. Sozialisationstheorien: Eine Einführung in den Zusam- Theory: Strategies for Qualitative Research. Chicago: Aldine. menhang von Gesellschaft, Institution und Subjektverdung. Reinbek: Lamnek, Siegfried. 2016. Qualitative Sozialforschung. (6th rev. ed.). Basel: Rowohlt. Beltz Weinheim Verlag. Weimann-Sandig, Nina. 2014. Betriebliche Mitbestimmung im Spannungs- Mangold, Werner. 1973. “Gruppendiskussionen.” In Handbuch der feld zwischen Liberalisierung und effektiver Interessensvertretung: empirischen Sozialforschung, ed. Ralf König (pp. 228-59). Stuttgart: Aktivierungspotentiale und Grenzen einer Neugestaltung betrie- Ferdinand Enke Verlag. blicher Mitbestimmung in Stadtwerken. Hamburg: Soziologische Pollock, Friedrich. 1955. Gruppenexperiment: Ein Studienbericht. Frankfurt: Fallstudien. Europäische Verlags-Anstalt. World Bank. 2012. Toward Gender Equality in East Asia and the Pacific: Przyborski, Aglaja. 2004. Gesprächsanalyse und dokumentarische A Companion Report to the World Development Report. World Methode: Auswertung von Gesprächen, Gruppendiskussionen und Bank: Washington, DC. http://documents.worldbank.org/curated/ anderen Diskursen. Wiesbaden: VS Verlag für Sozialwissenschaften. en/270511468246373406/pdf/734380PUB0Box30ward0gender0equal Rosenthal, Gabriele. 2002. “Biographische Forschung.” In Qualitative ity.pdf. Gesundheits- und Pflegeforschung, ed. Doris Schaeffler and Gabriele Müller-Mundt (pp. 221-32). Bern. Acknowledgments This note was prepared by the World Bank’s Social Protection and Jobs Global Practice (GSPDR). It was written by Achim Schmillen and Nina Weimann-Sandig, with guidance from Philip O’Keefe and inputs from Helle Buchhave of the World Bank’s Social, Urban, Rural, and Resilience Global Practice (GSURR). The note is published under the EAP Gender Capacity Project led by the EAP Social Development Team of the Bank. Its objective is to promote learning among clients, partners, and task teams about strategic regional priorities. For more information, please contact Ludy Anducta, Program Assistant, GSURR, at landucta@worldbank.org. 8