A good example provided by Taflinger (1996) mentions that research is finding out information not presently known. Struwig and Stead (2010: 3) explain that research can be distinguished from information gathering and decision-making, by three distinct but interrelated characteristics namely:
Research is based on an open system of thought.
Researchers examine data critically.
Researchers generalise and specify limits on their generalisations (Phillips and Pugh, 1994: 47).
A definition provided by Abrey (2015: 130), who cited Leedy and Ormrod (2010: 2), states that research is defined as a process entailing collection, analysis and interpretation of data with the intent of increasing understanding and insight into a specific phenomenon. Abrey (2015: 130) states that research can frequently be categorised as being either scientific (which is also known as pure/basic) research or applied research.
According to Shuttleworth (2008) pure scientific research is about explaining the world around us and trying to understand how the universe operates. Shuttleworth (2008) further explains that pure scientific research is about finding out what is already there without any greater purpose of research than the explanation itself. It is a direct descendent of philosophy, where philosophers and scientists try to understand the underlying principles of existence (Shuttleworth, 2008). Scientific research is systematic, based on fact and guided by testing theories and hypothesis, often generated by pure science and applied to real situations, addressing more than just abstract principles (Shuttleworth, 2008; Abrey, 2015:
130). As mentioned by Shuttleworth (2008), applied scientific research can be about finding out the answer to a specific problem.
5.2.1.1. Methods of data collection
A good measurement tool has three characteristics, namely reliability, validity and practicality. Reliability refers to the accuracy and precision of a measurement procedure, whereas validity is the extent to which a test measures what it is actually supposed to measure. Practicality on the other hand, is concerned with factors such as economy, convenience and interpretability (Schindler and Cooper, 2006). Both the reliability and validity of the measuring instrument must be assessed before researchers proceed to assess the strength of relationships in an empirical framework. A description of the statistical techniques used in this study to assess the reliability and validity of the results, is presented below.
5.2.1.2. Reliability of measuring instrument
Botha (2013: 357) who cites Collis and Hussey (2014: 53) mentions that in positivistic studies it can be seen that reliability is an important issue and thus involves assessing the degree of consistency between multiple measurements of a variable (Hair Jr, Black, Babin and Anderson, 2009: 3; Strang, 2015: 68). A definition of validity provided by Hair Jr et al. (2009: 2) defines validity as “… extent to which a variable or set of variable is consistent in what it is intended to measure,”.
The main aim of reliability is that it is concerned with estimates of the degree to which a measurement is free of random or unstable error and secondly that a measuring instrument is considered to be reliable to the degree that it supplies consistent results (Schindler and Cooper, 2006: 8; Botha, 2013: 357; Strang, 2015:
70). Botha (2013: 357) explains that internal consistency is a commonly used measure of reliability, which applies to the consistency among the variables in a summated scale. Schindler and Cooper (2006), Hair Jr et al. (2009) and Botha (2013: 357) provide an in-depth explanation of internal consistency and all agree that the rationale for internal consistency is that the individual items or indicators of the scale should all measure the same construct and thus all must be highly inter-correlated.
According Botha (2013: 358) who cited Schindler and Cooper (2006) explain that the Cronbach’s alpha technique is a type of reliability estimate or coefficient of internal consistency and is based on the average correlation of variables within a specific set of items measuring a construct. It can be seen that the reliability coefficients of less than 0.50 are deemed to be unacceptable, those between 0.50 and 0.60 are regarded as questionable, and those above 0.70 as acceptable (Botha,
2013: 358). Any coefficients greater than 0.80 are regarded as excellent (Botha, 2013: 358). In practice, it is generally agreed that the lower limit for the Cronbach-alpha coefficient is 0.70. In the case of exploratory research, this lower limit may be reduced to 0.60 (Hair Jr et al., 2009; Botha, 2013: 358). In this study, Cronbach’s alpha coefficients were used to measure the degree of reliability of the measuring instrument. Consequently, they were also used to determine which items would be included as measures of specific constructs.
5.2.1.3. Validity of measuring instrument
According to Collis and Hussey (2014: 4) validity is concerned with the extent to which the data collected is a true reflection of what is being studied (Botha, 2013:
358). A measuring instrument can only be deemed valid if it measures what the researcher thinks or claims it does and adequately reflects the real meaning of the concept under consideration (Schindler and Cooper, 2006; Botha, 2013: 358; Collis and Hussey, 2014: 8). A definition of validity provided by Hair Jr et al. (2009: 3) defines validity as “… extent to which a measure or set of measures correctly represents the concept of the study,”. In a study, the definition of a construct is proposed by the researcher and thus it must be matched to the selected indicators or measures which in turn determine the validity of the study to a great extent (Botha, 2013: 358).
For the purpose of this study, construct validity was utilised to determine whether the measuring tool actually fulfilled its purpose. According to Botha (2013: 358) who cited Zikmund, Babin, Carr, and Griffin (2013: 305) explains that construct validity
generated from a theory based on the concepts. Construct validity implies that the empirical evidence generated by a measure is consistent with the theoretical logic about the concepts (Botha, 2013: 359; Strang, 2015: 70). A good example of construct validity provided by Leedy and Ormod (2010) explains that when researchers ask questions or make statements as a way of assessing an underlying construct, they should have obtained some kind of evidence that their approach does, in fact, measure the construct in question (Botha, 2013: 359). According to Botha (2013: 359) who cites Venter (2003), a measuring instrument can only be considered to exhibit construct validity if the scale has both convergent and discriminant validity.
According to Botha (2013: 359) who cited Strang (2015: 70) explains that convergent validity refers to “… the degree to which scores on one scale, correlate with scores on other scales designed to assess the same construct,” (Botha, 2013:
359). If a known measure of a construct exists, researchers might correlate the results obtained using the known measure with those derived from the new measure, thus providing indications of convergent validity (Schindler and Cooper, 2006). Discriminant validity, on the other hand, as mentioned by Botha (2013: 358) who cited Strang (2015: 70) refers to “… the degree to which scores on a scale do not correlate with scores from scales designed to measure different constructs,”.
Establishing the discriminant validity of a measuring instrument will determine the extent to which each construct is separated or distinct from other constructs in the theory or related theories. In this study, the measuring instrument was developed based on constructs identified in theory, and consequently assessed the
discriminant validity in an effort to establish whether the measuring instrument adequately discriminated between the constructs being assessed.
For the purpose of this research, the multivariate technique of exploratory factor analysis was implemented to assess discriminant validity and this technique has been used by numerous researchers (Hair Jr et al., 2009; Botha, 2013). The primary purpose of multivariate technique of exploratory factor analysis is to define the underlying structure among the variables in the analysis (Hair Jr et al., 2009; Botha, 2013: 360).
5.2.1.4. Quantitative and Qualitative Methodology
Various methods are available for researchers to collect, analyse and interpret information. However, as Struwig and Stead (2010: 3) explain there is no commonly agreed method to acquire knowledge. According to Anderson and Poole (2009: 22) a crucial step in a study is the choice of methods. The main reason for choosing the correct method is simply because the wrong method may cause the whole study to be criticised on the grounds of inappropriate design, or even worse, as being unscientific or illogical (Anderson and Poole, 2009: 22). According to Abrey (2015:
131) who cited Verhoeven (2011), the two main approaches to research are the qualitative and quantitative methods. As explained by Anderson and Poole (2009:
22), quantitative research is typified by experimental studies in science-based disciplines where findings are usually expressed in numerical form. The other research method explained by Anderson and Poole (2009: 22) is qualitative research, which is characterised by ethnographic and historical studies where
5.2.1.5. Qualitative and quantitative approaches
Taylor (2005: 3) states that both quantitative and qualitative research share common ground. Both are concerned with reliability and study designs and the approaches are similar (Taylor, 2005: 3). Anderson and Poole (2009: 27) who share the same opinion as Taylor (2005) state that it is sometimes desirable to combine qualitative with quantitative research to maximise the theoretical implications of research findings. Anderson and Poole (2009: 27) further explain that although a combination of research approaches may be frowned upon by some in the research community because of the vastly different theoretical backgrounds and methods of data collection in each, a combined approach can prove valuable in certain projects.
According to Sheldon (2015) qualitative research follows a semi-structured discussion guide to ensure that all topics under consideration are covered and that the discussion stays relevant. However, the questioning is open and participants are encouraged to explore the reasons for their responses (Sheldon, 2015). The discussion process can reveal underlying views and motivations, behavioural triggers and barriers. It can explore reactions to messages, printed material, design features and new products, test understanding of terminology, help generate new concepts and much more (Sheldon, 2015).
During the data collection phase, the qualitative researcher as mentioned by Abrey (2015: 132) becomes immersed in the research environment, thereby experiencing the situation through interaction with research participants. The use of focus groups and in-depth interviews are regarded as a commonly used qualitative approach (Sheldon, 2015). Focus groups (also called group discussions) according to