• No results found

2.3 RESEARCH DESIGN

2.3.4 Research methods

2.3.4.1 Sampling

A sample is a sub-unit of a population or characteristics of the population selected to estimate the behaviour being researched (Babbie & Mouton, 2001; Mason, 2007). According to Babbie (2004) there are four different units of analysis frequently used by the research community, namely, individuals, groups, organisations and social artefacts. The primary unit of analysis for this study is workplace bullying from HR practitioners’ perspectives, hence they constituted the primary target population. HR practitioner’s perceptions, viewpoints, attitudes, feelings and experiences of workplace bullying constituted insightful and enriching data sources congruent to the set research objective.

My intention was to use a combination of non-probability sampling techniques, namely, purposive and snowball sampling. The latter “yields a study sample through referrals made among people who share or know of others who possess some characteristics that are of research interest” (Lindloff & Taylor, 2002, p. 124). Ultimately, however, I used only purposive sampling as participants were unwilling to refer me to any person(s) involved in bullying situation(s) due to the sensitivity of the phenomenon. According to Babbie (in Branch et al., 2007a, p. 109), purposive sampling “is suitable when it is not appropriate or feasible to interview all members of the population group.” This method enables the selection of participants based on the researcher’s judgement about the suitability of the participant to the purpose of the study (Davoudi et al., 2016). It is chosen because, a) it places emphasis on in-depth understanding and analysis of the research phenomenon (Patton, 2002); b) it does not draw on large or random samples, in accordance with the grounded theory (Terre Blanche & Durrheim, 2004); c) the sampling process is guided by the ongoing theory development (Gorra, 2007); and d) it has elements of theoretical sampling, that is, desirable participants are selected based on criteria specified by the researcher and the purpose of the study to build and expand on the substantive theory (Babbie, 2004; Henning, 2004; Silverman, 2005), either upfront or progressively (Miles & Huberman, 1994). Thus, qualitative purposive sampling is congruent with a grounded theory methodology, especially since it resonates the principles of theoretical sampling, which is a core grounded theory strategy. I explained what theoretical sampling means and also described how theoretical sampling was applied in this study in section 2.3.4.4 (iii).

Nine participants were interviewed to gain their perspectives, insight and understanding of workplace bullying in two tertiary institutions in South Africa. Application letters requesting permission to conduct the study were sent to both institutions for approval. Both institutions’ management and appropriate College Ethics Committees granted their permission in writing to conduct the study. In one institution, a list of HR practitioners’ contact details (names, levels, telephone numbers and email addresses) was obtained from the office of the HR Executive Director. I then recruited them through telephone and emails, explaining the research intention and requesting a meeting. In the other institution, HR practitioners were known to me and I approached them individually, either telephonically or in person, to ask if they would take part in the study. I then sent them a formal invitation4 via email. Nine HR practitioners in total, from the two institutions, agreed to be interviewed.

Eligibility to participate as HR practitioners in this study was guided by pre-determined purposive sampling criteria, known as exclusion or inclusion criteria which represent the defining characteristics specified by the researcher that identify suitable participants to be included in the study (Meline, 2006; Omona, 2013; West, Gartlehner & Mansfield, 2010). The exclusion or inclusion criteria could include factors such as employment status, position or organisation tenure, the participants’ work roles and history, desired understanding of phenomenon under study, age, gender and race(Van Spall, 2007). To be eligible for participation in this study, the HR practitioners had to be permanently employed in an HR position, with a minimum of three years of HR experience and performing general HR responsibilities. They were selected on the basis of relevance and acceptability (Robey & Dalebout, 1998). Table 2.1 shows the profile of research participants.

22 Table 2.1

Profile of research participants (n=9)

ID Gender Race Job title /

Position Organisational tenure Qualification Professional body / Membership

TS1 Male Black Team

Leader 8

Honours

Degree SABPP

NK2 Male White Manager 15 Masters

Degree SABPP

AD3 Female White Manager 18 Doctoral

Degree SABPP

TO4 Female Black Team

Leader 12

B.Tech

Degree SABPP

ET5 Female Black Practitioner 6 Honours

Degree SABPP

PI6 Female Coloured Practitioner 3 Diploma SABPP

LU7 Male Black Practitioner 18 Diploma IRASA

NM8 Male Black Team

Leader 20 Degree IRASA

ES9 Female Indian Practitioner 5 Degree SABPP

The sample of participants interviewed was fairly balanced between males (four) and females (five). Five participants were Blacks, followed by two Whites, one Coloured and one Indian. All the participants had some form of educational qualification, ranging from a Diploma to a Doctoral degree. Two of the nine participants reported were managers of the sub-units within the functions listed above. Three were team leaders reporting to HR managers and the remaining four were practitioners reporting to team leaders and/or managers. The majority of participants had been in HR positions for 11-20 years, two reported three to five years, two reported six to 10 years, five reported between 11 and 20 years. Most of the participants were members of the South African Board of People Practice (SABPP) as they were required by their institutions to register with HR-related professional bodies. The only exception to being registered with the SABPP were two practitioners who were members of the Industrial Relations Association of South Africa (IRASA).

With regard to purposive sampling within grounded theory, some qualitative researchers reiterated that researchers must have a smaller number of interviews rather than increasing the sample size to avoid the problem of running out of time and failing to analyse the data content appropriately (Bonde, 2013; Charmaz, 2014; Guest, Bunce & Johnson, 2006; Mason, 2010; Onwuegbuzie & Leech, 2007). According to Bonde (2013) in reference to Romney, Weller and Batchelder (1986), to obtain a satisfactory sample size, participants must be homogenous and possess expertise, experience or knowledge in relation to the phenomenon under inquiry. In the literature, some studies posit that homogenous participants with both common knowledge and experience about the research topic may lead to adequate sampling and data saturation occurring earlier in interviews (Bryman, 2012; Charmaz, 2014, 2012; Mason, 2010; Newman, 1957). On the contrary, Bonde (2013, p. 2) argues that participants “could also be made up of people with highly dissimilar characteristics (e.g., race, age, gender and so forth) who think and behave quite differently in relation to phenomenon of interest”.