Chapter 1 Introduction
3.5 Data collection and analysis
3.5.1 Data collection methods
3.5.3 Primary data
3.5.3.1 Population and sample selection process 3.5.3.2 Survey questionnaire design
3.5.3.3 Data analysis
3.5.3.4 Analysis and interpretation of data 3.6 Ethical considerations
3.7 Strengths and weaknesses identified in the chosen methodology 3.8 Conclusions
(Source: developed for this study)
3.2 Overview of theoretical framework
Through observation and the gathering of facts from various situations or business operations, information can be generalised as a concept and used to explain some characteristics of
certain phenomena. For more in-depth study, this involves the underlying theories concerned with verifiable concepts and empirical evidence (Ray 2008).
Theory as defined by Zikmund (2003, p. 41) is “a coherent set of general propositions, used as principles of explanation of the apparent relationship of certain phenomena”. Neuman (2006, p. 50) further defines social theory as “a system of interconnected ideas that condenses and organizes knowledge about social world”. Therefore, theories provide a framework that can guide business strategy by understanding operational insights useful for predicting future behavioural patterns.
Theory frames how researchers look at and think about a topic and give meaning to concepts, provide basic assumptions, direct researchers to important questions, and suggest ways to make sense of data (Neuman 2006).
As suggested by Neuman (2006), a theory comprises three parts: concepts, relationships and scope. Neuman (2006) further explains that concepts are the building blocks of theory, and are divided into symbols and definitions. Social theory contains concepts, a relationship among concepts, and a causal mechanism, or reason for the relationship. Scope of theory allows clearer communication to an audience.
Scientific method techniques are used to analyse empirical evidence. There are seven stages in an analytical process of scientific theory building: assessment of relevant existing knowledge of a phenomenon, formulation of concepts and propositions, statements of hypotheses, design of research to test hypotheses, acquisition of meaningful empirical data, analysis and evaluation of data, proposal of an explanation of the phenomenon and statement of new problems raised by research (Zikmund 2003).
Building blocks of scientific inquiry include the process of initially observing phenomena identifying the problem; identification of problems; constructing a theory to explain what might be happening; developing hypotheses; determining aspects of research design;
Sekaran (2003) goes on to suggest a relationship between a literature survey and theoretical framework is that literature surveys provide a solid foundation for developing the theoretical framework. That is, literature surveys identify variables that may be significant, as determined by previous research findings.
Deductive and inductive approaches are applied in building and testing of theory from two directions (Neuman 2006):
Deductive approaches start with an abstract, logical relationship between concepts and move toward empirical evidence, while inductive approaches begin with observations and move toward abstract generalisations and ideas. Preliminary relationships are identified after refining concepts from observations and developed empirical generalisations (Ray 2008).
3.3 Research design and procedures
Once the objectives of the research have been defined, the research design can be planned. One definition of research is “the process of systematically obtaining accurate answers to significant and pertinent questions by the use of the scientific method of gathering and interpreting information” (Clover & Balsley 1984, p. 1). As the outcome of any study is influenced by the choice of research design, determining the characteristics is a significant factor in the quality of the conclusion (Miller & Salkind 2002).
3.3.1 Research design alternatives
As there are a variety of different research approaches, it is helpful to categorise the different types of research. This thesis is concerned with the types of research applied in business. Business research can be classified on the basis of either technique or function (Zikmund 2003), and business research based on the purpose or function of the research can be further classified into exploratory, descriptive, or causal research.
1) Exploratory research is “initial research conducted to clarify and define the nature of a problem” (Zikmund 2003, p. 54). Exploratory studies are a “valuable means of finding out what is happening and gaining insights to assess phenomena in a new light” (Saunders et al. 2006, p. 96). This research does not look for conclusive evidence to determine a course of action.
2) Descriptive research is “designed to describe the characteristics of a population or phenomena” (Zikmund 2003, p. 55). The object is ‘to portray an accurate profile of persons, events and situations’ (Saunders et al. 2003, p. 97). This is often carried out to describe characteristics of groups within an organisation or community (Sekaran 2003). Descriptive research is less ambiguous and is carried out against a backdrop of prior knowledge of the central problem (McMurray et al. 2006) seeking to determine answers to who, what, when, where and how questions (Sekaran 2003).
3) Causal or explanatory research is conducted to establish that an activity occurs as a direct consequence of a particular activity (Zikmund 2003), and is conducted to identify the cause-and-effect relationships between variables, where the research problem has already been narrowly defined (Saunders 2003; Sekaran 2003), or to establish differences between groups, or the independence of two or more factors in a situation (Sekaran 2003).
3.3.2 Research methodology justification
A research paradigm is a framework or set of basic beliefs the researcher uses to understand the nature of reality in order to identify the relationship between variables, and to specify appropriate methods for conducting research (Collis & Hussey 2003; Clover & Balsley 1984; Zikmund 2003).
Two widely-accepted research paradigms are quantitative (hypothetico, deductive, positivistic) research; and qualitative (constructivist, inductive, phenomenological or interpretivist) research (Collis & Hussey 2003). Table 3.1 contrasts the main differences between these two research paradigms in terms of their assumptions of reality, research purpose, researcher role, methodology, data collection and data analysis techniques.
Table 3.1 Characteristics of quantitative and qualitative research paradigms
Features Quantitative paradigm Qualitative paradigm
Assumptions of reality Objective, real, independent of the observer, single, characterised by natural laws
Subjective, socially constructed, multiple, arising out of social interaction
Purpose Generalisability Contextualisation
Researcher role Objective, distant, non-interactive, value & bias free, no influence on research outcomes & data collected and analysed
Subjective, up close, personal values, beliefs & attitudes biasing the way of collecting and analysing data
Methodology Hypothetico-deductive Inductive/interpretive
Data collection Reduction/aggregation of data to
numbers Capture of lived experience of informants
Data analysis Falsification of null hypotheses with
statistical tests Identification of recurring themes & patterns to search for meaning (Source: adapted from McMurray et al. 2006)
Two principal approaches used in gathering information to provide insights into research questions are considered in Table 3.2
Table 3.2 Reasons for selecting quantitative or qualitative research paradigm
Criteria Quantitative paradigm Qualitative paradigm
Researcher’s world
view A researcher’s comfort with the ontological, epistemological, axiological, rhetorical and methodological
assumptions of the quantitative paradigms
A researcher’s comfort with the
ontological, epistemological, axiological, rhetorical and methodological
assumptions of the qualitative paradigms Training & experience
of the researcher Technical writing skills, computer statistical skills, library skills Literary writing skills, computer text analysis skills, library skills Researcher’s
psychological attributes
Comfort with rules and guidelines for conducting research; low tolerance for ambiguity; time for a study of short duration
Comfort with lack of specific rules and procedures for conducting research; high tolerance for ambiguity; time for lengthy study
Nature of problem Previously studied by others so that body of knowledge exists; known variables, existing theories
Exploratory research; variables
unknown; context important; may lack theory base for study
Audience for the study Individuals accustomed to supportive of
quantitative studies Individuals accustomed to supportive of qualitative studies (Source: Creswell. 2003, p. 9)
Quantitative research “emphasises the role of measurement and observation and is associated with the collection and use of numerical data” (McMurray et al. 2006, p. 69), and proceeds
Based on the information contained in Tables 3.1 and 3.2, this study has adopted a positivistic/hypothetico, deductive or quantitative approach to the research objectives as stated at 2.4.1 (page 39). Neuman (2003) states that variables and their relationships are the central idea in quantitative research, and that quantitative researchers primarily follow a deductive route moving from abstract ideas or variables to specific data collection techniques to precise numerical information, providing an empirical representation of the abstract ideas produced by the adopted techniques (Newman 2003).
This study has adopted a deductive approach to assess the main factors involved in the outsourcing decision as considered by accounting service providers and small businesses in Australia. It sets out to identify relationships between the variables in the theoretical model that were formulated on the findings of the relevant existing literature review.
3.3.3 Research technique justification
Emory (1991) classified business research based on technique into two types: observation and surveys. However, Zikmund (2003) expands this classification into four basic types: surveys, experiments, observation and secondary data studies.
1) The use of surveys is a research technique in which information is gathered from a sample of people through the use of a questionnaire (Zikmund 2003), and is one of the most popular research techniques used in quantitative business research (McMurray et al. 2006). “In using surveys the researcher systematically asks a large number of people the same questions and then records their answers” (Neuman 2006, p. 43).
2) Experimental research holds the greatest potential for establishing cause-and-effect relationships between variables (McMurray et al. 2006). The use of experimentation allows investigation of changes in one variable while manipulating other variables under controlled conditions (Zikmund 2003). This is supported by Neuman (2006, p. 41), who states that experimental research is “research in which the researcher manipulates conditions for some research participants but not others, then compares group responses to see whether it made a difference.”
3) Observation allows the researcher to observe and record the behaviour of others without relying on reports from respondents (Zikmund 2003). Observation may be unstructured, where a research question has not yet been formalised and the researcher may be aiming to develop a theory of what is going on in a certain situation, whereas the structured approach may be more suitable when the researcher is seeking to test hypotheses (McMurray et al. 2006).
4) Secondary data study is using “data previously collected and assembled for some project other than the one at hand” (Zikmund 2003, p. 63).
Surveys may be classified by the communication medium used: mail and self-administered questionnaires; telephone interviews; face-to-face interviews; and web surveys (Neuman 2006).
1) Mail survey and self-administered questionnaires are distributed directly or mailed to respondents. This survey method is cheaper than other methods, and allows the survey to be distributed to a wide geographical area. This method is also very effective, and can provide strong response rates.
2) The telephone interview is a survey method in which respondents are contacted by telephone to gather responses to survey questions. Although this method is quite expensive, it is flexible and represents better value than the face-to-face interview while providing many of the same benefits (Neuman 2006).
3) Face-to-face interviews are direct communications with respondents wherein interviewers ask respondents questions in face-to-face situations. This method can provide the highest response rate, although it is very expensive to administer (Zikmund 2003).
4) Web surveys use the internet as the communication medium. They are very fast and inexpensive, and allow for a flexible design of the survey with visual, audio or video capabilities. The main advantages of this method are the relatively short time required for implementation, and flexibility of design (Manning & Munro 2006).
Table 3.3 provides a comparison of features associated with the different survey methods
Features Mail
questionnaire Telephone interview Face-to-face interview Web survey Administrative issues
Cost Cheap Moderate Expensive Cheapest
Speed Slowest Fast Slow to moderate Fastest
Length (number of questions) Moderate Short Longest Moderate
Response rate Lowest Moderate Highest Moderate
Research control
Probes possible No Yes Yes No
Specific respondent No Yes Yes No
Question sequence No Yes Yes Yes
Only one respondent No Yes Yes No
Visual observation No No Yes Yes
Success with different questions
Visual aids Limited None Yes Yes
Open-ended questions Limited Limited Yes Yes
Contingency questions Limited Yes Yes Yes
Complex questions Limited Limited Yes Yes
Sensitive questions Some Limited Limited Yes
Sources of bias
Social desirability No Some Worse No
Interviewer bias No Some Worse No
Respondent’s reading skill Yes No No Some
(Source: Neuman. 2006, p.300)
In terms of research technique, this research study utilises the web survey research method The decision to use the web survey method is based on various factors including cost,
available budget, time constraints, access to research facilities, expertise of the researcher, the research objective, respondent demographics, and survey instrument complexity (Sekaran 2003).
The web survey research method provides a quick, efficient and accurate means of assessing information about the population through its ability to reach a geographically widespread population, interactivity and respondent anonymity (Zikmund 2003), and is more
appropriate where there is a lack of secondary data (Zikmund 2003). In this case, secondary data relative to F&AO in Australia is very limited.
The web survey method provides the opportunity to design and implement the survey questionnaire within a very short period of time, and also allows testing of the questionnaire before its being made available to the intended respondents, which provides the opportunity to evaluate any design and completion issues.
3.4 Research survey design
Drawing on material presented in Chapter Two, this section considers the research
questionnaire design and follows a three-stage approach comprising item generation, pilot survey design, and the main survey design.
3.4.1 Construct and item generation
The generation of items to measure the constructs and design the pilot questionnaire was achieved through the literature review undertaken in Chapter Two.
Seeking a high response rate, it was felt the survey questionnaire should be kept to a
reasonable size. Seventeen constructs and sixty two items were identified as being relevant to this study, and these would allow appropriate classification of data collected.
Seven-point Likert scale questions were used in the main body of the questionnaire in order to allow for more sophisticated data analysis than if categorical measurement scales were used. Neuman (2006) noted the real strength of the Likert scale measurement was in its simplicity and ease of use, and the fact it allowed for more comprehensive multiple indicator measurement when several items are combined, while Davis (2005) confirms reliability is high when it comes to the ordering of people with regard to a particular attitude.
The pilot questionnaire was divided into five sections, each designed to collect data relating directly to the research questions under study.
Section A (Outsourcing information) was intended to gather information regarding participants’ perceptions of the level of outsourcing undertaken by small business, as the level of outsourcing can be influenced by perceptions relating to decision, benefit and impact
Section B (Outsourcing decision factors) was designed to gather data relating to factors affecting the outsourcing decision by small business.
Section C (Outsourcing benefit factors) collected data regarding participants’ perceptions of the benefits of finance and accounting outsourcing by small business.
Section D (Outsourcing impact factors) provided analysis data on the perceptions of the impacts upon small business of finance and accounting outsourcing.
Section E (Background information) collected general demographic categorical data from respondents.
Thirty-eight finance and accounting service providers and five academics piloted the questionnaire. The academics came from diverse backgrounds with the majority possessing small business and research specialisations; the survey instrument was also reviewed by experts in survey design. Several iterations of the questionnaire were pilot tested, with the feedback from the accounting service providers suggesting the questionnaire was comprehensive and highly relevant.
Piloting was achieved by presenting the pilot survey instrument to the Annual General Meeting of a national accounting service provider in November 2006. Conference attendees were asked to complete the pilot survey, which included a section for comments and
recommendations regarding questionnaire layout, relevance and appropriateness of the questions to address any ambiguities.
While the small business survey questionnaire was based on the accounting service providers’ survey, the questions were rephrased to represent the small business owners’ perspective. The questionnaire items’ design is described as follows in the order the questions appear within the questionnaire.
Section A – Outsourcing information (Questions 1 & 2)
Questions 1 and 2 in Section A of the questionnaire collected data regarding participants’ perceptions on the level of outsourcing undertaken by small business. All survey
Table 3.4 Item measures of degree of outsourcing activities
Item Description
Question 1 Do you think small business outsources more or less of their accounting activities than other small business?
1a Level of outsourcing by small business clients compared with other small businesses
Question 2 To what extent do you believe small business would consider outsourcing the following activities?
2a Computer accounting help and training
2b Accounting systems installation and training
2c Monthly business performance reporting
2d Quarterly cash-flow reporting
2e Management accounting service
2f Budget and cash-flow service
2g CCH benchmarking and KPI service
2h Financial diagnosis using Optimist
2i Financial planning and forecasting service
2j Turnkey management systems
2k Job costing/management software
2l Business exit/succession planning service
2m Business appraisal service
(Source: developed for this study)
Section B – Outsourcing decision factors (Questions 3–7)
Questions 3 to 7 in Section B of the questionnaire collected data regarding participants’ perceptions relating to the factors affecting the outsourcing decision by small business. All survey questionnaires provided the opportunity to collect data that could be contrasted and compared. Likert scale questions were used in this section.
Table 3.5 Item measures of outsourcing decision factors
Item Description
Question 3 I consider the following activities as core activities in small business:
3a Payroll
3b Accounts receivable
3c Computer hardware support
3d Computer software support
3e Management accounting
Item Description
Question 4 The timing and co-ordination of the following activities are critical to the overall performance of small business:
4a Payroll
4b Accounts receivable
4c Computer hardware support
4d Computer software support
4e Management accounting
4f Internal auditing
4g Human resource management
Question 5 In deciding to outsource the internal accounting activities, what level of importance do you believe small business would place on the following criteria for selecting an outsourcing provider?
5a Prior experience with service provider
5b Cost of services provided
5c Flexibility of services provided
5d Testimonials from other businesses
5e Business ethics
5f Level of expertise of service provider
5g Industry knowledge
Question 6 There are many activities that small business DO outsource, or WOULD consider outsourcing because:
6a Outsourcing provides greater flexibility with accounting activities 6b Specialist service providers can provide a better service
6c Outsourcing can yield cost savings
6d Other businesses are using outsourcing service providers
6e Service providers are in a better position to manage changes in technology, etc 6f Outsourcing allows small business to minimise risk
6g Service providers are generally specialists in their field
Question 7 There are many activities that small business DO NOT outsource because: 7a It would be difficult to appraise the service providers’ performance
7b The service pr oviders could act in their own interest to the detriment of the business 7c Timing and co-ordination of activities are critical to the success of the business