Thesis Title: Subtitle

Table 1: Questionnaire - Age

Please select your age range.
Answer Options	Response Percent	Response Count
18-24	42.6%	66
25-30	26.5%	41
31-40	18.7%	29
41-50	6.5%	10
51-60	4.5%	7
61-70	1.3%	2
70+	0.6%	1
answered question		155

Table 2: Questionnaire - Gender

Please indicate your gender
Answer Options	Response Percent	Response Count
Woman	80.0%	124
Man	20.0%	31
answered question		155

Table 3: Questionnaire - Educational Attainment

Please select the highest level of education you have completed from the list below.
Answer Options	Response Percent	Response Count
Some highschool/secondary school	4.5%	7
Completed high school/secondary school	21.3%	33
Diploma/ associate diploma/ trade certificate/apprenticeship	14.2%	22
Graduate Diploma and Graduate Certificate	4.5%	7
Bachelors degree	34.2%	53
Postgraduate qualification	21.3%	33
answered question		155

Table 4: Interview Participant Demographics

Name	Gender	Age	Country	Interview Method	Marital Status	Educational Attainment
Natalie	F	25	Australia	In person	Married	Bachelors
Matthew	M	28	Hungary	Instant message	Single	Post-graduate
Sage	F	31	Australia	In person	In relationship	Bachelors

Carol	F	23	Germany/Hungary	Instant message	Single	Bachelors
Bird	F	29	Netherlands	Instant Message	Single	Post-graduate
Sally	F	24	Australia	In person	Single	Post-graduate
Camilla	F	30	Australia/Brazil	In person	Single	Post-gradate
Eva	F	25	Australia	In person	In relationship	Post-graduate
Irene	F	31	Australia/Norway	In person	Single	Post-graduate
Zoe	F	21	Australia	In person	In relationship	Bachelors
Marie	F	22	Germany	Instant message	In relationships	Bachelors
Paula	F	31	Australia	In person	Single	Bachelors
Mark	M	30	Australia	In person	In relationship	Post-graduate
Brendon	M	27	Australia	In person	In relationship	Bachelors
Chloe	F	32	England	Phone	Single	Bachelors
Kathryn	F	38	England	Instant message	Not disclosed	Bachelors
Joseph	M	26	Australia	In person	In relationship	High school
Candace	F	31	Slovenia	Instant message	Single	Bachelors
Andrea	F	37	Australia/Canada	In person	Single	Bachelors
Kate	F	48	Australia	In person	Married	Bachelors
Madeleine	F	24	Germany	Instant message	In relationship	Bachelors
Chloe	F	33	England	Instant message	Single	High school
Ethan	M	26	United States	Instant message	Single	High school
Martha	F	65	Australia	Phone	Married	Bachelors
Lisa	F	41	Australia	Phone	Single	Post-graduate

Mahesh	M	31	Australia	In person	In relationship	Post-graduate
Lara	F	37	Portugal	Phone	In relationship	Bachelors
Todd	M	28	Australia	In person	In relationship	Bachelors
Mandy	F	28	Australia	Instant message	Married	High school
Joshua	M	30	United States	Instant Message	Single	High school

Chain Referral Sampling: Techniques and Limitations
The sample above was generated through chain-referral sampling also sometimes referred to as a snowball sample. Before proceeding a distinction needs to be drawn between chain-referral sampling and respondent-driven sampling (RDS). While respondent-driven sampling is a form of chain-referral sampling, there are a few key differences. As with chain-referral sampling, RDS asks participants to recruit their peers into the study. These recruits are then asked to recruit their peers, and so on until the target sample size is reached (Paquette, Bryant and de Wit 2011). However, RDS is concerned with creating statistically significant samples in a way that chain-referral sampling is not (Kogan et al. 2011).
Chain-referral sampling means that research is conducted in a networked fashion and is a series of referrals that are made within a circle of people who know each other. Most commonly, chain referral sampling is used to access hidden populations and has been used in research examining drugs users (Paquette et al. 2011) and LGBT identities (Browne 2005), or groups at risk of contracting HIV (human immunodeficiency virus) (Heckathron et al. 1999). Chain-referral or respondent-driven sampling is useful for sampling populations that for various social, legal or moral reasons have very low visibility, thus presenting problems for locating and contacting respondents. This chain referral sampling in these instances is used out of necessity. For this research, chain-referral sampling helped maintain the characteristics of Facebook – that is, its networked nature – which is essential in creating a case study that deals appropriately with Facebook as a structuring object.
Chain-referral sampling is also an appropriate method when considering the size of the population in which I am interested (Facebook users). Facebook as an object of study is too diffuse and diverse to construct a rigorous purposive sample. However, creating a case study enables a deep examination of my network and allows the insight gleaned from this case to be theoretically inferred to other networks or individuals that have similar characteristics. For a project that examines Facebook, taking an approach that highlights the networked nature of Facebook was necessary. As will be discussed later, Facebook does have a tendency towards homophily given the emphasis of Facebook’s architecture on recreating offline social connections online. To control for these factors meant losing some of the characteristic of Facebook as a technology. Furthermore, Coleman (1958) argues that chain-referral sampling is a method uniquely appropriate to sociological research given its ability to sample natural interactional units (such as those found on Facebook). As such, this is not a study of Facebook as a case, but of a network on Facebook, which can be understood as a loosely configured interactional network centred on myself.

The use of chain-referral sampling in social science research is not without critique. Some of the criticisms levelled at chain referral sampling are: bias in the choice of initial subjects; differential recruitment by participants; differences in network sizes among participants; and a tendency towards homophily (Kogan et al. 2011). Homophily is a concept used to describe the homogeneity of social networks based on the idea that individuals are more likely to form and sustain social relationships with those who share similar characteristics (McPherson et al. 2011). Homophily occurs on both individual, intrapersonal and sociodemographic levels (McPherson et al. 2011). Homophilous networks are most strikingly visible on levels such as ethnicity, gender and occupational status (McPherson et al. 2011). The primary criticism of chain-referral sampling is that is that is not random and selects individuals on the basis of the social networks (Biernacki and Waldorf, 1981; Baxter and Eyles 1997; Faugier and Sargeant, 1997). The goal of this research is to engage in a sustained way with the structural logic of Facebook, which is the social network, in order to better understand some of the forces behind some of the patterns observed.