Phase two: Case study interviews

In the second stage of the study a case study was conducted. Interviews are widely used by researchers in qualitative research (Barbour, 2008; Bryman, 2008). Liamputtong (2009b) viewed interviews as one-on-one or face-to-face interactions between participants and researcher, providing the interviewee’s insights on range of relevant topics. In this research study, the case study comprised seven in-depth one-to-one interviews in six companies, ranging from large, midsize and small businesses that experienced ERP implementation and adoption challenges. According to Holstein (2003), the interview approach is an empirical data collection process in which an individual is encouraged to provide views on certain aspects in detail. Williamson (2002) listed some advantages of the interview approach in qualitative research:

• Interviews enable researchers to gather in-depth insights on issues.

• Interviews enable researchers to gather direct responses from participants.

• Interviews enable open communication and exchange of information.

• One-to-one interviews have better response rates than indirect communications (e.g. via mail or email).

• Questionnaires in interviews evolve due to one-to-one contact with interviewees.

• The interviewer can control the discussion and keep interviewees focussed on the issues being discussed.

• Unstructured interviews can be more flexible and expand on important aspects as various issues are discussed.

Some interview disadvantages were highlighted by Malhotra (1999), as follows:

• The limited skills and capabilities of the interviewer can be an issue.

• Lack of structure could make the outcome vulnerable to the interviewer’s influence on interviewees. This could result in quality being compromised.

Types of interviews

There are three types of interviews that could be considered for qualitative data analysis (Williamson, 2002c). These are:

• Structured interviews are standardised and/or scheduled before their inception. All interviewees are asked the same questions in exactly the same pattern or sequence. There may be some freedom provided to interviewees while they express their opinions, thoughts or views unrelated to the strict agenda followed by the researcher.

• Unstructured interviews are non-standardised, unscheduled and in-depth interactions with interviewees. The interviewer follows the flow of discussion and every interview answer could result in a new question. This type of interview is to gain new insights from interviewees and is appropriate for case studies to collective extensive data from key individuals. This type of interview is acceptable for interpretive research.

• Semi-structured interviews are based on a standard list of questions but the interviewer could follow the lead of interviewees or ask them additional questions to seek more detailed responses. This type of interview is closer to the unstructured approach than the structured one.

For the purpose of this research study, semi-structured interviews were used for the case study data collection. A list of questions was composed. However, due to the enormity of data sought, the interviewer remained open to discussion and allowed participants to share details as they considered them necessary or relevant.

Conducting a pilot

For the purpose of this research analysis, before conducting interviews, the researcher conducted a pilot interview with the first case organisation. The organisation was the only large size business case interviewed for this study and after completing the pilot, to validate its findings and obtain a detailed view on the organisational data, a second interview was organised with another official. The pilot interview helped with both practical and methodological purposes and assisted the researcher in gaining detailed insight into the case organisation. It also helped the researcher to understand the operational details of the case organisation and to refine the interview structure and case study protocols (Perry, 1998). The pilot case interview was a convenience selection as the researcher was working within that organisation. At the time of interview, the interviewee (a colleague of the researcher) played a vital role as the Technical Resource Manager of ERP applications implementation at the organisation. Primarily due to the professional relationship the researcher had with the interviewee, the pilot case interview was productive and the researcher was able to explore a range of aspects relating to the organisation’s ERP implementation experiences. In addition, the interviewee offered access to relevant organisation documentation, as deemed necessary for the research analysis. It was understood by the interviewee that the researcher was in the early stages of establishing the instrument for interviews and was testing the model. Due to the sheer volume of data being analysed, the researcher tested the time limits and ensured that the interview was completed within an hour. The pilot interview was conducted using a semi-structured approach and the pattern of interviewing as well as the instrument remained consistent throughout the rest of the interviews.

The pilot case interview was conducted in August 2011. The interview was recorded and transcribed. In addition, the researcher made notes about the theme from data and issues relating to research design and procedures (Yin, 2003). The researcher notes were later scanned and kept in an electronic file along with the interview audio recording. The data collected from the pilot was included in the data collection and analysed with other case study interviews.

Conducting interviews

For the purpose of this study, interviews were conducted at several different locations based on the participants’ availability and convenience. It is understood from the literature that the location of an interview could have influence on the data being collected and the context of 126 | P a g e

the interview, as this is considered to be a social interaction amongst two individuals (Neuman, 2005). Furthermore, it was suggested by Neuman (2005) that the interview should be conducted at a private or quiet location, such as a home. In this study, most of the interviews were conducted at the participants’ office meeting rooms, based on the availability of each participant. In two instances, interviews were conducted in a cafeteria in Melbourne, as suggested by participants due to their tight schedules.

According to Cavana (2001), there might be some non-verbal behaviour issues that could impact on an interview. Some strategies and guidelines were suggested to counteract such behaviour, such as:

• The interview pattern: clear patterns of interview interactions are required, by having good communication with interviewees through managing language barriers (if any) and encouraging participants to freely provide their insights and in-depth information.

• Listening: the feedback provided by participants in response to questions should be carefully comprehended, clearly interpreting the essence of the comments made.

• Paraphrasing: the interviewer must precisely paraphrase the true message communicated by the interviewee.

• Probing: relevant questions should be asked to dig deeper and produce more in-depth insight on the relevant information.

All interviews were carefully planned, organised and scheduled before their inception. A list of individuals who were relevant to the area of interest was made and each individual was informally contacted and invited to participate in the study. As a result, some individuals declined and some agreed to participate. Every potential participant was requested to provide a date, time and venue for their liking to schedule an interview session. The consent information that outlined the purpose of this study and a statement that their participation would remain confidential was provided to each participant. The researcher was aware of the deficiencies of the interview process such as bias, interviewer characteristics and the effects of interviewer on the discussion. To cater for these challenges, the researcher maintained impartiality, remained focused on the topic content and professional throughout the interview process. No cost reimbursement was claimed from the researcher’s University; as all interviews were conducted in Victoria, Australia and the researcher bore the costs of data collection.

Interview questions

The interview questions were developed within two categories or phases. In the first phase, standard organisational questions were asked to obtain an insight on business operations and to categorise the case study organisation as small, midsize or large. Initial questions asked were about the size, geographic location, business type/nature, the type and kind of ERP applications implemented, the time and cost spent on each implementation (if multiple), whether the implementation was internally controlled or outsourced, and the implementation methodology used.

In the second category, questions were asked about the ERP adoption model content, factors, stages of implementation, and the relevance of each factor within each stage. Interviewees were asked to comment on the sequence of activities in each stage of the implementation and relevant factors as previously identified from the research literature and the expert panel. The second interview stage was executed as semi-structured, with questions asked in a sequence relating to the model while allowing the interviewee to provide feedback as they deemed necessary. In some instances, interview questions were paraphrased differently to cater for the different characteristics, knowledge and background of the interviewees (Manaster, 1972).

Recording interviews

During the interview process, it is difficult for a researcher to capture all details correctly; hence recording the interview could be advantageous. It is possible that the researcher may forget all critical details of the interview should they decide not to record it appropriately (Flick, 2009). It is an ethical responsibility of the researcher to seek authorisation or approval from the participant before recording of the conversation could start. Should a participant refuse to be recorded and/or not be comfortable with the recording, this would be a challenge for the researcher and threaten to disrupt conversation (King, 2010).

For the purpose of this research, the audio recording feature of the researcher’s Blackberry mobile was used to record conversations. In addition, the researcher took extensive written notes while conducting the interview and maintained a balanced approach to ensure that the interview followed a sequence and was not disrupted. At the start of each interview, participants were provided with a hardcopy of the questions along with the proposed model. The interviewees were asked to provide feedback on the factors and were encouraged to add, delete or move factors from a stage if they thought it necessary. The researcher immediately 128 | P a g e

noted any changes on a hardcopy. The participants also assisted the researcher to write notes and provide clarification as deemed necessary.

The hardcopies used by the researcher to record interview findings were later scanned and remained as the electronic evidence for later analysis.

Ethics approval

As this research required interaction with people and the handling of data obtained from the public domain, the researcher obtained the appropriate ethics approval from the University’s ethics committee. The application was submitted to the University’s Human Research Ethics Committee in early 2007 and the approval was granted in April 2007. A further application was submitted with amendments in 2010, due to changes in the data collection process to include the case study interviews. The ethics approval for the second application was granted in 2010.