**3.3 Target population**

O'Leary [33] defines population as the total unit of a particular class or group from which a sample is drawn. Bryman [34] describes a population as a collection of people or items considered for a research study. The study was conducted in the Northern Cape; however, participants were selected from various provincial government departments responsible for delivering public infrastructure projects. 28 participants from eight Departments (see **Table 1**) within the Northern Cape Provincial Government and responsible for the execution of public infrastructure projects in their respective departments were interviewed. They perform roles such as project managers, quantity surveyors, architects, planners, engineers, middle and senior management, and project stakeholders.

### **3.4 Sampling method**

The participants for this study were chosen using the purposive sampling technique. The researcher first identified the relevant population, which comprises the entire number of elements, to extract a sample [35]. In this case, individuals from the Northern Cape provincial government departments are responsible for delivering public infrastructure projects. A reasonable population sample had to be drawn


after identifying the target demographic. Purposeful sampling is a technique that is frequently employed in qualitative research [36]. Purposive sampling allows researchers to find and choose people or groups with exceptional expertise or experience in the phenomenon under investigation [37]. The research focuses on the infrastructure construction projects within the Northern Cape Provincial Government. Thus, project implementers and managers from various departments in the case study province were purposively selected for the interview. According to O'Leary [28], sampling is the "process of selecting elements of a population to participate in the study." The researcher invited 32 participants for an interview, of which 28 participated. According to Leedy and Ormrod [38], a qualitative study among a homogeneous population using interviews requires a sample size of 5–25. In Camngca et al.'s [39] study, 8 participants were used, and a saturation point was realised at the 7th interviewee. Again, Smith and Amoah [40] used 16 participants for their qualitative study and reached a saturation point at the 12th interviewee. A saturation point for this study was realised at the 22nd participant; thus, 28 interviewees used for the analysis are justified. According to Mayring [41], what is essential is when saturation is reached during a qualitative interview.

#### **3.5 Data collection methods**

The research study made use of a semi-structured interview guide. The researcher conducted semi-structured interviews. The semi-structured interviews were meant to get the participants' verbal, reliable, and valid information. Letters were sent to the Heads of Department offices for the various departments to get permission and set up interview appointments. The researcher visited the targeted individual's offices to conduct the interviews at an agreed convenient time. The participants were contacted telephonically to confirm their availability before the researcher visited their offices. The researcher took notes during interviews and used a tape recorder. An interview is a data collection exercise whereby the researcher asks questions and records verbal answers from the participant [42]. Where there were ambiguities and discrepancies in the participant's responses, the researcher sought clarifications to ensure consistency in the interviewees' responses to the questions asked. This also enriched the data and information provided by the interviewees. The interview took about 30–45 minutes per participant. The saturation point (where no new information emerged) was reached at the 22nd participant. The working departments of the participants interviewed are indicated in **Table 1**. In the discussion, the participants were asked to indicate their demographic data, including their position and experience in the organisation. The participants were then asked to express their views on the risk factors that usually confront them in the project execution process, whether risk management is included in the project plan, stakeholders involved in risk assessment and risk management practices in the organisation, among others. However, this study focuses on the risk factors confronting them in the project execution.

#### **3.6 Data analysis methods**

Data analysis dissects the gathered information into components to address the research questions or achieve the research goals [43]. Data editing and coding were done by repeatedly listening to the recorded participants' responses and notes taken *Risk Factors Affecting Public Infrastructure Projects DOI: http://dx.doi.org/10.5772/intechopen.112002*

during the interviews. The gathered data was examined using the content analysis method. This method allows the researcher to review transcriptions of recorded data carefully. Data gathered through the interviews were analysed using Polit and Beck's [44] qualitative contents analysis processes, as indicated in **Figure 2**. The demographic features of the participants are indicated in **Table 2**.


#### **Figure 2.**

*Qualitative contents analysis. Source: Polit and Beck ([44], p. 542).*


#### **Table 2.**

*Participants' demographic data.*

**Step 1**: The researcher repeatedly listened to each recorded interview to understand the interviewees' statements. This was done by downloading the recorded interviews into the computer. The researcher then played the interview loudly and listened attentively to have a general overview of what the participants stated.

**Step 2:** The researcher transcribed the summary of each interview in a notebook and typed it into a Word document. This also assisted the researcher in further reading the participants' narrations to understand and make the statement meaningful.

**Step 3:** The researcher carefully analysed each transcribed data to identify any differences in the responses. This was done by carefully reading and comparing the participant responses to questions asked and notes made during the interview.

**Step 4:** Where differences were identified, the researcher replayed the recordings for further clarification and referred to the notes made during the interview to assist in reconciling responses.

**Step 5**: The researcher identified common meanings from the participants' responses and entered them into an Excel spreadsheet. This was done by using keywords or themes to define participants' responses.

**Step 6:** The researcher identified similar meanings from the themes identified in each interview. This was done by comparing various themes stated by each participant to see where similarities exist.

**Step 7**: The researcher finally merged similar themes. This was then by grouping similar themes into one to reduce the volume of themes listed. Then sub-themes were identified and grouped under the main themes.

Among the 28 participants in the study's total sample, 64% identified as male and 36% as female. Again, the position compositions of participants indicate that the majority (20%) are engineers in the departments involved in the government project execution, whilst 18% are project managers and deputy directors, respectively. Also, participants' years of experience executing public infrastructure construction projects indicate that 50% have over 10 years, while 29% have 6–10 years. This shows that participants have experiences and knowledge of the risk factors causing failure in government infrastructure projects.
