Appendix 1: Participants’ Characteristics Participants in Baroda, India Name Sex Age Econ Religion/ Alcohol/ Living Father/Mother Own profession Rlp status class* caste^ Smoke arrangement profession Nilesh M 25–30 M Patel NO With in-laws Business person/HM Family business LM (1 yr) Jambli F 20–24 M Brahman NO With in-laws Accountant/HM Web design LM (1 yr) Aditya M 25–30 M Gujarati Jain YES/NO Joint Business person/HM Family business AM (3 yrs) Geet F 25–30 M Kutchi Jain YES/NO Joint Business person/HM Homemaker AM (3 yrs) 164 Hiren M 25–30 UM Khadayata NO With in-laws Business person/HM Family business AM (2 yrs) Swati F 25–30 UM Patel NO With in-laws Business person/ Travel agency AM (2 yrs) Fashion designer Lena F 25–30 M Brahman NO Nuclear University Prof/ Dance teacher LM (5 yrs) Teacher – HM Priya F 20–24 M Patel NO Nuclear Shopkeeper/Teacher Student AM (6 mths) Tarun M 25–30 M Patel NO Nuclear Teacher/HM Pharmacist AM (3 mths) Krishna M 20–24 LMi Sindi NO Nuclear Purchase Officer/HM Student S (Father) Durish M 20–24 M Lohana YES/NO Joint family Business person/HM Student R (4 mths) Toni M 25–30 UM Leva Patil YES Nuclear family Business person/HM Family business S Rahul M 20–24 UM Jain YES/NO Nuclear family Business person/HM Family business S Nirali F 20–24 M Leva Patil NO Joint family Civil Servant/Clerk Student nurse S Muktha F 20–24 M Tailor NO Hostel Tailor/Tailor – HM Student S Kareena F 25–30 UM Jain NO Flat Doctor/Doctor Bank S Rekha F 20–24 LMi Muslim NO Nuclear Teacher/HM Student S Ismaili family Seeta F 20–24 LMi Leva Patil NO PGH/Flat Blue collar Unemployed/ R (1 yr) worker/HM Student *Economic class based on my own judgement, includes type of house (bungalow, flat, area), cars/motorbike and travel abroad M = Middle UM = Upper middle LMi = Lower middle. ^As defined by participants AM = Arranged marriage, LM = love marriage S = Single, R = In relationship (as reported by participants) HM = Homemaker. Participants in London, UK Name Age Sex East Econ Religion/ Ethnic Living Father/Mother Own Relationship African Class* Caste^ area Arrangement profession profession Status Roots Lona 25–30 F NO M Patel White With spouse Grocery shop Finance LM (1 year) owners Sohan 25–30 M NO M Patel White With spouse Business Finance LM (1 year) Person/ Administrative assistant Ameera 25–30 F YES UM Muslim Asian/ With spouse Entrepreneurs Medical LM (1 year) Gujarati doctor 165 (continued ) Continued 166 Name Age Sex East Econ Religion/ Ethnic Living Father/Mother Own Relationship African Class* Caste^ area Arrangement profession profession Status Roots Mahendra 25–30 M YES UM Vaishya White With spouse Finance/ Business LM (1 year) Homemaker Darsha 25–30 F NO M Potter Asian/ With fiancé Factory owner/ Solicitor Engaged Gujarati Homemaker Pretak 25–30 M NO M Sudra Asian/ With fiancée Unknown IT Engaged Mochi Gujarati Prity 20–24 F YES M Lohana Asian/ With parents Engineer/ Journalist Relationship Gujarati (nuclear) Homemaker (5 years) Renu 20–24 F NO M Jain Asian/ With partner Journalist/ PR Cohabiting (3 Gujarati Teacher years) Rama 20–24 F YES M Brahman White With parents Bank Manager/ Adminis- Single (nuclear) Homemaker trative assistant Naveen 25–30 M YES LMi Sudra Mixed With parents Teacher/ Student Single Mochi ethnic (nuclear) Homemaker area Nihal 20–24 M NO LMi Leva Patel Mixed With friends IT/Homemaker Lawyer Relationship ethnic (1 year) area Yogesh 20–24 M YES M Lohana Asian/ Alone Dentist/ IT Single Gujarati Homemaker *Economic Class based on my own judgement, includes house and area grew up in (e.g. wealthy area, or council estate), and parents occupation. UM = Upper middle, M = Middle, LMi = Lower middle. AM = Arranged marriage, LM = Love marriage, R = In relationship, S = Single (as reported by participants) HM = Homemaker. ^As defined by participants. Appendix 2: Data Analysis Procedures The analysis was conducted using the NVIVO computer program, version eight (QSR International 2008). This section explains the data analysis procedures I went through. It is ordered sequentially into separate ‘steps’ of analysis but in fact the analysis went through a much more chaotic back and forth process. Step one – immersion in the data After each interview I listened through the recording at least once, sometimes making notes of follow- up questions for a later interview. Then I transcribed the interviews, or had a professional transcriber take a first attempt and then ‘fix it’ the way I wanted it, including non- verbal queues and the context of the interview from my field notes. Conventions in transcription notation were taken from Silverman (2001) and from theatre play scripts. After the interviews were tran- scribed I then read through them again with the intention to become immersed in the data through reading and rereading (Barrett 1996; Silverman 2001). At this stage I also took notes on emerging themes or ideas sparked by the transcripts and field notes. Step two – coding There were two main phases to coding; in the initial phase I coded each line in a sub- sample of interviews and field notes. The sub- sample included interviews of two male participants in India and two in the UK, and of two female participants in India and two in the UK (total eight participants), and field notes from the first month in each site. The codes emerged from the data itself, rather than from a previously devised frame of codes. Line by line coding is used to capture the ‘essence’ of each statement; the code should capture the ‘meaning or action in the line, it is the first step in interpretation’ (Charmaz 2006: 45). In the second phase I brought together all the codes that had emerged during this initial stage and subsumed them under focused codes. Focused codes are ‘more directed, selective and conceptual’ than the initial line by line codes (Charmaz 2006: 57). While the line by line codes often focused on actions or intentions within a line, the focused code captured larger sections of data in a slightly more abstract way. These focused codes were then used to code all the data – while still allowing new codes to emerge. Step three – memo writing, categories and concepts The coding is interspersed with memo writing – that is reflective writing on the emergence of important themes and codes. In memo writing ‘you stop and 167 168 Appendix 2: Data Analysis Procedures analyse your ideas about the codes in any – and every – way that occurs to you during the moment’ (Charmaz 2006:73; see also Glaser 1998). These memos should help in the abstraction of codes and ideas about the data. They also help in identifying gaps in the data or analysis, or in pointing the way towards the most salient concepts which emerge from the data. Memos also facilitate the constant comparison for which grounded theory is so well known: data from different individuals and in different contexts are compared and questions are asked of emerging theories or concepts, constantly refining and reworking the understanding of the data (Charmaz 2006). The data from India and the UK were analysed in one ‘file’ but following the grounded theory methodology, I con- stantly compared the data and concepts from the two contexts, trying to under- stand how Gujaratis in the UK were similar or different to those in India, and to speculate about why this might be. This comparative element to the analysis was especially helpful in crystallising the emerging concepts. To some degree it helped ‘make strange’ some views, especially those amongst the UK participants which I often found similar to my own. These comparisons were worked over within the memos that I wrote. The memos then helped to elevate codes into categories. Categories should ‘explicate ideas, events, or processes in your data – and do so in telling words. A category may subsume common themes and patterns in several codes’ (Charmaz 2006: 91). Categories emerged from the focused codes which I felt best represented the data, or at least the story that was emerging. These often evolved from memos which became the basis of findings chapters. For example, the term ‘arranging love’ emerged from my understandings of ‘focused codes’ around courtship in Baroda. It suggests both how participants in Baroda appear to manipulate their feelings for their partner, and the integration of love with arranged marriage. Categories were then raised to concepts; in interpretive grounded theory, theo- retical concepts enable an understanding of the relationships between the cat- egories. A concept subsumes categories and has ‘analytical weight’, it should help you to understand the connections between the categories and bring the data together into a complete story (Charmaz 2006). The process is one of increas- ing abstraction, but grounded in the data collected. The concepts integrate the whole book; each findings chapter leads into one another. They are linked both thematically and theoretically. Appendix 3: Participants’ Ranking Participants were asked to rank in order of importance the most important qualities in a spouse, with 1 representing the most important. Ranking by Baroda participants 12345678910 Nilesh Virginity Kind Family Education Broad- Good- Intelligent Settled in Cooking Family background minded looking job Ability Wealth Aditya Education Intelligent Kind Broad- Good- Family Cooking Family Settled Virginity 169 minded looking back- ability wealth in job ground Tarun Virginity Family Education Good- Kind Cooking Broad- Intelligent Family Settled background looking ability minded wealth in job Toni Education Intelligent