This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
This chapter will discuss the methodology of this research study. The research process summary will be followed by discussion of research philosophy, and then the research approach that will be adopted in the current research. Various data types, data collection methods and sampling techniques used in this research will be discussed. Finally, various limitations of this research will be covered.
3.2. Research Process
A widely accepted definition of research as presented by Collis & Hussey (2003) is that research is a methodical and systematic process of investigation and enquiry that increases knowledge. The research process explains how the research has been conducted. Research process contains several stages which are common to all investigations carried out on scientific methods.
The stages of the research process are listed below:
Identify the Research Topic
Define the Research Problem
Determine How to Conduct Research
Collect Research Data
Analyse and Interpret Data
Write Dissertation/ Thesis
Fig. 3.1: Research Process (Collis and Hussey, 2003).
Stages One and Two have been investigated in the previous chapters, while the next two stages will be reviewed in the present chapter, and the data analysis and interpretation will be examined in chapter five. Final step is writing the thesis.
3.3. Research Philosophy
The research philosophy or research paradigm refers to the progress of scientific practice based on the nature of knowledge and the assumptions or feelings of people about the world. In brief, the research philosophy depends on the researcher's way of thinking about the knowledge development.
According to Saunders et al. (1997), two views dominate the literature about the research process; phenomenology and positivism. They are different, if not mutually exclusive, views about the way of knowledge development and judgement of knowledge as being acceptable.
However, researchers have added one more view to research philosophy, i.e., realism. A brief discussion about each philosophy follows:
Rubin and Babbie (1997) defined positivism as a paradigm introduced by Auguste Comte, that illustrates that social behaviour can be studied and understood in a scientific and rational manner, in comparison to explanations based on religion or superstitious beliefs.
Remenyi et al. (1998) states the assumption of positivism that the researcher is independent of the subject of research and neither affects nor is affected by it.
Thus, many writers concerned with the social sciences have, whilst supporting a positivism approach for the natural sciences, called into question its relevance where the focal points of investigation are human beings in terms of their behaviour as opposed to their observable physical characteristics that comply with the laws of nature. The consequence of this is a body of theory related to research methodology termed phenomenology.
As claimed by Gorner (2001), Phenomenology is a special kind of study related to consciousness. Instead of being attached to physical basis of our mental states, rather it is concerned with the causes of our mental states. It is exclusively concerned with consciousness as the subject of consciousness is able to be aware of it.
Every individual behaves differently from others. According to Saunders et al. (2000), this research philosophy is efficient at finding the reality of a hidden hypothesis. In other words, the phenomenological approach is not about understanding what is happening; rather than it is about finding why something is happening. It is ideal for research trying to find the solution to why something is happening and what solution can be found. Phenomenology helps to understand the crucial research hypothesis.
According to Collis & Hussey (1997), phenomenology is the predictive type of research that will analyze from the basis of the hypothesis.
Hence, after carefully looking at both, the researcher believes that phenomenology is the best suitable approach for this research study.
3.4. Research Approach
While conducting a research, two approaches can be followed: (i) deductive and (ii) inductive.
3.4.1. Deductive Approach:
The deductive approach is described by Saunders et al. (2000) as one of the two qualitative approaches that is concerned with shaping and adopting the qualitative research process and aspects of data analysis using existing theory. While following the inductive approach, the research project seeks to build up a theory through data collection and is adequately grounded through many relevant cases.
As Robson (1993) describes, deductive approach has five sequential stages which are listed below:
Deducing a hypothesis from theory. A hypothesis is a testable proposition that describes the relationship between two or more events or concepts.
Articulating the hypothesis in operational terms (indicating exact way of measurement for variables) which propose a relationship between two specific variables.
Testing this operational hypothesis that will involve an experiment or some other form of empirical inquiry.
Examining the specific outcome of the test. It will either result in confirming the theory or indicating the need for any modification.
If necessary, modify the theory in the light of findings. Afterwards an attempt is made to verify the revised theory by going back to the first step and repeating the whole cycle.
In brief, the deductive approach is a scientific research which involves the development of a theory that is subjected to rigorous testing.
3.4.2. Inductive Approach:
Saunders et al. (2000) describe the inductive approach as a process which involves data collection and leads to development of theory as a result of data analysis.
Hyde (2000) explains inductive approach as a theory-building process which starts with observations of specific events and attempts to generalize the phenomena under study.
These two approaches have no rigid differences, but the inductive approach is associated with phenomenology paradigm and the deductive approach is normally associated with positivism paradigm.
The following table enumerates the major differences between the two approaches:
Moving from theory to data.
The need to explain casual relationships between variables.
The collection of quantitative data.
The application of concepts to ensure clarity of definition.
A highly structured approach.
Researcher independence from what is being researched.
The necessity to select samples of sufficient size in order to generalize conclusions.
Gaining an understanding of the meanings humans attach to events.
A close understanding of the research context.
A more flexible structure to permit changes of research emphasis as the research progresses.
A realisation that the researcher is part of the research process.
Less concern with the need to generalise.
Table 3.1: Differences between Deductive and Inductive Approach (Hyde 2000).
For the current research, the inductive approach will be followed by developing a functional model of ERP through the literature review, and then by testing it.
3.5. Data Types
The next step is data collection after the selection of research philosophy and research approach. In this section, the sources of data are discussed, followed by the differences among various data collection methods and techniques, and finally the different types of data collection techniques used in business studies.
Hussey and Hussey (1997) describe the two types of data as following: (i) qualitative data, and (ii) quantitative data. Qualitative data is described as data with non-numerical properties whereas Quantitative data is concerned with data collected in form of numerical terms.
Though, the phenomenology paradigm of inductive approach will be followed in this research, there is still a combination of quantitative and qualitative method that inputs data generating activities. This method helps to introduce an element of data triangulation that assists in validating the research work.
Triangulation is defined by Denzin (1970) as a combination of methodologies in the study of the same phenomenon. This will lead to higher level of reliability and validity of the research. There are two basic sources of data: primary data and secondary data.
3.5.1. Primary Data
Collis et al. (2003) describes primary data as the original data, which is collected at source. The researcher will collect primary data by conducting quantitative research through questionnaires to HP's employees regarding their opinions about the impact of ERP systems on management practices.
3.5.2. Secondary Data
According to Hague (2002), "It is a collection of information which has already been published."
Collis et al. (2003) describes secondary data as data which already exists, like books, films and documents such as annual reports, published statistics, accounts of companies and internal records ( e.g., personnel records) kept by organisations.
As explained by Saunders et al. (2003), the major advantages of secondary data are:
It saves time and money.
It provides high quality and reliable data.
The main disadvantages of secondary data are:
The collected data is for different purposes.
Access to data may sometimes be difficult and expensive.
The aggregations and meanings may be unsuitable to the research under study with no real control over data quality.
Secondary data in this research study will be collected from sources like books from the library, newspapers, HP and SAP company websites, and articles from the Emerald and Mintel reports.
3.6. Data Needs
ERP systems are attracting a great deal of interest and investment across a range of industries, especially high-tech multinational companies, in order to have dramatic effect on customer retention rates and profitability. This clear growth and importance of ERP systems in this competitive business environment led to the present research into ERP systems at a high-tech multinational company and particularly at Hewlett-Packard (HP).
This research study has been carried out basing on phenomenological research perspective. The nature of data to be collected in this research is combination of quantitative and qualitative. The analysis method will be statistical. Therefore, the data collected needs to be mostly in quantitative form which is accomplished through surveys and questionnaires.
3.7. Selection of Organisation
Hewlett Packard (HP) has been chosen as the case study subject for this research report. Currently HP is one of the major companies in the United Kingdom offering best features on computers and printers to customers. Enterprise Resource Planning supports Hewlett Packard in the following business functions: Market Basket Analysis, Customer location, Customer Management Systems, Sales Force Automation, Opportunity Management, E-commerce support, Telesales, Direct- mail, and Direct Catalogue sales. With the aid of effective consumer data and information, Enterprise Resource Planning can enable the executives to have more effective supply and demand forecasting, and they can come up with new product launches and new marketing information campaigns.
3.8. Criteria for choosing Organisation
Case study research as cited by Cassell & Symon (2004) is very commonly used for organization studies and can be depended as a rigorous research strategy in itself. Case study research is made up of a comprehensive examination of collected data. Its aim is to analyse the processes and context of the issues under study. This approach is particularly suitable for obtaining answers to research questions that demand in-depth review of organisational procedures. The criteria for choosing the organisation for this research was established as:
The selected organisation must be using ERP systems.
The organisation should be big so that the sample size is large enough to ensure generalisation of the research.
The organisation should be a well-known brand in its market to ensure reliability of the data.
The employees of the organisation should be skilled enough to understand the importance of ERP in order to ensure validity of the research data.
Hewlett Packard is a well known company who invests in technology and operates in 170 countries all over the world. The company keeps on searching and exploring the opportunities that how the new technologies can assist people and organisations in addressing current challenges of time, and comprehends their possibilities, goals and objectives.
Hewlett Packard is giving the best value of money to their customers by lowering the prices every now and then. In this way they can lock in their customer and lock out their competitors. HP online shopping offers the price comparison software for businesses and private customer. The software helps customers know that HP is giving them the best deals.
Currently HP is one of the major companies in the United Kingdom offering best features on computers and printers. HP has got customer data from different sources weather it's online shopping or purchasing from store or customer complain in their call centre. But this data is neither integrated nor properly utilised yet. HP can drive its customer service relation with the help of (Enterprise Resource Planning) database through measurable performance. Because success can only be achieved if the organization acts before its competitors and respond fast and at the right time to the ever changing customer needs and demands.
HP can target its customer by using the database because every customer is unique and its preferences are also unique. It can focus its marketing effort with the help of intelligent (Enterprise Resource Planning) data base system to a particular customer. In this way it can easily differentiate and segregate its profitable customer. Once they have a clear view of their customer with the help of the database, then it will be very easy for them to develop the right marketing mix plan (Product, Price, Place, Publicity, People, Physical evidence and Process) about their customer. If HP has the real time right information about its customer needs and preferences, It will be very easy for its Executives Directors to create and formulate smart decisions about the companies organisational aims and objectives.
3.9. Data Requirements and Collection Process
This section will discuss various data collection methods and techniques, the differences in these methods, followed by data collection methods used in this research.
Research methods are referred by Ghauri et al. (2005) as orderly, systematic, and focused collection of data with the purpose of obtaining information to solve or answer a particular research problem.
Data collection methods are separate from data collection techniques. In data collection methods, data are collected through historical reviews and analysis, surveys, field experiments and case studies. In contrast, data collection techniques are step by step procedures that need to be followed in order to collect data and to analyse them for solving the research question.
In business research studies, two data collection techniques are used which are:
Qualitative techniques or qualitative research,
Quantitative techniques and quantitative research.
The main difference between qualitative and quantitative research is the procedure followed for data collection.
3.9.1. Qualitative Research
Qualitative research is conducted to collect data about how people think, feel, and the explanation of their behaviour. It includes all techniques which attempt to illustrate the view of those being studied without assuming that these views are necessarily held by others who have not been investigated.
There are two categories of qualitative research. One, continuous research, takes place at regular intervals over an extended period of time so that trends in the data can be identified. Predictions can be made based upon the trends observed.
Second, Ad Hoc research refers to a one-off investigation to study an issue that has arisen and for which further information is needed to help in decision making. it cannot be automatically assumed that the findings will hold true for the future because the scope of research is limited to time period in which it is conducted.
Different qualitative techniques that are used to collect data are:
Semi-structured and individual in-depth interviews,
188.8.131.52 Focus Group
Focus group is described as an interview conducted in a non - structured and natural manner simultaneously by a trained moderator for a small group of participants.
According to Calder (1977), a focus group is used to generate or evaluate new ideas for products and product uses and explains the result for other studies.
Normally, a focus group has three stages. First, a group member discusses products used for a particular situation/need with little intervention by the moderator; secondly, members discuss how they rate alternative products; finally, the moderator probes their feelings in order to uncover what they favour one product over other.
According to Lehmann D., 1997, focus group is flexible tools, they allow a detailed questionnaire, and they can take advantage of unexpected responses that were initially thought unimportant. The situation offers the individual member the opportunity to express his opinion, which was not possible to say personally.
184.108.40.206 Semi - Structured Interviews
Saunders et al. (2000) describes that in this method, there are themes and questions to be covered by researcher; however, the themes and questions are different for every interview and the order of questions is different. Similarly, the researcher may omit some questions in relation to the research topic.
220.127.116.11 In-depth interviews
This type of interview is described by Lehmann (1997) that a probing question is directed at a single subject by a single interviewer. These types of interview require at least one hour and require a high demand of pay. These interviews are derived from 1950s psychology and have enjoyed considerable popularity. Initially, they were used to prove Freudian assumptions (sexual in nature) by Ernest ditcher, but now they are applied for laddering and repertory grids to gain in-depth understanding of interviews (Lehmann D, 1997).
18.104.22.168 Projective Techniques
Projective techniques are used in creative sessions. They include sentence completion, spontaneous drawing and collage making, word and picture sorting, and personification. These techniques are used for drawing out attitudes from the respondent's mind that are hidden and unformed. These are helpful in identifying responses to questions that are beyond conventions and expectations (Mariampolski, 2001).
Collis & Hussey (2003) describe ethnography as an approach to understand the observed patterns of human activity through the use of shared and socially acquired knowledge.
This process is time consuming and spans over a long period of time.
22.214.171.124 Case Study
According to Collis & Hussey (2003), the case study method is categorized as exploratory research which is used in deficient areas of knowledge where the number of theories is few.
Case study method is argued as a very worthwhile way of exploring an existing study by Saunders et al., (2000).
3.9.2 Quantitative Research
Quantitative research is typically used in conclusive research where the research purpose is specific, in order to verify insights and to aid in selecting a course of action. In the situations with clear data needs, well defined data sources and large size of sample (in order to permit generality of findings), this method is used. Conclusive research further involves data collection being "ridged" (a well laid out procedure) and analysis is formal (Parasuraman, 1991).
Recommendations are more final than provisional. Conclusive research is a type of research that intends to verify insights and provide help to decision makers in selection of a certain path of action (Parasuraman, 1991).
Therefore, the major aim of quantitative research is to provide help to decision makers in choosing the best strategy of action in any certain circumstances, as accurate facts and figures are essential.
There are two approaches used in quantitative research. They are the survey and observation methods.
Surveys are mainstays of marketing research. Surveys are relatively cheap and easy to administer and are also the only means of getting thoughts and attitudes measured.
The survey method is the data collection method in which the researcher directs his/her questions towards some relatively large groups of people, for exploring the issues (Jankowicz, 1995).
Ghauri & Gronhaug (2005) explains that this method is very effective for collecting attitudes, opinions, and descriptions, as well as for getting cause and effect relationships.
In conducting surveys, there are many contact approaches. One major approach is face to face surveys (also known as personal interview surveys). In this method, interviewers collect the information at the location convenient for the respondent. The survey can also be conducted through telephone and drop off, call back, pre-existing panels, group interviews, location interviews and finally through mail/e-mail.
Observation is just what the name implies: observing people, objectives, or events. Observation can be made by human resources or mechanical devices.
Ghauri et al. (2005) describes observation method as the method in which data collection involves observation of other people's behaviour by listening and watching in such a way that permits some type of learning and analytical interpretation.
Questionnaires are described by Sharma (2001) as formal set of questions used to extract information from a group of target audience. The questionnaires are distributed personally or can be mailed to the participants. It's an efficient and useful method for collection of data.
There are two categories of questionnaires: (i) Structures Questionnaires and (ii) Unstructured Questionnaires.
A Structured Questionnaire contains questions and permitted responses posed in a predetermined order. Such type of questions are also called closed questions because they only allow the respondent to select one or more among the answers already given by the researcher.
In the case of Unstructured Questionnaire, there are no limited choices, instead the respondents are expected to answer according to their own frame of mind and only the context of the question is stated in the questionnaire. These types of questions are also called open-ended questions. This type of questionnaire is mostly used for exploratory type of research where structured questioning doesn't work.
For this research, the researcher will conduct quantitative research by using a survey technique through a structured questionnaire. The researcher will collect data (filled questionnaire) from the respondents through a survey at the respondent's place and through email. The questionnaire design will be discussed in the next step.
3.10 Questionnaire design
A questionnaire is a formalised method of data collection from respondents. Measurement is the main function of questionnaires. Questionnaires are used for measurement of attitudes, past behaviours, and respondent characteristics.
Designing a questionnaire is a skill that is learned through experience not by studying a set of sequential instructions. Kinnear and Taylor (1996) say that the only way of developing this skill is to write questionnaires, use them in a series of interviews, analyse their weaknesses, and revise them.
The reliability and validity of data collected in the research, depends on the questionnaire design and the rigour level of pilot testing. As described by Foddy (1994), in order to ensure that a research is valid and reliable, the following four stages must occur:
Researcher is clear about the required information and designs a question.
Respondent decodes the question in the way the researcher intended.
Respondent answers the question.
Researcher decodes the answer in the way the respondent intended.
Questionnaires have different types such as open verses closed-end questionnaires, direct versus indirect, aided versus unaided, phraseology, response format and general suggestions type. The researcher will use a closed-ended questionnaire to conduct the research.
3.9.1. Scale Questions:
For collecting attitudes and belief data, normally the scale or rating questions are used. The most common approach that is used is the Likert-style rating scale. Interval sales are most used in Likert scales. The interval Likert scale is used to measure the extent of agreement or disagreement of a person to the questions. The scale which is used most commonly is 1 to 5 corresponding to different levels of agreement. Normally the scale in use will be
1 : strongly disagree,
2 : disagree,
3 : not sure,
4 : agree, and
5 : strongly agree.
Likert scale was developed in 1932 being the five-point bipolar response format that is familiar to most people today. Questions based on this scale always ask people to answer by indicating how much they agree or disagree to any specific statement, believe any fact to be true or false and approve or disapprove any idea.
Nunnally (1978) explain that the ends of the scale can be expanded by adding "very" for creating 7-point scale, which helps to improve the level of reliability by reaching upper limits of reliability.
It is best to use a scale as wide as possible because the responses can always be collapsed later on into compressed categories for the purpose of analysis.Â The scale used by the researcher in the present research was a 7-Likert scale, and with the scale representing 1 as strongly agree and 7 as strongly disagree.
3.9.2. Pilot study
The researcher used a pilot test before going to collect data with a questionnaire. The pilot study is useful to check the reliability and validity of questionnaires. Five questionnaires were sent for pilot testing by the researcher. Feedback was asked from three professors in business school through the research supervisor and two SAP working people in HP. After getting some of the suggestions from their response, the researcher modified the questionnaire design and formulated a final questionnaire to conduct the research.
According to Saunders et al. (2000), sampling techniques provide a diverse array of methods through which the researcher is able to reduce the amount of data needed for collection by only choosing data from a sub group instead of all possible instances or cases.
Churchill et al. (1995) describe that the process of drawing a sample consists of following six steps:
Defining the population.
Identifying the sampling frame.
Selecting a sampling procedure.
Determining the sample size.
Selecting the sample elements.
Collecting data from the designated elements.
In this research, population means the employees who are the users of ERP systems (SAP) at the HP factory at Erskine. A sample was selected from this population. For the process of selecting the sample, there were two main types of sampling designs: (i) Probability sampling and (ii) Non-probability sampling.
3.10.1. Probability Sampling
Churchill et al. (1995) say that in probability sampling method, the calculation is made of a likelihood that a probability sample will include a given population because the final sample are selected through a specific process objectively and not according to desires of the researcher or field worker.
There are mainly five types of probability sampling. Table 4.2 compares briefly these methods. Convenience sampling technique of non-probability sampling was used for the current study.
Sampling Frame required
Size of Sample needed
Geographical area to which suited
Ease to explain to support workers
Advantages compared with simple random
Accurate and Easily accessible.
Better with over a few hundred.
Concentrated if face-to-face contact required, otherwise does not matter.
High if large sample size or sampling frame not computerized.
Relatively difficult to explain.
Accurate, easily accessible and not containing periodic patterns. Actual list not always needed.
Suitable for all sizes.
Concentrated if face-to-face contact required, otherwise does not matter.
Relatively easy to explain.
Normally no difference.
Accurate, easily accessible, divisible into relevant strata.
Better with over hundred.
Concentrated if face-to-face contact required, otherwise does not matter
Low provided that list of relevant strata available.
Relatively difficult to explain.
Better comparison across strata.
Differential response rates may necessitate re-weighting.
Accurate, easily accessible, relates to relevant clusters, not individual population members.
As large as applicable.
Dispersed, if face-to-face contact required and geographically-based clusters used.
Low provided that lists of relevant clusters available.
Relatively difficult to explain until clusters selected.
Quick but reduced precision.
Initial stages: geographical.
Final stage: only needed for geographical areas selected.
Initial stages: as large as applicable.
Final stage: suitable for all sizes.
Dispersed if face-to-face contact required, otherwise no need to use this technique.
Low sampling frame for actual survey population only required for final stage.
Initial stages: relatively difficult to explain.
Final stage: easy to explain relatively.
Difficult to adjust for differential response rates. Substantial errors possible.
Table 3.2: Probability Sampling Techniques (Saunders et al., 2000).
3.10.2. Non-Probability Sampling
Non-probability sampling involves personal judgement at some stage in the process of selection. This technique has a main assumption that the sample will be randomly chosen.
Non-probability sampling techniques are of five types. The following table 3.3 compares briefly these techniques:
Likelihood of sample being representative
Types of research in which useful
Control over sample contents
Reasonable to high, although dependent on selection of quota variables.
Where costs are constrained/data needed very quickly, so an alternative to probability sampling needed
Moderately high to reasonable.
Low, although dependent on researcher's choices: Extreme case
Where working with very small samples.
Focus: unusual or special. Focus: Key themes.
Focus: importance of case.
Low, but cases will have characteristics desired.
Where difficulties in identifying cases.
Low, but cases self-selected.
Exploratory research needed
Very little variation.
Table 3.3: Non-Probability Techniques (Saunders et al., 2000).
3.10.3. Sample Size
Fifty questionnaires was distributed to the Personnel department of HP, Erskine, for data collection from each department. The researcher received feedback from thirty-one respondents. Out of them, three respondents from the Personnel department, eight respondents were from the IT department, two respondents were from the Financial department, ten respondents were from the Production department, six respondents were from the Marketing (Sales and Distribution) department and two respondents were from the Warehouse department. The conclusion was drawn from the collected data.
3.11. Analysis of Data
After collecting data from selected sample, the author used SPSS 12.0.1 software for the analysis of data.
SPSS means 'Statistical Package for Social Studies'. SPSS Inc. being founded in 1968, has now become one of the leading world-wide providers of predictive analytical solutions and softwares.
SPSS predictive analytical software is useful in anticipating change, managing more effectively both daily operations as well as special initiatives, and realizing positive, measurable benefits (source: www.spss.com)
There are several data analysis tools such as SNAP/SQUAD, SPSS 12.0.1, ORIGIN, SIGMAPLOT9.0 and TPL TABLES 5.1 etc.
The author has chosen SPSS 12.0.1 for data analysis because it is a widely used tool in the market and also is easily available to the author. The process of data analysis is clearly explained later on in the following chapters.
3.12. Limitations of Research Methods
Some limitations restrict the conduct of the research. These limitations include, firstly, a restricted sample size due to limitation of resources and time. Secondly, the author was unable to maintain balances between the numbers of respondents from each department.
Thirdly, the author had chosen only the HP, Erskine factory for data collection because it was near to place where the author lives and was easily accessible. Also, the response rate is very low due to respondents' available time to address the questionnaire and their lack of interest in responding. These are the restrictions and inaccuracies that might have transpire during the course of research.