Practical Example

The study below was conducted over four years in two urban Aboriginal health centres, located in two inner cities in Canada. Browne et al. (2016) studied the health status inequities affecting indigenous populations and provided an evidence-based framework and specific strategies to promote health care. The researchers conducted over 900 hours of intensive participant observation focused on staff members’ interactions and patterns of relating with patients and other staff during clinical encounters and in waiting rooms, staff meetings and case- management discussions (Browne et al., 2016).

Practical example

Brotsky and Giles (2007) used a controversial method of data collection in the study, “Inside the ‘Pro-Ana' Community: A Covert Online Participant Observation”. Many health professionals have become alarmed at the emergence of the pro-ana community; pro-ana referring to the promotion of behaviors related to the eating disorder anorexia nervosa (Brotsky & Giles, 2007). Typical pro-ana communities are interactive in the form of bulletin boards or message forums, where users post information in non-synchronous fashion, organized into “threads” by site owners. (Brotsky & Giles, 2007).

Brotsky and Giles (2007) acknowledged the evident ethical considerations brought on by covert participation observation. They argue that deception is necessary to gain meaningful access to pro-ana sites as it felt highly unlikely that access would be granted to a researcher openly disclosing the purpose of the study due to the high degree of defensiveness within pro-ana communities (p. 96). “In this instance, was deception justified? We believe that it was, given the charges laid against the pro-ana community (that they are effectively sanctioning self-starvation), and the potential benefit of our findings to the eating disorders clinical field” (Brotsky & Giles, 2007, p.96).

Research Steps

1. Problem formulation: Brotsky and Giles (2007) focused on three research questions:

• What benefits do pro-ana site users derive from their membership of the community?
• What beliefs do pro-ana site users hold about eating disorders, and how do these shape their attitudes towards “recovery” (and therefore towards the health and medical professions more generally)?
• Do the beliefs of the pro-ana community have a consistent, shared viewpoint? (p. 97).

2. Identify and become familiar with suitable online communities: After an initial study of pro-ana interactive virtual communities, Brotsky and Giles (2007) narrowed the field to a total of 12 separate communities. According to Akturan (1983), the traffic in the community should be substantial and the number of postings should be high to ensure rich and detailed data.

3. Accessing the community: Brotsky and Giles (2007) entered a password protected pro-ana community by assuming the identity of a fellow eating disordered site user. The study followed the stages as outlined:

a) Induction phase: Researcher constructs a plausible persona that enables them to obtain access

b) Interaction phase: Researcher enters the community as a plausible member.

c) Exiting phase/ Post-engagement phase: Researcher reflects on experiences and collects outstanding data (p. 97-98). Sangasubana (2011) claims that exiting from virtual ethnography can be more difficult than other research methods as the researcher and the accessibility of the community continue to exist in the virtual world.

4. Data Collection: Using non-participant observation, Brotsky and Giles (2007) first spent several weeks’ analyzing the sites to identify typical characteristics, such as “cyber names,” ages, height and weight statistics, long and short-term weight goals, and other attributes common to pro-ana community members (p. 97-98). A typical pro-ana site user persona was developed:

Age: 20

Sex: Female

Height: 5'6'

Eating Disorder: Anorexia

Highest Weight: 138 lbs.

Current Weight: 104 lbs.

Lowest Weight: 97 lbs.

Short-Term Goal Weight: 100 lbs.

Long-Term Goal Weight: 90 lbs. (Brotsky & Giles, 2007, p. 98).

Using this persona and covert participation observation, Brotsky and Giles participated as a member of the various groups over a period of two months, depending on the degree of acceptance within the group and interaction by site members in general. Data included chat room transcript, instant messenger dialogues, and bulletin board exchanges (p. 96-99).

5. Data Analysis and Interpretation: Data from the Brotsky and Giles (2007) study was analyzed using a variety of qualitative methods appropriate for the different types of data. Thematic analysis was used to identify discussion topics, concerns, and by evaluating users’ contributions as positive or negative. Following extensive reading of archived discussion forum content from pro-ana sites, the authors posted statements about “ana” which were consistent with typical viewpoints expressed in the community as a whole, enabling them to gauge the responses of others (p. 99).