6 Symbolic Interaction Volume 36,Number 1,2013 on their marriage timing and marriage chances(Wong 2003,2005;Ravanera and Rajulton 2007;Ono 2003;Korabik 1993;Ng and Ng 2009).These guiding questions included: (1)Do the career women want to get married in the first place? (2)If so,then what were the factors that hindered them from realizing their marriage goals? a.Putting career over marriage/family life?(voluntary) b.Difficulty in finding partners?(involuntary) (3)What type of partners were they looking for? With each stage of sampling and coding,new questions were added onto the interview schedule according to the evolving theoretical needs of the study(Charmaz 2006).For instance,in stage two of interviewing,questions related to parental expectations were added onto the schedule due to the prevalence of such a theme in the pilot or first stage of interviewing.The filial-related questions that were added were: (4)What criteria do your parents look for in your marriage partner? (5)Have they ever introduced any partners to you? (6)Have there ever been any disputes regarding your differences in partner choice criteria？ Sampling Procedures and Concept Formation At the outset,it was determined that all of the informants in the sample would have university-level education or above and work as professionals or managers in global firms in Shanghai.This was in keeping with the leading characteristics of sheng nu or"leftover women"who were hypothesized to have difficulties finding marriage partners because of their superior economic qualifications (Chinese Ministry of Education 2007).The age group and the marital statuses of the informants were not determined until after the pilot stage of interviewing.This was in keeping with the openness and fluidity of the grounded theory method,which encouraged concepts to emerge from the data with each stage of iterative sampling and coding.Besides the sampling criteria of"ethnicity"(Chinese),"educational level"(university or above),and"professional status"(professionals,managers,senior level executives, and entrepreneurs),the rest of the sampling criteria was left to be relatively open to ensure that it was guided by the empirical data,in keeping with the rationale of grounded theory,which was to let the hypotheses and concepts emerge and be induced from the data,instead of predetermining hypotheses in the onset and using the data to verify them like in quantitative approaches
6 Symbolic Interaction Volume 36, Number 1, 2013 on their marriage timing and marriage chances (Wong 2003, 2005; Ravanera and Rajulton 2007; Ono 2003; Korabik 1993; Ng and Ng 2009). These guiding questions included: (1) Do the career women want to get married in the first place? (2) If so, then what were the factors that hindered them from realizing their marriage goals? a. Putting career over marriage/family life? (voluntary) b. Difficulty in finding partners? (involuntary) (3) What type of partners were they looking for? With each stage of sampling and coding, new questions were added onto the interview schedule according to the evolving theoretical needs of the study (Charmaz 2006). For instance, in stage two of interviewing, questions related to parental expectations were added onto the schedule due to the prevalence of such a theme in the pilot or first stage of interviewing. The filial-related questions that were added were: (4) What criteria do your parents look for in your marriage partner? (5) Have they ever introduced any partners to you? (6) Have there ever been any disputes regarding your differences in partner choice criteria? Sampling Procedures and Concept Formation At the outset, it was determined that all of the informants in the sample would have university-level education or above and work as professionals or managers in global firms in Shanghai. This was in keeping with the leading characteristics ofsheng nu or ‘‘leftover women’’ who were hypothesized to have difficulties finding marriage partners because of their superior economic qualifications (Chinese Ministry of Education 2007). The age group and the marital statuses of the informants were not determined until after the pilot stage of interviewing. This was in keeping with the openness and fluidity of the grounded theory method, which encouraged concepts to emerge from the data with each stage of iterative sampling and coding. Besides the sampling criteria of ‘‘ethnicity’’ (Chinese), ‘‘educational level’’ (university or above), and ‘‘professional status’’ (professionals, managers, senior level executives, and entrepreneurs), the rest of the sampling criteria was left to be relatively open to ensure that it was guided by the empirical data, in keeping with the rationale of grounded theory, which was to let the hypotheses and concepts emerge and be induced from the data, instead of predetermining hypotheses in the onset and using the data to verify them like in quantitative approaches
Understanding Sheng Nu("Leftover Women') 7 The pilot stage of sampling was conducted to"obtain an overview of the overall process"'and to "determine the dimensions and boundaries ..of the project" (Morse 2007:235).In the pilot stage,I interviewed 20 women and men(with university education and working as professionals and managers)who were in the age range of 27-59,and who were single,in a relationship,married,divorced,and separated.After coding the interview transcripts using ATLAS.ti qualitative data analysis software, and analyzing the data manually for concepts,I decided that men's views did not contribute much to women's views on marriage and partner choice,and hence they were omitted from the sample.Other criteria that were refined after this initial stage were the age range and marital statuses of the women,as it was found that those who were never-married,and in the age range of 28-33 had more relevant and interesting relationship experiences to share,and hence the married,divorced,separated,and older informants were omitted.The second stage of"purposeful sampling"thus required sourcing out informants with this new set of criteria in accordance with the analyses of the first stage.In each stage of coding the data,the technique of "constant comparison"(Strauss and Corbin 1990,1994;Corbin and Strauss 2008) was employed to compare the codes and the data against each other in order to arrive at the most significant,representational,and conceptually relevant codes that will eventually become"conceptual categories'that will be integrated into the coding paradigm.The ATLAS.ti computer software program played an important role in identifying which codes had the largest number of occurrences and hence had the largest potential to be developed into analytical concepts and conceptual categories-although not replacing the concurrent manual analyses of the data. Eventually,after the first two stages of sampling and coding,analytical concepts and conceptual categories had been developed.The two main conceptual categories included"parents'emphasis on the economic status of women's partners"and"male superior norms and patriarchal perceptions of men."The third and fourth stages of"theoretical sampling"were subsequently conducted to fill in the dimensions and properties of these conceptual categories (Charmaz 2006).In the third stage of theoretical sampling,informants who hadhad experiences of being introduced to men by their parents (known as"matchmaking")were sourced out to fill in the dimensions and properties of the conceptual category of"parents'emphasis on men's economic status"-until the point of saturation,that is,when no new knowledge about the concept can be found.Similarly,in the fourth and final stage of theoretical sampling,informants who were at the top of their professions and would be hypothesized to pose the most threats to men's accomplishments were located in order to fill in the dimensions and properties of the conceptual category of"male superior norms and patriarchal perceptions of men
Understanding Sheng Nu (‘‘Leftover Women’’) 7 The pilot stage of sampling was conducted to ‘‘obtain an overview of the overall process’’ and to ‘‘determine the dimensions and boundaries ... of the project’’ (Morse 2007:235). In the pilot stage, I interviewed 20 women and men (with university education and working as professionals and managers) who were in the age range of 27–59, and who were single, in a relationship, married, divorced, and separated. After coding the interview transcripts using ATLAS.ti qualitative data analysis software, and analyzing the data manually for concepts, I decided that men’s views did not contribute much to women’s views on marriage and partner choice, and hence they were omitted from the sample. Other criteria that were refined after this initial stage were the age range and marital statuses of the women, as it was found that those who were never-married, and in the age range of 28–33 had more relevant and interesting relationship experiences to share, and hence the married, divorced, separated, and older informants were omitted. The second stage of ‘‘purposeful sampling’’ thus required sourcing out informants with this new set of criteria in accordance with the analyses of the first stage. In each stage of coding the data, the technique of ‘‘constant comparison’’ (Strauss and Corbin 1990, 1994; Corbin and Strauss 2008) was employed to compare the codes and the data against each other in order to arrive at the most significant, representational, and conceptually relevant codes that will eventually become ‘‘conceptual categories’’ that will be integrated into the coding paradigm. The ATLAS.ti computer software program played an important role in identifying which codes had the largest number of occurrences and hence had the largest potential to be developed into analytical concepts and conceptual categories—although not replacing the concurrent manual analyses of the data. Eventually, after the first two stages of sampling and coding, analytical concepts and conceptual categories had been developed. The two main conceptual categories included ‘‘parents’ emphasis on the economic status of women’s partners’’ and ‘‘male superior norms and patriarchal perceptions of men.’’ The third and fourth stages of ‘‘theoretical sampling’’ were subsequently conducted to fill in the dimensions and properties of these conceptual categories (Charmaz 2006). In the third stage of theoretical sampling, informants who had had experiences of being introduced to men by their parents (known as ‘‘matchmaking’’) were sourced out to fill in the dimensions and properties of the conceptual category of ‘‘parents’ emphasis on men’s economic status’’—until the point of saturation, that is, when no new knowledge about the concept can be found. Similarly, in the fourth and final stage of theoretical sampling, informants who were at the top of their professions and would be hypothesized to pose the most threats to men’s accomplishments were located in order to fill in the dimensions and properties of the conceptual category of ‘‘male superior norms and patriarchal perceptions of men.’’