Small and Medium Enterprise (SME) creation is an important factor that drives developing world economies. Curiously, African SMEs, while being significant in number, tend to be insignificant in terms of contribution to GDP. Overall, they contribute only 1% to GDP growth on an annual basis. Further, the creation and operation of SMEs in Africa appears to be continually stifled by both intrinsic and extrinsic factors. The study uses quantitative and qualitative methods to investigate determinants of SME success and failure in Africa as well as the characteristics of successful and unsuccessful African entrepreneurs. The methodology chosen for the study was a mixed methods approach. The quantitative findings of a survey of 705 businesses in Africa were complemented by qualitative personal interviews of sixteen successful entrepreneurs. Principal Component analysis and regression analysis are used to analyse the questionnaire data and build a predictive model of business profitability. The quantitative results are more powerful for predicting individual SME success rates than they are for predicting country specific SME success rates.
The mixed method approach utilized in this study found significant results with respect to the goals of the research. There appear to be significant variables that create barriers to the creation and operation of African SMEs. These variables most significantly include infrastructure, managerial skill, education, access to capital, and access to qualified staff. These barriers also form a pattern of extrinsic and intrinsic determinants to success. The extrinsic determinants are infrastructure, access to capital, poor government policies, and competition from imports while the intrinsic determinants are education, staffing, and managerial skill. This research finds that there are likely to be distinctive differences between the background, experience, skills and knowledge behind successful SME creation and unsuccessful SME creation. These differences are most significantly managerial skill and education. Finally, this study finds that there are distinctive differences between successful business creators and unsuccessful business creators with respect to their motivations. These differences were found to be opportunism and the concern for the welfare of those surrounding the SME creator. In conclusion, it can be said that SMEs could be created with a view of participating in the export-growth path either directly by means of producing for the international market or indirectly by means of sub-contracting.
4.1) DEVELOPING THE THEORETICAL FRAMEWORK
The section will form a framework for the construction of the conceptual model of SME ownership, success and failure study. To achieve this, three relevant dimensions will be presented, forming a holistic yet structured and simple construct of model believed relevant for addressing issues of SME creation, success and failure in Africa and at the same time feasible for this research. The framework of relevant variables is presented and addressed as a starting point for the construction of the conceptual model.
As the research design stipulates, the model mapped for the questionnaire survey of African SMEs participating in this research project was used for comparison (data analysis) in Chapter 5, with the conceptual model developed in this chapter. The objective was to gain deeper understanding, knowledge and insight into how to identify, describe and analyse the multidimensional relationships between internal and external African SMEs environment and the construct of growth, survival and failure factors that constitute and promote entrepreneurial activities within the SMEs.
4.1.2) Three Dimensions of SMEs Characteristics Forming a Model
The survival and growth of a small company depends on a series of internal and external environmental factors (Arbaugh and Michael, 2003). Building on the research ECA (2001), this research project takes the position that there are three dimensions of SMEs characteristics relevant to studying African SME ownership, success and failure. These three dimensions are presented and addressed below, forming a holistic, yet simple construct that can be applied when approaching the conceptual framework based on the objectives of the research as stated in Chapter 1.
Source: Margaret (2004)
Based on the above dimensions, this thesis primarily used an explanatory approach That is, a specific number of personal, business and entrepreneurial characteristics of SMEs were studied at start-up and related to eventual outcomes (survival, growth, failure).
While there is a mass of literature concerning SMEs and entrepreneurship there is no generally agreed theoretical framework for carrying out research in this field. In order to conduct this study based on the three dimensions and to also understand the vast literature on business creation, success and failure, a framework based on the extant literature was developed. The framework has been designed specifically to facilitate explanation of the three dimensions and also to aid research relating SMEs and may not be considered useful where large businesses are being investigated. The research framework is shown in diagrammatic form in Figure 4.2 and it is a development of the model used by Watson, Hogarth-Scott and Wilson (1998) with contributions from the work of Fick (2002, P16-18). The model classified the three dimensions into external and internal factors, that is, factors that the SME has no direct control over them and factors that are within the SME direct manipulation.
Factors Affecting Internal and External Environment of SMEs
Source: Developed for this research from the model of Watson, Hogarth-Scott and Wilson (1998).
Mazumdar and Mazaheri (2003) had undertaken similar research as the aims of their research were “to discover the types of crises faced by infant businesses, how these crises were met, and the factors affecting growth and decline – particularly with the business people themselves”. Further, it was their objective to discover whether within his findings any indications as to the most appropriate ways of training those intending to start a small business and supporting them once in business” (ibid, 130).
The difference between my thesis and theirs is that they used an action research methodology which involved a large degree of participation in the businesses and offering guidance. Their approach does however have useful and significant insights to the particular methodology this thesis will adopt. Mazumdar and Mazaheri (2003) also support the methodology for this thesis as they assert that, “Methodologically the suggestions of Bygrave (1989) for research into entrepreneurship are supported. That is in an emergent field we were less concerned with sophisticated statistical analyses and theoretical models and more with empirical models … based upon longitudinal field research” (ibid, 132). This suggests similar techniques of analysis in the sense that this research will use a model developed from the work of Watson, Hogarth-Scott and Wilson (1998), which is used to understand the entrepreneur’s behaviour towards SME ownership, success and failure. The method for this thesis analysis attempts to investigate the African SMEs entrepreneurs with similar depth as stated previously.
The three dimensions that this research project used are the business, the entrepreneurial and the personal (individual) dimensions, these are presented in figure 4.1. They formed interdependent strategic components in the construction of the conceptual model for the study of SME creation, success and failure.
126.96.36.199) Variables Relating to the SME (Business Characteristics) Dimension
Entrepreneurial processes focuses on the form managerial behaviour, which are associated with entrepreneurial opportunity-seeking firms. Here the focus of this research is on the policies and actions of entrepreneurs in the SMEs who may or may not be the owner managers of the SMEs, rather than their personal attributes and influences which is another dimension of the study. Entrepreneurial management is regarded as a process in two senses. First the actions that entrepreneur takes in order to create and exploit profit and growth opportunities. Second the managerial styles, which must evolve in the process of pursuing these opportunities, in response to problems that entrepreneur face at certain points in the development of the business (Jalloh and Falola, 2002, p78).
Miller (1983) claimed there is a continual need for innovation and pursuit of new opportunities in companies. Covin and Slevin (1991) describe SMEs success, survival and growth based on product innovation, good planning, good management and opportunity recognition. He added that poor management decisions are the major causes of business failure and small firms have limited resources and little margin for error (Christopher and Erik, 2003).
SMEs should be able to have an edge in innovating although they may not afford well-paid scientists and engineers, support staff, modern facilities and equipment, and extensive research and development. These firms should be bold enough to cope with the risks of innovation and be able to come up with an array of innovative and effective financial management style for success and growth. Although, innovation in SME is a process that involves an enormous amount of uncertainty and chance (Utterback 1996). Owualah (1999) think it is virtually impossible for SMEs to be truly innovative. Many entrepreneurship researchers and practitioners believe that a non-large firm cannot innovate because of the need for stability and risk avoidance. However, research shows that SME entrepreneurs prefer to take moderate, calculated risks (Chalfin, 2000). High-risk projects leave too much to chance and low-risk projects do not provide the needed challenge or enough potential return.
Pursuit of new opportunities or proactiveness can also be described as “bias for action”, particularly in the marketing arena (Carson et al (1995). Sandberg (1992) interprets this to mean that the firm is first in offering new products and services rather than responding to competition. Many use the term proactive to describe the firms practising entrepreneurism (Kinunda-Rutashoya et al., 2001; Zahra, 1993; Covin and Slevin, 1991). Such firms try to improve their processes, products, and staff to increase the customer’s value (Carson et al (1995). Perhaps the most critical antecedent of SMEs action is the categorization of strategic issues into opportunities and threats (Christensen, 2003). As with intentions, opportunities are constructed, not found. (Kinunda-Rutashoya et al., 2001; Mintzberg, 1994). The above discussion addressed innovativeness information gathering, opportunity pursuit and opportunity alertness (proactivness) as important and fundamental characteristics of business which will be used as business dimension variables in this research.
188.8.131.52) Variables Relating to the Entrepreneurial Dimension
This section will present the variables that are relevant to the entrepreneurial dimension. These variables are presented in the conceptual construct of the entrepreneurial dimension, in figure 4.2, in both the internal and external environment of the SMEs.
The seven culture norms described by Stephen and Jonathan (2003) have high relevance for the entrepreneurial dimension. Each of the norm dimensions regulates behaviours that are related to one or more of SMEs stages of the entrepreneurship process addressed earlier. Several of the norms tend to create a climate that fosters the generation and testing of new ideas and thus increases the initiation of success and avoids failure. The seven culture norms are:
- Value for innovation as a practice and as a source of competitive advantage.
- Norms encouraging creativity among SMEs
- Norms encouraging search for innovation opportunities from external source
- Norms supporting information-sharing between entrepreneurs
- Norms that promote tolerance for failure when creative ideas/projects are not successful
- Norms that encourage the open-minded consideration of new ideas and projects
Whereas entrepreneurial posture summarizes beliefs and attitudes of only top management in SMEs who may not necessarily be the creators or owners of the SMEs, the cultural variables summarize beliefs and attitudes of the whole SMEs participants socialized into the SMEs entrepreneurial behaviour as represented in figure 4.2. Finally, norms related to providing aid and support to SMEs participants charged with the implementation of ventures that have successfully passed through the creation stage help to make innovation a successful part of the ongoing activities of the entrepreneurial firm. Each norm is presumed to have a direct, positive relationship to one or more of the phases of entrepreneurial development.
This research suggests that the entrepreneurial dimension in Watson, Hogarth-Scott and Wilson’s (1998) model should also include a linkage connecting opportunity exploitation and development, and innovations for SMEs success, growth and development. This linkage highlights a critical entrepreneurial test for the African SME’s new businesses/products, i.e. whether critical African SMEs stakeholders accept the concept of opportunity exploitation and development, and innovation or not. The link completes a positive feedback loop which shows increases in performance due to successful opportunity exploitation and development, and innovation, whereas performance declines due to unsuccessful opportunity exploitation and development, and innovation tends to weaken entrepreneurial activities in SMEs.
Thus, this study has built on ECA (2001a)) research to recognise an additional array of characteristics of entrepreneurial African SMEs that characterise normal small business and promote entrepreneurial success. These characteristics include:
- Technology (Imported or Adopted)
- Labour and Financial base
- Strategies and plans
- Management style and resources
- Innovation Strategies
- Lifestyle/Consumption patterns
- Organisational characteristics
- Purchase behaviour
- Influx of Foreign Goods/Services
Other relevant entrepreneurial variables are:
- Increased levels of entrepreneurial training permit SMEs managers/creators to propose and test more new ideas.
- Increased availability of financial resources to entrepreneurs facilitates the ‘championing’ of innovative products. Therefore, more new ideas can be initiated and more projects started since resources available to SMEs entrepreneurs are more widely dispersed than if all goes to big organisations.
- The effective source of information facilitates the unfettered exchange of information, which may result in the generation of more new products for more successful running of the SMEs.
- Increased governments support in terms of products consumption, tariffs and taxes especially in the manufacturing sector may increase the commitment of SMEs members to come up with innovation projects, thereby facilitating the growth and survival of SMEs
These relevant variables supporting and affecting the entrepreneurial behaviour in SMEs were used as variables that are related to the entrepreneurial dimension which influence SMEs success, survival and growth.
184.108.40.206) Variables Relating to the Personal Dimension
This section will present the variables that are relevant to the personal (individual) dimension. These variables are presented in the conceptual construct of the entrepreneurial behaviour in the study framework presented in figure 4.2.
Researchers who have investigated the role of individual characteristics in the creation of a new business have focused either on the differences between entrepreneurs and successful business managers or differences between entrepreneurs and the general population (Kinunda-Rutashoya et al., (2001). Freses (2000) concluded that there are few psychological characteristics that distinguish entrepreneurs from practicing business managers but felt that locus of control and achievement motivation are important factors in the decision to start a business. Other recent findings include variables such as energy level, conformity, and need for autonomy (Kalanje, 2002); need for achievement, need for autonomy, dominance, high energy level, and persistence (Manu, 1999); a desire for personal control (Chrisman, Bauerschmidt and Hofer, 1998); and the desire to build something of one’s own (Knight, 1987).
This thesis focuses on a set of individual traits i.e. characteristics of an African entrepreneur. The characteristics are found by looking into the present literature review and the research focus of this thesis and are shown in figure 4.2 and are as below:
- Experience (local and Overseas)
- Socioeconomic background
- Skills and Knowledge
- Personality attributes, traits
- Values, expectations/necessities
- Demographic characteristics
The major reason for selecting these characteristics was to provide a foundation for matching certain personal entrepreneurial characteristics as bases/predictors of SMEs ownership, growth, development and failure. A set of hypotheses that link the individual, the SMEs, and African entrepreneurship could then be tested.
4.1.3) Entrepreneurial Opportunism as a Common Variable
Covin and Slevin (1991) suggested an integrative model that explains the association between a company’s entrepreneurial opportunism and its external environment, strategy, internal factors, and organisational performance. The concept of entrepreneurial opportunism as a philosophical position is important to SMEs entrepreneurship because it captures the propensity of entrepreneurial efforts by SMEs entrepreneurs to provide the resources and business activities that facilitate ‘entrepreneurial opportunism’. Sandberg’s (1992) conceptualisation of entrepreneurial opportunism includes five components: 1) The entrepreneur’s propensity to touch risky businesses; 2) the extent and frequency of product innovation; and 3) the pioneering nature of entrepreneurs tendency to engage in proactive competition with industry rivals 4) the ability of entrepreneurs to practice financial probity and prudence 5) and the ability to shake–up personal interest when taking business decisions.
Entrepreneurial opportunism represents an overall ‘strategic philosophy’ regarding the role and function of entrepreneurial activities like opportunity identification, exploitation and development, and innovation among others in a firm’s grand strategy. Entrepreneurial opportunism is considered to influence firm performance directly and to be influenced by environmental characteristics, the business and mission strategies of the firm, and SMEs variables such as resources and competencies, competition and management styles (Margaret, 2004). In this research project the term ‘entrepreneurial opportunism’ is used as the policy-making ‘broad based strategic concept’ as intended by Sandberg (1992), for the three dimensions within the construct, i.e. the business dimension, the entrepreneurial dimension and the personal (individual) dimension.
In brief, entrepreneurial opportunism represents the strategic mindset of the entrepreneur which expresses intent and preparedness to embrace and support entrepreneurial actions as a means to create viable and sustainable SMEs through competitive advantages. Entrepreneurial opportunism is also closely linked with the entrepreneurial orientation of a firm and is the tactical expression of the entrepreneurial opportunism. Entrepreneurial orientation then refers to the managerial practices and decision-making activities that motivate the entrepreneur with the necessary decisiveness, energy and support to get the SME started and to keep it on track.
4.1.4) Summary of Theoretical Framework
This section developed and presented a holistic, yet structured construct of three interdependent dimensions; the business dimension, the entrepreneurial dimension and the personal (individual) dimension from the internal and external environments of the SMEs, all believed necessary for approaching the phenomenon of SMEs creation, success and failure study as shown in the developed model presented in figure 4.2. The section also introduced the term ‘entrepreneurial opportunism’, which is closely linked to the important term ‘entrepreneurial orientation’. From the overall discussions, it can be concluded that there are many internal and external barriers to SMEs creation, success and survival in Africa that need to be studied. In order to make the project more meaningful and to identify whether there was a predictable pattern for the research questions that were derived from the research hypotheses the following factors were fully exploited as reasons for SMEs creation and ownership in Africa:
- the reasons leading to SME ownership in Africa,
- opportunity search intensity,
- opportunity identification and recognition,
- strategy followed by the SMEs,
- methods of strategy development adopted,
- the degree of innovativeness and,
- the managers and shareholders role (contributions).
Also, an array of factors that were from the three dimensions of the study were used to conduct an empirical and qualitative analysis for the study of success and failure of African SMEs.
4.2) DEVELOPING THE RESEARCH DESIGN
Research is a process of planning, executing and investigating to find the answers required for agiven phenomenon. When conducting a research project, it is important to choose the correct techniques for collecting and analysing data. Techniques must not only be appropriate, but the data must be valid and reliable. Boundaries must be identified, such as problems that could be associated with access or time restrictions. This project is designed systematically not only to understand the increasingly complex African business world but also to achieve reliable answers and information that can be used and presented for all to understand and believe.
While the previous section established a theoretical foundation for this research including the construction of a conceptual study model for SMEs creation, success and failure, this section develops an appropriate research design in consideration of the research hypotheses and objectives, which were stated in Chapter 1. The research design is a plan outlining the methods and procedures for collecting and analysing data to answer a specific research question (Hussey and Hussey, 1997).
This design has five major components. These components are reported and defined below with brief conclusion regarding the researcher’s choice within each component:
- Research Paradigm: A paradigm holds the concept of the enquirer’s basic beliefs about the world. The researcher of this project has come to adhere to the positivist paradigm for the social and management sciences.
- Methodology: The overall approach to the research process, from the theoretical underpinning to collection and data analysis. Based on the positivistic approach, qualitative methods recognise a subject to subject relationship between the researcher and the area of research. This researcher believes that an objective relationship to research, and in particular to the area of this research project, is both possible to obtain and desirable. Additionally, both qualitative and quantitative methods were used to enhance the analytic results from primary data collection.
- Methods: The overall approach to primary data collection. This researcher concludes that the present research project would obtain the most relevant objective data for analysis by using non-participant questionnaire/interview survey techniques and activities.
- Data analysis: The approach used to analyze the collected data. This researcher chose the established small business firm as the unit of analysis along with a cross-sectional approach to the questionnaire survey method. It also use qualitative analytical technique to analyse the data obtained from the face to face interview with successful African entrepreneurs.
- Reporting of the research: The present research has used professional software, ‘SPSS – (Statistical Package for Social Science)’ for the purpose of storing and analysing complex quantitative data, to help ensure and enhance the reliability of this research.
There are choices to be made as to the strategic approach to the research. The options available proposed by Saunders et al (2002) consist of: Experiment, Survey, Case study, grounded theory, ethnography, Action Research, Cross-sectional and longitudinal studies and exploratory, descriptive and explanatory studies. For this study there will only be a consideration of the following due to their usefulness and suitability:
|Experiment||Involves the selection of samples and allocation of these samples to different experimental conditions with an introduction of planned change on one of the variables.
|Survey||Allows the collection of a large amount of data and usually based on a questionnaire and can be of either qualitative or quantitative or a combination of the two. This approach offers great scope for finding of the results the title questions
|Case study||This approach may include questionnaires and interviews. This strategy will be of particular interest if a rich understanding of the context of the research and the processes being enacted are required, and also provides answers to ’why?’ as well as ‘what?’ and ‘how?’
|Ethnography||This is firmly associated with an inductive approach and its purpose is to interpret the social world the research world inhabits in the way in which they interpret it, and is very much connected to qualitative data and observation.|
A questionnaire and interview survey approach was chosen as the methods for this research and reasons for this choice is as explained above and below. Primary research is designed around the key issues as highlighted in developed conceptual study model in figure 4.2 that is using identified variables in the three study dimensions (business, entrepreneurial and personal) and focuses on analysing African entrepreneurs’ success in real life instances. It will also test African entrepreneurs’ perceptions of their barriers and failures, and ways they feel they can be overcome, if at all.
The quantitative aspect of the research was based on African SMEs (more than 10 and less than 200 employees, which is inline with SMEDA – Small and Medium Development Agency of Nigeria definition of SMEs and adopted for this thesis) as it was felt this provides great scope for a real life look at the situation of African small businesses entrepreneurship. It was businesses of this size which need higher entrepreneurial activities in order to develop and succeed.
However, for the qualitative aspect which is the study of the history of successful African entrepreneurs, the approach used was the interview method and no restriction on the type of business is placed, that means it included entrepreneurs that were operating in the small, medium and large African firms. The only barrier was for a successful entrepreneur in large business to be included, he must have started as an entrepreneur in the SME business and must have developed the SME to the large business level.
220.127.116.11) Data Collection Methods
Primary data is information collected by the researcher. There are three main ways of collecting the information.
The first way to collect primary data is by conducting semi-structured and in-depth interviews. The data colleted from this method will be much more reliable and valid than when conducted by unstructured interview method. There are three types of interviews: structured, semi-structured and unstructured. Structured interviews consist of predetermined questions whereas semi-structured interviews consist of a list of themes and questions that the researcher hopes to cover. Unstructured interviews are less formal and a discussion takes place covering a general area in which the researcher is interested. The usefulness and amount of information obtained using semi-structured and in-depth interviews as a way of collecting data will depend on the level of competence of the interviewer and the willingness of the interviewee to provide information.
The second way of collecting primary data is through observation. It involves “the systematic observation, recording, description, analysis and interpretation of people’s behaviour” (Saunders et al., 2002). It is particularly applicable if the research concerns what people do.
The third way of collecting primary data is by using questionnaires. It is one of the most widely used methods and is an efficient way to collect responses from a large sample. They can either be self-administered or interviewer-administered and questions can be open or closed (list, category, ranking, scale, quantity and grid).The five main types are online, postal, delivery and collection, telephone and interview schedule. The design of the questions and the structure of the questionnaire are extremely important as they can have a great affect on the validity, reliability of the data collected and the response rate.
18.104.22.168) Questionnaires and Interviews
The two main methods used to collect the qualitative and quantitative data for this research project were questionnaires and interviews. The techniques for each of these methods are discussed below.
22.214.171.124.1) Questionnaires design and administration
Questionnaires can be considered to be the same as structured interviews, and therefore have the same uses and limitations, although questionnaires are often thought of as something that is filled in by the respondent. There are many different and conflicting definitions of the questionnaire, although deVaus, (1991, cited in Saunders et al., 2002, p. 278) has defined it as a “term to include all techniques of data collection in which each person is asked to respond to the same set of questions in a pre-determined order”. This can therefore include telephone, face-to-face and mail questionnaires. According to Dillman (1978), as cited in Saunders et al., 2002), there are four main types of variable which the data collected by questionnaires can be divided into: attitudes, beliefs, behaviour and attributes.
Questionnaires are used in situations where there are many simple questions to be asked (which would therefore make it unsuitable for interview research). Although questionnaires are cheaper than interviews, they are much less flexible. There are also associated advantages and disadvantages of each type of questionnaire. For example, postal questionnaires are relatively cheap and quick, but there is a very low response rate compared to phone and face-to-face surveys (Creswell, 2003).
There are many different types of questions. Open questions may be used particularly in unstructured questionnaires, in which there are no predetermined answers, simply a blank space for a written statement. They can be used for obtaining people’s opinions and perspectives in more detail than a closed question can, and are particularly good when the researcher is unsure how the respondent will react and therefore could not offer predetermined answers (Saunders et al., 2002). However, they are also limitations, as people may have difficulties completing them and they obviously take much longer to complete than closed questions, and time is an extremely important issue. The results from open questions are obviously more complex to analyse (Easterby-Smith et al., 2002) as they generally provide qualitative data.
Closed questions, on the other hand, are quick and easy to both complete and analyse (Easterby-Smith et al., 2002), although the data will not be as in depth and therefore may ultimately be less useful. It is also possible that respondents will give them less thought. In the case of scales in which respondents are given a statement and asked how strongly they agree or disagree using a ranked scale of agreement, there may be the problem that people may constantly choose an average value if they cannot decide. Closed questions may also present the issue of bias to the analysis of the questionnaire.
One of the major limitations with questionnaires is that it is actually very difficult to construct a clear and practical questionnaire that will be as simple as possible to complete. Considerations must be given to clarity, layout, use of language, and adequate completion instructions, not to mention the actual content of the questions. It is usually advisable to test the questionnaire thoroughly before use, as studies have shown that poorly designed questions can have an adverse affect on questionnaire results (Sudman and Bradburn (1974) and Fowler and Mangione (1990) cited in Fowler, 1995). There is the problem of deciding what makes a good question, however, according to Fowler (1995, p.344) “A good question is one that produces answers that are reliable and valid measures of something we want to describe”. Most questionnaires can be classified as either descriptive or explanatory.
“Analytic or explanatory surveys attempt to test a theory by taking the logic of the experiment out of the laboratory and into the field… a descriptive survey is concerned primarily with addressing the particular characteristics of a specific population of subjects” (Margaret 2004, p.133). The greatest use of questionnaires is in the survey form. Deciding which type a survey is, will ultimately determine the design of the interview or questionnaire. As people are asked the same set of questions it allows for an efficient way of collecting responses from a large sample and then can be used in a quantitative analysis. Fowler (1995) suggests it is much harder to produce a good questionnaire than is initially thought. It needs to be designed to enable the researcher every opportunity to gather relevant data to help answer the research question. The questionnaire needs to be correct first time as revisiting someone with an adjusted questionnaire is highly unlikely.
The questionnaire needs to have;
- Carefully designed individual questions
- Clear layout of the questionnaire form
- Lucid explanation of the purpose of the questionnaire
- Pilot testing.
It is important for anyone undertaking a research project to remember that “before adopting any method of data collection, it helps to be clear about the overall objectives of the research” (Easterby-Smith et al, 2002, p.85), emphasising that the methods of data collection must above all be the most appropriate methods for the aims and objectives of the research. Ultimately, a ‘multi-method’ approach may be the most successful way of gaining sufficient and accurate data (Saunders et al, 2002), as can be seen by examining the case study method.
For this research, the most structured and respected form of this data collection would be through a questionnaire and in specific, with the use of Likert Scales. A Likert scale allows for the respondent to answer a statement with their degree or agreement or disagreement to it. 0 on the scale means a strong disagreement to the statement and a response of scale 5 indicates a strong agreement to the statement.
This can be seen below:
Strongly disagree strongly agree
This scale does however present its limitations. With the nature of the Likert scale, respondents are likely to adopt middle group as opposed to outright agreement or disagreement. However some of the questions indicate a choice between 0 and 10 and in few cases some questions require a yes or no answer. The questions were designed to collect the data required to make sufficient and respected conclusions in relation to the subject. Each question was designed carefully and specifically to test the key findings of the literature review.
To ensure an acceptable quality and avoid ambiguity, the questionnaire was prepared in a language that the respondent would understand easily or will be made to understand through language interpreter (for this research, languages spoken in the area covered by this project are English, Arabic and French). It was also prepared in such a manner that it would not lead the respondents to a particular answer (Fowler, 1995). Moreover, the questionnaire was based largely on forced-choice questions (though there were a small number of open-ended questions). A closed or forced-choice question is one in which a number of alternative answers are provided from which respondents are to select one or more. (See Appendix A for a copy of the survey instrument.)
126.96.36.199.2) Sample questionnaire
Pilot testing of the questionnaire took the form of survey of 10 African SMEs entrepreneurs in Nigeria and Egypt. They collectively failed to see any major way in which it could have been improved. On reflection, this could have been because it is perfect as the questions are simple and unambiguous or could have been a lack of understanding of the subject or my objectives of the research. The questionnaire remained unchanged for the research. However, the questionnaires were worded and terminology used in such a way that meant their completion could be done without guidance. This needed to be the case as is shown below, as the researcher wasn’t always present at the time of completion of the questionnaires although in over 75%, the researcher was physically present.
There were three main approaches that have been identified in carrying out the questionnaire and these have been highlighted in figure 4.5.
Approaches to Administering Questionnaire
|Bias (from Interviewer)||3||2||1|
|Control over collection||2||1||3|
|Depth of questioning||1||2||3|
|Follow up ability||2||1||3|
|Hard to recall data||2||3||1|
|Rapport with respondent||1||2||3|
|Speed of obtaining results||2||1||3|
Key: 1 = Good, 2 = acceptable, 3 = Bad.
Source: Taken from Frey (1983)
Although Frey (1983) highlights that the best approach is to either carryout the questionnaire in person or over the telephone, it was decided to administer the questionnaires face to face but in some cases it was decided to leave the questionnaires with the entrepreneurs to complete in their own time. The reason behind leaving the questionnaires was in most cases at the instance of the person filling the questionnaire and because the researcher didn’t want to disturb or interrupt any work load the entrepreneurs may have had as always claimed by them.
188.8.131.52.3) Sample Survey Population
The questionnaires were given out to creators/owners or people that are deemed fit (as indicated in the survey questionnaire). The questionnaires were administered on entrepreneurs that have established or purchased SMEs employing more than 10 and less than 200 employees, which as the SMEDA definition classified as SMEs in Africa. The dates the questionnaires were administered were between 25th April 2006, and 30th October 2006.
A stratified random sample of 1,200 SMEs was drawn from a certified list of business names provided by small and medium development agencies of each of the five countries (Nigeria, Ghana, Mali, Egypt and Zambia) because of the vast territorial area, language barriers and sample size, this study employed field assistants for a period of one month in each country, including an Arabic translator in Egypt and a French translator in Mali. The structured questionnaire was administered using a single key respondent (i.e. a creator and/or principal owner) in each of the selected SMEs. During the seven month of data collection period, 19 responses were eliminated as they indicated the business was no longer an independent trading entity. A further 23 respondents were not creators / or the principal owners of the business, and 413 were regarded as non-respondents.
After a series of telephone notices (to fix a date and time for the face to face administration of the questionnaires) and remainders (for questionnaires that were left with the SMEs under the study), 766 valid questionnaires were obtained from a valid sample of 1,200, a 63.90% valid response rate. This response rate compares favourably with similar studies, which generally have much shorter and less detailed research instruments (for example Chames’s (1999) study of micro-enterprises in West Africa recoded a response rate of 55.7% and McCormick’s (1997) study of Nairobi’s small manufacturers recorded 68.0% response rate). 61 of these Respondents who reported they had only inherited an established business or did not meet the criteria distinguishing success or failure and were excluded from any further analysis. Therefore the final sample consisted of 705 African entrepreneurs. A further breakdown of the valid responses shows 246 came from Nigeria, 131 from Ghana, 122 from Mali, 107 from Zambia and 99 from Egypt. Also, the 705 respondents consisted of 543 successful and 162 failed businesses. The scale used in classifying a business as a successful or a failed business is defined in the success versus failure data analysis section of chapter 5.
The interview can be defined as “a purposeful discussion between two or more people” (Kahn and Cannell 1957, in Saunders, et al 2002, p242). This is perhaps too simplistic a definition however, as there are many different types of interviews, each with its associated uses and limitations. The main feature of the research interview is the generally high level of interaction between the researcher and the interviewee or respondent, although this depends to some extent on the structure of the interview.
Interviews can be used to obtain detailed data regarding personal experiences, opinions and attitudes, which may be of a sensitive or confidential matter. They allow scope to ask for elaboration and also, in the case of semi-structured interviews, to be flexible with questions asked and issues raised.
Interviews are often considered to be one of the most successful and useful data collection techniques, but they are not as simple as they seem, and have many associated limitations and complexities (Ghauri and Gronhaug, 2005; Easterby-Smith et al., 2002). For example, it is imperative that the interviewer fully understands the research topic and aims of the research, and is familiar with the respondent so they can ask appropriate questions and gather as much useful data as possible. Furthermore, interviews are extremely time-consuming. Not only can each in-depth interview last several hours in itself, after the interview has taken place, the transcription of any recordings must be done, which can be a monotonous and inevitably time-consuming task.
The issue of confidentiality is minimised in interview situations, as people generally feel more comfortable with personal interaction than disclosing sensitive information on a questionnaire, for example, although some people may still object to being recorded. It is extremely important for the researcher to be as objective as possible, which may be very difficult, particularly in semi-structured interviews in which they are expected to maintain the momentum and steer the conversation without using leading or biased questions.
184.108.40.206.5) Interview Sample population
Interview method using narrative research approach was used to critically study the profile of 16 African successful entrepreneurs (4 from Nigeria and 3 each from Ghana, Mali, Eggpt and Zambia) for the purpose of understanding their route to success which is part of the objectives of this research as mentioned in chapter 1. In addition to the interview method, the use of secondary data was employed in analysing these great African successful entrepreneurs. The interview was conducted in the business location of each of the selected entrepreneur. The 16 successful entrepreneurs were selected from the countries that questionnaire survey was also carried out and these countries are Nigeria, Ghana, Mali, Zambia and Egypt. It was also conducted at the same period the survey was done. The bases for the selection of the successful entrepreneurs are based on the answers provides in the questionnaires and information gathered from secondary data relating to the business empire of the selected entrepreneurs.
To start each interview with the selected successful African entrepreneur, the ‘informer interview’ technique (Lieblich and Rivka, 1998) was used. This qualitative interview technique is relevant when the researcher is interested in a phenomenon that has already happened or when the researcher her/himself does not have the opportunity to study the phenomenon. This is a particularly good technique in the beginning of research projects when the researcher is not clear on what questions to ask (Lieblich and and Rivka, 1998).
The open interview is another example of a qualitative interview, which is sometimes called the qualitative research interview. This type of interview is used to develop a deeper understanding of a person’s behaviour, actions, motives and personality (Lieblich and Rivka, 1998 p. 28). The main purpose of this technique is to collect descriptions of the interviewee’s own world so that the researcher can interpret the described phenomena better. It is a difficult technique but very valuable. The strength of such interviews is the possibility they offer for discovery. Many times in the present research project, the open interview approach provided unexpected comments from the interviewees, which in turn provided valuable input for further discussions in other relevant directions. Knowledge many times is wider and deeper that can be revealed by the open-interview technique.
The semi-structured interview is used in the same situations as the open interview technique. The difference is that the researcher has developed a greater understanding of the area studied. There were usually some areas that the researcher knew he wanted to take up for discussion with interviewees, and here an interview guide was used to guide the interviews. The results of the interview enabled the researcher to comprehensively write the profiles of the selected entrepreneurs using narrative research technique. Narrative research technique is a form of qualitative research approach that centred on study of human experience (Clandinin and Connelly, 2004). It is a presentation of coherent series of actions which the narrator connects with others. In narrative research, researchers described the life of individuals, gather and write stories about people’s lives, and narrate individual experience. Typical distinctions that can be made about narrative research as a form of qualitative research are (Clandinin and Connelly, 2004):
- Narrative Research specifically focuses on study of single individuals.
- It focuses on gathering data through collection of individual’s stories.
- It engages in the reporting of individual experience.
- It thoroughly discusses the meaning of those experiences for the individual.
Narrative research is carried out when the individual is willing to tell his individual stories and experiences. It also brings the researcher closer to the participant, although he may not influence the individual story (Lieblich and Rivka 1998, p83). It attempts to put matters in chronological order. Narrative Research first overview came about in 1990 when Clandinin and Connelly cited in Clandinin and Connelly (2004) took an overview of narrative research in the field of education and trends they discovered that influenced the development of narrative research are:-
- Increased emphasis on teacher reflection
- Emphasis placed on teacher knowledge
- Attempt to bring teachers voices to the forefront
In narrative research design, the researcher focusses on questions that relate or recall the individual story which may be in the form of biography and autobiography among others. It is also expected that the researcher should focuses on recording the life history and personal experience of the participant without influencing the participant, although he can ask questions targeted at specific answers.
Major characteristics of narrative research designs are:-
- Individual experiences
- Chronology of the experiences
- Collecting Individual stories
- Coding of themes
- Context of setting
- Collaboration with participants.
The narrative approach was used to interview and analyse the 16 African successful entrepreneurs out of the 20 targeted. The routes to success followed by each of them are reported in chapter 7.
220.127.116.11) Secondary Data Collection Sources
Secondary data is “marketing research information available to researchers that they have not gathered themselves” (Oxford Business Dictionary, 1996). It can be ‘raw’ (unprocessed) or ‘compiled’ (summarised). Once the data has been collected, it must be evaluated for its suitability. This will include analysing its reliability and measuring any bias it may have. However, it is better to have some data which is not completely reliable or without bias than having no data at all. There are three different types of secondary data: Documentary, Survey-Based and Multiple-Source. There are many advantages of secondary data. It may have fewer requirements, it is unobtrusive, it may allow for longitudinal studies, it can provide comparative and contextual data, it can result in unforeseen discoveries and it is permanent. However, it may have been collected for a purpose which does not match your need, access to it may be difficult or costly, aggregations and definitions may be unsuitable and the initial purpose may affect how the data is presented.
For this research, secondary data was collected as a complement and/or support to the primary field-data collected. However, no secondary data was engaged in the empirical (quantitative) analysis of the stratified sample of 705 Reponses recorded on the questionnaire survey but used in the narrative research analysis of the 20 great African successful entrepreneurs. The secondary data that was collected was both qualitative and quantitative including annual reports, organisational charts, financial plans, business plans, operational plans, marketing material and other material of strategic interest. These data were collected as part of the process of collecting primary data, and they validate and complement the data collected from observing and interviewing. The main purpose of collecting secondary data was to develop and increase the researchers’ understanding of the situation studied, which in turn helped this researcher to develop a deeper contextual understanding of the research object within the particular research situation.
18.104.22.168) The Researcher’s Role
A researcher can take at least two types of roles. These can be described as 1) participant observation or even action research if the objective also is to implement change within the research setting, and 2) outside observation and/or non-participant observation. This thesis uses the latter approach. Neither of these two approaches can be considered as objective since the researcher in the analysis involves his or her own conceptual understanding, references and prior understanding of the area of research. However, this researcher believes that conducting a thorough non-participant survey will enhance the level of objectivity.
The non-participant questionnaire surveyor technique preserves some level of interference from the personnel in the studied firm. Field operatives will view the non-participant researcher not as one of themselves, but as an outsider who is no threat to them.
The merit of this approach is that the researcher is seen as not having a direct personal stake in various interpretations and outcomes, and thus personnel will often be relatively frank in expressing their views in the questionnaire without the influence of the researcher even though he is physically present at the time of filling the questionnaire, his presence will only facilitate the speed of filling the questionnaire and its accuracy of the answers.
There are disadvantages in participant questionnaire/interview surveyor or in participant observation, e.g. the involved researcher will be perceived as having a direct personal stake in various views and activities, and other personnel may be more guarded in their expressed answers or interpretations as a consequence. Unless participant observers or action researchers hide their research motives, which could be considered an unethical position (Ghauri and Gronhaug, 2005). A final problem with the role of the involved researcher is the extreme difficulty of reporting the part he played in the various matters under consideration. Self-reporting faces the twin dangers of over-modesty and self-aggrandisement, and it is particularly difficult to steer a middle path between these two extremes (Robson, 2002).
The above reasoning led this researcher to the conclusion that the present research project would obtain the most relevant objective data for analysis by using non-participant questionnaire/interview survey techniques and activities.
4.2.3) Data Analysis
This subsection will address and present the framework used for the data analysis discussed in Chapter 5, 6 and 7.
The main challenge to qualitative and quantitative data analysis is that there is “no clear and accepted set of conventions for analysis corresponding to those observed with qualitative or quantitative data” (Robson, 2002, p. 370).
One approach is to quantify the collected data, either formally or informally. This researcher believes, however, that such an approach is undesirable and that by turning both the qualitative and quantitative data into numerical data, the researcher would not be stretched far beyond his positivist paradigm position.
The entire collection of field data was generated from the chosen countries of analysis. The research questions and the data collected together enabled the researcher to determine at what point data has been collected from sufficient respondents to enable appropriate analysis. Most researchers are vague when it comes to suggesting how many actual respondents to study (Fowler, 1995). As discussed earlier, the research performed in this research area does not show consistency regarding industry type, or level of profitability of the SME. It is therefore purely on logic and apparent relevance that the following characteristics of the unit of analysis were stated.
22.214.171.124) Timeframe: Longitudinally vs. Cross-sectionally
Ghauri and Gronhaug (2005) claim that research can be conducted either cross-sectionally or longitudinally. Cross-sectional studies collect the data at a single point in time, making the analysis valid to that particular point in time. In a longitudinal study, field data are collected at different times or from different interviewees. Longitudinal studies are often used when the researcher wants to study changes in the research setting. This researcher rejects this approach, recognising the difficulty of eliminating the effect of external factors if the study is conducted over time (Easterby-Smith, et al 2002). This researcher also believes that it is not necessary to observe changes in the research setting to answer the research Hypotheses stated for this research. Therefore, a cross- sectional approach to the questionnaire survey method stated above was chosen.
As a conclusion to the discussion concerning Data analysis, the inclusion of Nigeria, Ghana, Mali, Egypt and Zambia in the studies in this research addressed all of the established priorities mapped out. Based on this fulfilment of priorities, it can be argued that the bases for Data analysis was justifiably chosen and included based on relevance to the stated purpose of this research.
4.2.4) Reporting of the Research
The issue of how to report fieldwork is important in all research, but it can be argued that it is particularly critical in positivist questionnaire survey research. Positive researchers do not always readers that they are reporting their interpretations of other people’s interpretations instead, they claim to report facts. However, this research uses positivist approach and the credibility of the research and for the researcher has already been described in detail in this section.
126.96.36.199) Validity and Reliability
Validity and reliability are essential elements to consider in research design, as they refer to the degree of confidence researchers and academics can have in the results of a study (Ghauri and Gronhaug, 2005; Margaret, 2004). Validity refers to minimising research errors, ensuring the accuracy of the results. Validity is concerned with the extent to which the research findings accurately represent what is happening in a situation; in other words, whether the data collected are a true picture of what is being studied (Hussey and Hussey 1997). Reliability refers to achieving consistency in the results of the research, if the research has to be repeated (Owojori, 2005). This is a more important issue in a positivistic study (Hussey and Hussey, 1997).
The two elements, validity and reliability, are addressed separately below with regard to the research design.
Validity is primarily concerned with whether or not the findings are about what they appear to be about and whether or not the research accurately describes the subject. Robson (2002) also suggests threats to the validity; History, Testing, Instrumentation, Mortality, maturation and Ambiguity about casual direction. Three types of validity need to be considered in survey study research: internal validity, external validity and construct validity (Owojori, 2005 p 137).
Inner validity or internal validity concerns the question of whether the findings are correct in relation to the reality (Fowler, 1995). Is the researcher actually studying what s/he believes s/he is studying?
Margaret (2004) draws the conclusion that there is no objective or universal way of guaranteeing validity; there are only interpretations of it. To ensure internal validity of this research, a five-country survey study with a holistic orientation has been kept in focus. A holistic understanding will enhance the level of internal validity (Kiggundu, Abbas and Charies, 1998). As with most survey studies (Fowler, 1995) the internal validity for the conclusions stated in Chapter 5 can be regarded as high.
External validity or outer validity is concerned with the extent to which the results from one study are generalisable to other situations. Since the aim of this research has been to put forward a hypothesis that is generalisable to other settings or populations, the demand for high external validity is necessary (Fowler, 1995).
The aim of this research has been to achieve enough external validity so that the outcome of the research will be accepted and will support further within this young area of research. Since this research has used a five-country survey study with a cross-analysis approach, the external validity is high. At the same time, this researcher acknowledges and accepts that both quantitative and qualitative approaches set some limits on the methodology, expecting validity to be as high as would be expected with a positivistic approach.
Construct validity refers to ensuring that suitable operational measures are used for the concepts studied (Ghauri and Gronhaug, 2005). This is an important aspect of positivistic studies, and not as relevant in interpretative studies. However, applying the combination of deductive research and SPSS software (Statistical Package for Social Science)’ in an iteractive process has ensured an acceptable level of construct validity.
Reliability is generally not difficult to achieve in survey-study research, as it is highly likely that, if the research were to be repeated, the researcher would be confronted with the same events, individuals and groups (Owojori, 2005). Reliability implies that or can be assessed by posing the following questions proposed by Easterby-Smith et al (2002:41):
- Will the measure yield the same results on different occasions?
- Will similar observations be made by different researchers on different occasions?
Robson (2002) suggests that there are 4 threats to reliability; subject error, subject bias, observer error and observer bias all of which should be considered when designing the research.
There are several techniques to ensure reliability in survey-study e.g. verifying data interpretations, triangulation and consistency checking (Owojori, 2005 p 164). Another way to enhance the level of reliability in survey research is to establish a survey-study database. This allows access to others, enabling others to independently check interpretations and findings (Ghauri and Gronhaug, 2005). The present research has used professional software, ‘SPSS – (Statistical Package for Social Science)’ for the purpose of storing and analysing complex quantitative data, to help ensure and enhance the reliability of this research.
Stating the underpinning assumptions that the researcher has and that s/he states the theoretical perspectives chosen will also enhance reliability. The researcher should in detail describe how the study was conducted and how the conclusions were drawn on the basis of the available information (Fowler, 1995). This researcher has endorsed these suggestions.
The section will continue with a discussion of the ethical considerations and conclude with a summary.
188.8.131.52) Ethical Considerations
Since ethical issues are an important consideration for any research, these concerns were closely adhered to as set out by the research committee of Portsmouth Business School of Portsmouth University. “In the context of research, ethics refers to the appropriateness of your behaviour in relation to the rights of those who become the subject of your work, or are affected by it”. (Saunders et al 2002). Ethical issues must consider issues such as plagiarism, honesty of the data and the privacy of the subject.
Whilst designing and conducting the research the respondents must be kept informed of the aims of the research as well as ensuring their confidentiality at all times even in the presentation of findings. Only respondents that have given their consent will be approached throughout this research
Finally, on deciding the approach to use it is important to consider views expressed by Margaret (2004) who argues that “needs, interests and preferences (of the researcher) …are typically overlooked but are central to the progress of fieldwork”.
4.3) SUMMARY OF RESEARCH METHODOLOGY CHAPTER
This section has detailed the comprehensive research design that was developed to carry out this research. The research design is summarised below.
This researcher chose to take a positivist/deductive approach, with a holistic understanding for the research area in focus. The purpose of the research was to draw general conclusions and empirically test a priori derived hypotheses, with the view of gaining a deeper understanding of African SMEs entrepreneurship.
In the positivistic view of this researcher, a positivistic analysis would be most rational for this research and therefore, this section concluded that a positivistic approach is suitable and relevant for this researcher in addressing the chosen area of research. As to ontological assumptions, this researcher believes that the reality is subjective and multiple as seen by participants in a study, i.e. that the world/reality is socially constructed and can be truly understood only by examining the perceptions of the human actors. The researcher therefore believes that the world is created by individuals through language and actions based upon the pre existing values and norms of those individuals or group of individuals.
The researcher has therefore positioned himself closest to the position of a subjective idealist. Reflecting on the epistemological assumptions of the researcher, the researcher took a positivistic stance acknowledging that the researcher will interact with and affect (leave ‘footprints’) on what is being studied, that facts and values are intertwined, and that both are involved in the creation of scientific knowledge. The researcher also believes that the researcher’s non-involvement with what is being researched is possible and although the researcher carries values and opinions, the researcher can be unbiased towards what is being studied. At the same time the researcher believes that being unbiased is scientifically motivated and necessary in order to judge what facts are important for getting what fact/s are relevant to the phenomena that is being explored. The researcher believes and accepts that human interests and objectives drive science. The researcher also acknowledges that for a researcher, obtaining an objective relationship to the research and the area of study is possible, and is also desirable.
This research focused on collecting primarily qualitative and quantitative data through questionnaires and interviews collection techniques as suggested by Kiggundu, Abbas and Charies (1998). Considering this study’s stated focus, an explanatory approach was chosen and applied. The explanatory study was performed with the aim of including managerial implications, recommendations and suggestions (to be discussed in the last Chapter 7), which made the research somewhat deductive. The collected field data was treated deductively with the researcher using insights from the available literature and research when performing the analysis. A combination of deductive research and SPSS software was used in an iterative process.
Since the researcher believes that the phenomenon of African SMEs entrepreneurship is best observed and analysed within its natural environment i.e. from actually operational African firms, a survey-research approach conducting a cross-sectional study approach was used. Within the survey, a non-participative questionnaire technique was applied to collect the field data.
The method chosen for analysis was statistical analysis of the generated data using SPSS (Statistical Package for Social Science) software. An analysis of the frequency of responses to the questionnaire and the mean of individual variables was undertaken in order to describe and understand the sample. In addition to the statistical analysis, a qualitative analytic method for identifying themes among the successful entrepreneurs (16 participants) responses was also used.
The comprehensive research design developed in this section has produced results showing an acceptable level of reliability and external validity relevant for this type of research design. The internal validity is as high as in the research projects using a survey-research approach.
The next three chapters, Chapters 5, 6 and 7, presents the field data collected using the research design developed and presented in this section.