3.3. Research process onion
3.3.7. Research process, techniques, and procedures
3.3.7.1. Description of participants
125 3.3.5.4. The methodological approach in qualitative research
Qualitative research incorporates a range of methods and practices with some common characteristics. Qualitative research is inductive and develops theories (Coyle & Tickoo, 2007;
Mohajan, 2018). Qualitative research methods are flexible, and analysis is not dependent on statistical procedures (Antwi & Hamza, 2015; Creswell & Poth, 2016). Qualitative research considers the researcher as a research instrument (Anderson, 2017).
Qualitative research includes various methods, namely case studies; ethnography;
phenomenology; action research; and ground theory (Mohajan, 2018). The case study method emerged as the most appropriate method. It aligned with the study's purpose of developing a social performance measurement framework for small and medium social enterprises in KwaZulu-Natal.
The case study approach provided in-depth and rich descriptions, explorations, and understandings of social performance measurement of small and medium social enterprises.
126 participant responded that their organization focus on delivering community services in the education of skill and training sector in SA. However, there is no data or information specifically about SMSEs delivering community services in the skills and training sectors in SA, specifically in KZN. Therefore, it was impossible to determine the industry's size. The researchers purposively focused on SMSEs identified and registered for the Local Economic Development’s (LED) programme for UKZN by the eThekwini Municipality’s LED department. Purposive sampling allowed the researcher to choose cases that exemplified the features or procedures the research is concerned about (Chimhundu, 2018). This type of sampling suits a qualitative case study because it allows the researchers to pursue categories where “ the process being studied is most likely to occur”(Denzin & Lincoln, 1998;202). Selecting cases purposively makes the research “a piece of information- a rich case study to explore” specified research issues (Saunders et al., 2009:142).
The number of cases in multiple case studies varies in the literature. It ranges from ideas that only provide criteria without signifying a specific number (Lincoln & Guba, 1985; Patton, 1990) to those that suggest the approximate number of cases (Miles & Huberman, 1994; Yin, 2003). Based on those authors, the recommended case range falls between two and four cases as the minimum, and ten to fifteen cases as the maximum (Perry, 1998). Furthermore, according to (Perry, 1998:19),
“the rigorously analytical method of case study research [is] usually based on many interviews within four to fourteen cases”, following the interview protocol. Additionally, Eisenhardt (1989:545) highlighted that, “with fewer than four cases, it is often difficult to generate theory with much complexity, and its empirical ground is likely to be unconvincing.” Initially, 54 SMSEs that undertook the Champion Programme of the LED of UKZN from 2015 to 2019 were approached. Factors such as nature and scope of the study (Morse, 2015), purpose and research question of the study (Mason, 2010; Patton, 2015), methodology chosen and type of data collected (Creswell, 2016), and data saturation (Guest, 2006; Guest, 2020; Hennik & Kaiser, 2022) determine the sampling plan and size. Based on the recommended range and the inclusion and exclusion criteria set for this study (Chapter 1) and purpose of the study, out of 54 SMSEs, 10 were selected. It is crucial to choose people or sites that help understand the central phenomenon (Creswell, 2013). A small sample size in qualitative research allows the researcher to be focused on the in-depth understanding in a particular social and cultural context (Subedi, 2021:6).
Following this, the founders/managers/executive directors of selected SMSES were invited by email to participate in the study, beginning in August 2019. Ten SMSEs were contacted and agreed
127 to participate in follow-up emails and by phone. The number of participant (the size and composition of the sample) in qualitative research depends on the problem under study (Patton,2015; Subedi, 2021). In qualitative study, the researcher needs, and judgment play a major role in constituting how many samples are required and who can participate as a sample in particular study (Creswel1, 2016; Patton, 2015; Subedi, 2021). Therefore, Stakeholders are not included in the study as a sample because the researcher beliefs relevant information needs, their involvement and influence of the stakeholders was briefly elaborated by the SMSEs. The sample was homogenous in terms of the enterprises' scale and the way of performing the measurement.
Creswell (2013:155) stated: “it is essential that all participants have [similar lived] experience of the phenomenon being studied”. However, the selected SMSEs perform various community services and have different legal forms of registration, based on South African business registration. Prior thematic saturation and the inductive thematic saturation point was reached after seven SMSEs.
Saturation refers to “the point in data collection when no additional issues or insights are identified and data begin to repeat so that further data collection is redundant, significantly that an adequate sample size is obtained” (Hennink & Kaiser, 2022: 116). In qualitative research, saturation has become a important component that helps make data collection robust and valid. Furthermoore, saturation is “the most frequently touted guarantee of qualitative rigor offered by authors to reviewers and readers" (Morse, 2015:587). Guest et al. (2006:60) refer to the saturation as ‘the gold standard by which purposive sample sizes are determined in health science research. Different scholars refer to saturation as a ‘point’( Hennink & Kaiser, 2022), as a ‘rule’(Denny 2009; Sparkes et al. 2011), or an ‘edict’ (Morse 1995), of qualitative research, and it features in a number of generic quality criteria for qualitative methods (Leininger 1994; Morse et al. 2002).
Following this, this study adopted a simple method to assess and report saturation in qualitative research developed by Guest et al., (2020) (see Figure 3.2). The method consists of three elements:
the base size, the run length, and the relative amount of incoming new information, or the new information threshold (Guest et al., 2020). This method can be applied to either a prior thematic saturation (where it relates to the degree to which identified codes or themes are exemplified in the data in sampling or inductive thematic saturation (where it relates to the emergence of new codes or themes in the data analysis) (Saunders et al., 2018). The base size refers to how the research define the body of information already identified in the dataset to consequently use as a
128 denominator (Guest et al., 2020). Base size defines the minimum number of data collection event (interviews) researchers should review/analyse to calculate the amount of information already obtained (Guest et al., 2020). The base sizes used as a denominator in the saturation ratio. The run length is the number of interviews within which the researcher looks for and calculate new information (Guest et.al., 2020). The number of new themes found in the run defines the numerator in the saturation ratio. According to Guest et al., (2020) new information threshold is a point where we accept as a new information is obtained. It is subjective decision for the researcher in determining whether the collected information is new or similar to the existing information.
Figure 3.2: Summary of process, base line and run length options (Guest et al., 2020)
Previous studies in qualitative research (Morgan, 2002; Guest et al., 2006; Guest et al., 2017) indicated that in qualitative dataset, new information is generated early and generally follows an asymptotic curve, with a relatively sharp decline in new information occurring after just a small number of data collection/analysis events (Guest et al., 2020:1899). Depending on this reason, this study chosen to test 2, 3 and 4 interviews as base sizes from which to calculate the total number of unique themes to be used in the denominator of the saturation ratio. Based on the conclusion of Guest et al., (2020) the unit of analysis for base size is the data collection event; the items of analysis are unique codes representing themes.
129
Base line 2 data collection
events
3 data collection events
4 data collection events
Run length 2 data collection
events
3 data collection events
N data collection events
New information threshold
< n% new information
< 5% new information
No new information
Level of confidence saturation reached
Figure 3.3: Saturation assessment parameters and level of confidence saturation reached.
Figure 3.3 indicate the base line, run length and new information threshold parameters used to obtain the saturation ration with the level of confidence saturation reached being < 5% of new information. The steps used to calculate the saturation ration is explained in the next section.
Interview number 2 3 4
New themes per interview 28 8 4
Number of base themes 40
Step 1: Find the number of unique themes for the base: Start by looking at the firs three interviews conducted and summing the number of unique themes identified within this group. The resulting sum, 40, is the denominator in our equation.
Interview number 4 5
New themes per interview 4 2
New themes in run 6
STEP 2—Find the number of unique themes for the first run. The study is using a run length of two, so include data for the next two interviews after the base set–i.e., interviews 4 and 5.
After reviewing those interviews, the study identified four new themes in interview 4 and three new themes in interview 5. The number of new themes in this first run is six.
Step 3: Calculate the saturation ratio
Saturation ratio = the number of new themes in run/ number of base themes = 6/40
130 = 15%
The quotation indicates 15% new information, which is not < 5%, thus the process will continue.
STEP 4 –Find the number of new unique themes for the next run in the series. For the next run the study add the new themes for the next two interviews, 5 and 6 (note the overlap of interview 5), resulting in a sum of three.
Interview number 5 6
New themes per interview 2 1
New themes in run 3
STEP 5—Update saturation ratio. the number of new themes in the latest run
(three) and divide by the number of themes in the base set (40). This renders a quotient of 7%, still not below our <5% threshold. The process will continue to the next run.
STEP 6 –Find the number of new unique themes for the next run in the series. For this third run add the number of new themes identified within interviews 6 and 7.
Interview number 6 7
New themes per interview 1 0
New themes in run 1
STEP 7—Update saturation ratio. the number of new themes in the latest run (one) divided by the number of themes in the base set (40) equals.
Updated Saturation ratio = 1/40
= 3%. At this point the proportion of new information added by the last run is below <5% threshold was established, so the data collection stop here after the 7th interview is conducted and the amount of new information is minimizing to the level where it can be concluded as saturation has been reached.