3 CHAPTER THREE: THE RESEARCH METHODS AND PROJECT
3.2 Triangulation
Section 1.3.1 discusses how research can withstand external scrutiny that includes terms such as validity, reliability and generalisability (Easterby-Smith et al., 2002). One item that may influence this scrutiny is a bias that exists along the course of the research. A technique to mitigate this bias is called triangulation. It is, therefore, important to recognise the potential bias that may exist within the research setting; and to design a procedure for the triangulation.
A few aspects of my research may create some potential biases during the research period. Four main potential biases are identified and the triangulation associated with each of them is explained in this section. One, it is the potential for a cultural bias. My research involves knowledge management systems that are implemented across geographical boundaries. Data collection in Schlumberger, Power International and Friends Provident, therefore, involves users who work in different parts of the world. While this data collection design brings rich insights, it also carries a potential cultural bias during the data collection. Two, it is the organisational politics bias. The research subject involves KMS that, in general, is a sensitive issue to discuss within organisations due to the possible dichotomy of opinions and the amount of funds invested in KMS. Interviewees may be very cautious in answering questions. Three, it is the researcher’s bias. I was a senior manager in Schlumberger when the Project One took place in the research site even though I was never part of the InTouch development and/or deployment team. This situation may bring some advantages such as access to data and interviewees; and at the same time this situation carries disadvantages that I may assert my bias during data collection and data analysis. Four, it is the interviewees’ personal bias. When interviewees are very much involved in the research subject they have in-depth knowledge that is required for an exploratory nature of the thesis; at the same time they may exercise a certain bias toward the research subject depending on their position related to it.
Beneficial Results
Consequences/ High-level strategy elements
Attributes of
Knowledge management Attributes Consequences Values The Creation, Mobilisation,
Table 3-5 lists these potential biases and the techniques to address each of them, including triangulation.
The potential
bias The situation in the research setting The techniques to mitigate the bias Cultural bias The research
subjects are KMS used across the globe. Therefore, potential bias exists in data collection from different geographical areas.
The design of data collection includes interviewees from different geographical locations across the globe: Asia, Europe, Africa, America, Pacific. In Friends Provident case, interviewees are from the United Kingdom only because the KMS is used locally. However, this approach creates another issue of the cultural bias from each interviewee. Two areas are particularly addressed in the attempt to overcome this bias: 1) all interviews are focused on a common unit of analysis, i.e. the business process. Therefore, cultural bias is reduced because the practices within the process are standardised across the globe; 2) documents such as the KMS technical descriptions, the Infosys/Wharton award document, and SPE publication are used to confirm or reject culturally-biased data.
Politics bias Knowledge management is a sensitive subject in companies. Therefore, interviewees may apply politics bias when answering questions.
Interviewees are selected from different ranks and different functions in each of the participating companies. Most of the data collection is conducted through individual interviews. This reduces the tendency of organisational script where each respondent says the same thing. Data from the anonymous internal InTouch surveys are used to triangulate the data in the attempt to overcome this politics bias.
Researcher’s
bias The researcher was a senior manager in Schlumberger when Project One took place. Significant challenges for bias may be: 1) Pre- understanding; 2) Role duality; 3) Organisational politics that refers to the potential threat the
researcher-manager may bring to the organisation.
A technique “mirroring or reflecting” is applied during data collection (see section 3.1.3). A statement prior to beginning the interviews that the data is solely used for academic research should reduce the bias from both the researcher and the
interviewees. Verification of the transcription of the interview data by the interviewees also reduces this bias. In Project Two, the researcher bias is also avoided by referring back to the interviewees for futher interpretation of the data previously collected in Project One.
The researcher’s bias is very much avoided by applying triangulation. Both internal and external reports are examined to confirm the researcher’s analysis or reject it when items are conflicting and suggesting that the researcher’s bias is involved.
Interviewees’ personal biases
Respondents in data collection are very much involved in the Knowledge Management Initiative, the research subject, and, therefore, they may have certain personal biases toward the subject.
Interview technique includes repeating the search for answers from the interviewees through different questions. The personal biases of the interviewees are also addressed through triangulation. Extract of data from different interviews are cross-checked to each other. In the case that the data is unique it is then cross-checked with the technical description of the KMS, the survey data, the clients’ presentation and, when available, the external documents. In the case that triangulation does not confirm support the collected/analysed data; a further discussion is conducted with different respondents.
As shown in Table 3-5, the available data in the company’s archives used for triangulation include internal documents such as technical descriptions for the KMS, clients’ presentations and the internal survey data in Schlumberger and Power International. The external documents that are used for triangulation include Infosys/Wharton award document and Society of Petroleum Engineers (SPE) publication about InTouch, the Schlumberger KMS for its technical service delivery process.
The procedure I followed to carry out the triangulation can be described with examples as follows:
1. A couple of respondents from Schlumberger claimed that the success of InTouch had very little things to do with Information Technology (IT). They claimed that with traditional telephone technology the system should also function well. This claim is interesting as the majority of the respondents agree that IT contributes significantly to InTouch operations, exemplified by one respondent who suggested that “Technology is the price of admission since we would not get far without it”. The first triangulation for such a claim is to compare this claim with the data from other respondents which differ in opinion. In addition, the data from the Infosys/Wharton award document explicitly stated that the Schlumberger intranet technology played a critical role for InTouch. Furthermore, several minutes of the meeting documents dated 1993 to 1998 explicitly show the effort made in aligning intranet to InTouch operations needs. Therefore, the data that claim IT is not significant have to be rejected.
2. Training was mentioned as one critical attribute by all respondents. I also learned that Training is one aspect that Schlumberger puts a lot of attention to. At the same time, the researcher is also a believer in training. Therefore, mentioning training can be considered as both organisational script and also a response that the researcher would like to hear. The question is whether training is really a critical attribute for operationalising InTouch. I went through the internal documents and verified the numbers of training conducted related to InTouch and I found that in average, three training sessions per week were conducted in each site during the period of 1993 to 1998. The data from the InTouch user survey also indicated the importance of training in operationalising InTouch. With this triangulation, a decision was made that Training is one important attribute for InTouch operations.
3. Operationalising KIAT in Power International showed that the Director of the Software Services claimed that leaders in Power International walk the talk as far as Knowledge Management System (KMS) is involved, “we really have the commitment from the management, and they keep talking and promoting this”. The issue is that the director himself is one of the top management and, therefore, it raised the question of the interviewee’s personal bias. This particular item was then triangulated with the extract from the other six respondent interview data. These six respondents were KMS users. The data confirmed that the leaders in Power International support fully the company KMS.