Although I argue that the interviews are best suited to my research, I acknowledged that ethnography could also be an appropriate method to use. According to Jerolmack and Khan (2014), ethnography has the ability to best explain social actions, as they are observed in their everyday contexts, which is their natural setting. However, a practical issue, namely language barrier, vastly influenced the method choice. Observing family communications without understanding them would not make sense and excluding non-English speaking families would massively limit the scope of my research.

Regarding the sampling procedure, I applied the sampling for range technique because it guarantees the involvement of a certain number of people from identified subcategories of the studied group (Small, 2009). It was really important to have students from different countries of origin in the sample because I believed that different cultural backgrounds might account for some patterns in family communicating habits. Gender also played an important role in sampling because gender differences could also be the root of dissimilarities in contacting habits. Therefore, I wanted to have both male and female participants. Although sampling for the range was the most efficient for the goals of the research, my sample suffered from problems of representativeness and bias (Small, 2009). Moreover, all the students I interviewed were postgraduates, which potentially leads to sample bias, creating a limitation of the research.

My position as an insider in the researched topic enabled me to easily access participants. I used my social networks to recruit potential interviewees. I asked university peers, either online or personally, whether they would like to take part in my research. Haraway (1988) argues that no objective knowledge production exists, therefore to decrease subjectivity paid specific attention not to involve close friends in the study. I was afraid that my background information about the interviewees and their families would influence the interview – by asking or not asking about particular issues – and the research outcome. Accordingly, the ideal participants were those, whose families I have never heard about previously. When contacting the potential interviewees, I briefly introduced the project and made