Dissertation Research Design

Sample Dissertation Methodology: Quantitative Survey Strategy

1.1 Introduction

This research project has been one of the most thought-provoking and challenging feature of the master’s course. It provides a chance to endorse, simplify, pursue and even explore new facets of one’s research topic. The research approach adopted is an important aspect to increase the rationality of the research according to Cresswell (2007). The research ‘onion’ is a methodology that was developed by Saunders et al (2003).According to the research ‘onion’, as shown in figure 4.1, the entire process is in the form of an onion comprising of various layers. The research philosophy, research approaches, research strategies, time horizons and the data collection method form the different layers of the onion depicting each of the research process. The process involves peeling each layer at a time to reach the centre which is the actual question of the research. For this research philosophy of interpretivism was chosen along with deductive approach and mainly using quantitative techniques for data collection and analysis (Saunders et al., 2009).

The chapter details the research process adopted and continues with an explanation of the data collection and data analysis methods employed by the researcher including a justification for the approach and method.

The sampling method used by the researcher is discussed and justified and the chapter continues with a commentary of the limitation of the study design.

Finally the issues of observer influence are covered as in the ethical approach to the research and a summary of the chapter is presented

1.2 The Research Philosophy

Research philosophy forms the outermost layer of the research ‘onion’. There are three views based on the way knowledge is developed and corroborated. Individuals or groups rely upon their individual experiences, memories and expectations to derive logic from situations occurring in the society. This logic gets revised over a period of time with new experiences which in turns leads to different interpretations. Therefore it is essential to determine and understand the factors that impact, govern and affect the interpretations of individuals.

According to Denzin and Lincoln (2003) interpretivists believe in multiple realities. Hatch and Cuncliffe (2006) have described how interpretivists try to draw meaning from realities and further creat new ones to analyse the different point of views and to validate them against academic literatures. Since the aim is to interpret the thinking of ‘social actors’ and gaining insights using their pointo of views, it cannot be generalised (Saunders et al. 2007). Remenyi et al. (1998) described an interpretivist as one who tries to ascertain the details of the situation with the underlying motive to unearth the working logic behind the situation.

Eriksson and Kovalainen (2008) point out a flaw which researchers need to take care of while adopting the interpretivism. They say that because of the closeness of the researcher and the researched, there is a likelihood of a bias in the interpretation. The solution is self-reflection

This research attempts to ascertain a relationship, if any, between knowledge management framework in an organisation and the behaviour resulting from the knowledge management practices. This approach adopted by the researcher requires to ‘get close’ to the participants and try and throw light on their acumen of the reality. Thus it can be said that the researcher adopts a interpretivism philosophy.

The Research Approach

The next layer of the research ‘onion’ is the research approach. The design of the research project determines the choice of research approach adopted. If the research involves developing a theory and hypothesis (or hypotheses) and design a research strategy to test the hypotheses then the approach classifies as a deductive approach. On the other hand the inductive approach involves data collection and developing a theory based on the analysis of the data.

In an inductive approach a theory follows the data collection where as it is vice versa in case of a deductive approach. According to Saunders et al (2003), researchers in the 20th century criticised the deductive approach stating that deductive approach help establish cause-effect links between specific variable without taking in to account the human interpretation. Saunders et al. (2000) suggest that researcher should be independent of what is being observed, which the deductive approach dictates. Robson (1993) suggests that the deductive approach is a theory testing practice which arises from an established theory or generalisation, and tries to validate the theory in context to specific instances.

According to Jashapara (2004) Knowledge Management, the central topic of the research, has been around since ancient Greece and Rome and it further mentions that knowledge management is growing at an exponential growth with a lot of literature available. As Creswell (1994) suggests that a deductive approach would be a better approach in such a scenario. Since the data collection for this research involves online surveys by professionals, time is a valuable commodity. In a deductive approach, data collection is less time consuming and works on a ‘one take’ basis, which is also beneficial for the participant of the survey. Following a deductive approach ensures a highly structured methodology (Giles and Johnson, 1997) and can also be basis for future research adopting an inductive approach.

1.3 Research Strategy

The research strategy provides a rough picture about how the research question (s) will be answered. It also specifies the sources for data collection and hindrances faced throughout the research like data access limitations, time constraints, economical and ethical issues. Saunders et al. (2003) explain that the strategy is concerned with the overall approach you adopt while the tactics involves the details like data collection methods (questionnaire, interviews published data) and analysis methods. There are several strategies that can be employed and they can be classified based on the approach, deductive or inductive, adopted.

This research adopts deductive approach. Survey strategy is well suited for this approach. A large amount of data was required to determine the relationship, if any, between the constructs defined in the literature review. According to Saunders et al (2003) and Collins and Hussey (2003) surveys allows data collection and can be addressed to a sizeable audience in a very cost-effective way. Surveys are mostly done in the form of questionnaire, as questionnaire provides standardised data making it easy for comparison. One drawback is the time spent to construct and test a questionnaire. In a survey there is a huge dependence on the participants to answer the questionnaire causing unnecessary delays. There is also a limitation on the number of questions that can be included in the questionnaire. This limitation is from the respondents’ perspective if the researcher wants a high quality of response from the participants.

Owing to the nature and amount of size required, statistical analysis of data, time available for the research and for economic reasons the survey strategy has been adopted for this research.

Choice of research method

According to Saunders et al. (2003) the research methods are in accord with the methods and used for data collection and analysis. Quantitative research is associated with numeric data collection and analysis while, ‘qualitative’ methods are inclined towards non-numeric or data that is gained from inference. However a combined approach can also be adopted as suggested by Tashakori and Teddlie’s (2003). The main advantage is that the researcher can get a different perspective while attempting to answer the research questions and also make more reliable interpretations, ’triangulation’ (Saunders et al. 2009).

For this research data was collected via online questionnaire and was statically analysed and represented using graphs. Number crunching methods are generally used in business and management studies. This method is primarily contributed to quantitative analysis. To answer the research question data was also collected from theories and case studies and analysed qualitatively. To present the analysis in a structured manner and articulate the inferences from the theories and statistical analysis could only be done by means of words (Saunders et al., 2009). By making use of qualitative methods the data could be categorized under “knowledge management environment”, “organisational knowledge behaviour” and “Individual knowledge behaviour” and with the aid of narrative an attempt to establish relationships, if any, between them (Saunders et al., 2009, p.516).

1.4 Time Horizons

Saunders et al. (2009) suggest that a research can be depicted in a snap look alike or can have a diary like perspective. A ‘snapshot’ horizon is termed as a cross sectional whereas the diary perspective is termed as longitudinal. Further Saunders et al (2003) suggest that the time perspective to research (cross-sectional or longitudinal) is independent of the research strategy.

Longitudinal research is adopted when change or development that occurs over a period of time is to be studied. Adam and Schvaneveldt (1991) suggest that in longitudinal studies is very useful in studying human behaviours and development. Longitudinal studies do have a limitation when time is a constraint. In cross-sectional research, a certain phenomenon is studied at a particular point in time. This research tries to explore the relationship between organisational environment and its effect on organisational behaviours in the context of Knowledge Management. It is aimed to find the relation at the present time so a cross-sectional study is adopted. According to Easterby-Smith et al. ( 2002) surveys are preferred in cross-sectional studies. However Robson (2002) g=further says that qualitative methods can also be adopted in cross-sectional studies by considering interviews carried out in a short span of time.

1.5 Secondary Data Collection

According to Saunders et al. ( 2003) secondary data includes both quantitative and qualitative data. Secondary data is usually used in the form of case studies or survey-based research in management and business research. Saunders et al. (2003) have classified secondary data under documentary data, complied data and survey-based data as shown in figure 3.1

For this research the primary data collection was using online questionnaires. However documentary secondary data was also unsed in conjunction to the primary data. The purpose of making use of secondary data was to explore the existing literature and explore the various facets of knowledge management. Documentary secondary data like books, journals articles were used in this research to define the three constructs explained in chapter 2. Also secondary data was used to explore the literature to define the research question. Books by noted authors and academic journals such as Emerald journals, swetswise e-journals, ebsco host were refereed for the purpose of data collection.

The reliability and validity of secondary data relates to the methods by which the data was collected and the source of the data. A quick assessment of the source can ensure validity and reliability of the data. Dochartaigh (2002) suggests the testing of reliability and validity refers to testing the authority and reputation of the source. Articles and papers found in Emerald and Ebscohost are likely to be more reliable and trustworthy which can be inferred from the continued existence of such organisations. Dochartaigh (2002) furthers the point of assessment by looking out for copyright statement.

1.6 Research Sample

Saunders et al. (2003) differentiated sampling techniques as probability sampling and non-probability sampling based on their generalizability. Probability sampling meant that the research question could be answered and generalized across the target population, based on the responses from the sample size. Time was a constraint owing to the business of the participants who belonged to Knowledge intensive industry, selecting a sampling method was a challenge. According to Easterby-Smith et al. (2002), sampling methods must reduce the amount of data to be collected by focusing on the target population rather than a random sample population.

Snowball sampling was selected to ensure that maximum participants could be reached. The research was carried within 7 organisations across 5 countries. The researcher could not personally know so many professional from IT and other knowledge intensive industry, so a few managers was contacted who subsequently forwarded the questionnaire to others with in their respective organisation resulting in to a homogeneous sample (Babbie, 2008). Manager also had to be contacted since all participants could not be addressed directly due company policies restricting external emails.

Since the questionnaire was target at the users of knowledge management tools and practices with in the organisation, the researcher requested the managers to forward the questionnaire across the organisation independent of the managerial status. Sample selection was continued till 20 responses from each organisation were received. 140 samples have been considered for this study.

1.7 Primary Data collection

Questionnaire is a form of data collection in which all the respondents are asked the same set of questions in a pre-set order (deVaus, 2002). Robson (2002) suggested that questionnaires are not effective in a descriptive research as it is requires many open ended questions to be answered. All the participants should interpret the questionnaire in the same manner; the data collected can be reliable. If the questionnaire is worded correctly, less effort is required to administer the questionnaire (Jankowicz, 2000). Questionnaire can be classified as shown in the Figure below. The differentiation is based on the level of interaction between the researcher and the respondents.

The research has an international orientation to it. The respondents are based in 5 countries and it was not feasible for the researcher to meet each respondent. So a self-administered questionnaire was the most appropriate option. Time and monetary constraint further helped to narrow down the survey to an online questionnaire where the questionnaire was forwarded to the emails. Email offers a better reliability as the respondents would access their own emails and respond to the questionnaire (Witmer et al., 1999). In this case the questionnaire was sent to the managers who further forwarded the emails to their colleagues. In this scenario online questionnaire was a more feasible option because it is easy to forward emails, unauthorised access to emails would be difficult and the responses would go directly to the researcher without them being disclosed or discussed with.

The questionnaire has been divided into two parts. The first part consists of information regarding demographics such as organisation location, age, tenure in the organisation and job role. The second part consists of questions related to organization’s knowledge management practices, knowledge behaviour and use of the knowledge. The data required for the research required responses from managerial and non-managerial employees working in a knowledge intensive environment. It was required to create an accurate cause – effect relationship of the KM practices with respect the organisational environment and behaviour of employees. This required honest responses about the KM practices.

Appendix shows the questions that were asked to define the relationship amongst the construct defined in the literature review. Likert scale has been used to score each question and score will be given from strongly disagree(1) to strongly agree(5) to. In the questionnaire 1 question has been framed using negation and in a reverse order. Podsakoff et al. (2003) suggest that this should be done to ensure that respondent pay attention while reading the question. All questionnaires were returned within 72 hours. Considering the incentives and time constraints for the respondents the questionnaire was designed so that it does not take more than 8-10 minutes to be answered.

1.8 Dota Analysis methods

Qualitative and quantitative data has been used in this research. Qualitative data has been used to study the literature about knowledge manangement and define the constructs that for the basis of the research question. Quantitiative data was collected primarily with the help of questionnaire.

1.9 Methodological Review

Saunders et al. (2003) emphasis on two aspects of data collection: validity and reliability. The validity and reliability of secondary data has been explained in SECTION. Saunders et al. (2007) suggest that in case of a questionnaire pilot testing should be done to ensure the validity of the question and the reliability of the data subsequently collected. The questionnaire used for the survey has been tested on a group, to test the comprehensibility of the content and the logic of the questions. Bell (1999) suggests that a trail run should never be compromised even if time is a constraint. While testing the questionnaire the respondent were asked regarding the time taken to complete, ambiguity of the questions, if any questions caused a uncomfortable feeling or awkward state of mind and the last was the structure. Validating the questionnaire ensures that the response for each question and the motive for the question are the relevant (Saunders et al. 2000).

Reliability of the questionnaire depends on the consistency of the response to the same questions. To ensure this the questionnaire must be answered twice by the respondent at differing time (Easterby-Smith, et al. 2002). This may be difficult due time constraints but should be done. Mitchell (1996) suggests that the responses of the questions should be checked for consistency within the subgroup. In this research the questionnaire has been divided in to 4 sections. During the pilot testing the responses where checked for consistency with in each section to ensure the reliability. The results can be generalised to an extent due to the sample size and inferences are gathered based on the statistical analysis. Steps have been taken to ensure the anonymous nature of the questionnaire so that the responses are honest and unbiased.