A need for the multiple methods approach

In recent years, the multiple methods approach has been proposed to overcome the above-mentioned challenges of self-report assessments (Hofer, 2004; Muis et al., 2016). Epistemic conceptions literature contains several variants for the term multimethodology, such as triangulation, multi-method, and mixed-method. Sometimes these terms have been used as synonyms, sometimes not. Among researchers there is no all-encompassing agreement about the meanings and definitions of these terms (Teddlie & Tashakkori, 2010). One alternative is to consider the term “multimethod” as referring to research involving multiple qualitative or quantitative methods in data collection and analysis, while “mixed-method” can be used to refer to research in which both quantitative and qualitative methods are combined in the same study or in a series of studies (see also Creswell, 2010; Leech & Onwuegbuzie, 2009). The present discussion pertains to this definition.

As mentioned before, there is a long history of using multiple methods in research on epistemic conceptions (see Hofer, 2004; King & Kitchener, 2004; Kuhn, 2005). In the existing literature, several advantages attributable to the mixed- or multi-method approach have been reported. Firstly, through the use of the mixed- or multi-method approach, it is possible to select and integrate the appropriate methods to gain a more thorough picture of the phenomenon (Ghel-bach & Brinkworth, 2011; Karabenick et al., 2007; Muis et al., 2016). Secondly, the use of multiple methods allows a researcher to minimise the weaknesses or complement the strengths of particular methods. Thus, the advantage of combining the different assessment methods by which individuals’ conceptions of knowledge are measured is that it offers multiple insights into and thus a more comprehensive view on individuals’ conceptions of knowledge and knowing. As Muis and colleagues (2016) have argued, the mixed-method approach (i.e., combining surveys with interviews) provided much deeper nuances of the nature of students’ conceptions. This kind of information would have been impossible to gain by means of self-reporting only. Thirdly, through the use of the mixed- and multi-method approach it is also possible to identify and address new and unexplored aspects of students’ conceptions and gain a more complete understanding of the phenomenon (see Onwuegbuzie, Johnson, & Collins, 2009). Finally, the mixed- and multi-method approach provides a way to enhance the cognitive validity of surveys which focus on abstract constructs, such as conceptions of knowledge and knowing (Ghelbach & Brinkworth, 2011; Karabenick et al., 2007).

It is important to notice that the mixed- and multi-method approach can be employed in data collection, data analysis, and merging interpretations (Creswell, 2010). Epistemic conceptions literature features several variants of multi-and mixed-method research. In essence, these variants can be roughly divided into two main strategies, namely sequential and concurrent strategies (see Figure 11.1). To our knowledge, most mixed- and multi-method research on individuals’ conceptions of knowledge have followed a sequential strategy including at least two separate phases or sub-studies. In this approach the

Example 1, Sequential strategy

The study consists of separate phases where each dataset is analysed separately. The follow-up phase produces answers to the questions which remain unanswered in Phase 1.

Phase 1:

A survey to describe students' conceptions of knowledge

■> Findings

Phase 2:

Follow-up ■> study, student interviews

■> Findings

Example 2. Concurrent strategy

A study in which two or more data collection and analytical methods are used simultaneously. Different stages of study are merged.

Online enquiry task with think-aloud method

> Findings

Student interviews

Observation

FIGURE 11.1 Examples of sequential and concurrent designs for research on individuals’ conceptions of knowledge researcher typically uses one method to collect and analyse the data. After that, the results that need additional explanations are identified and the researcher accordingly selects a second method. The purpose of the second method, i.e., the follow-up phase, is to provide a better understanding of the research problem than using only the first method. As an example, qualitative analysis of interview data is often used to complement and extend the findings of quantitative analysis of questionnaire data (e.g., Hofer, 2006; Hyytinen et al., 2016; Muis et al., 2014). In these cases, a sequential strategy is used to develop and revise quantitative data collection instruments.

A concurrent strategy, by contrast, is used to confirm and cross-validate findings, to seek information at different levels, and address different questions and perspectives (e.g., Barzilai & Zohar, 2012; Hofer, 2004; Hyytinen et al., 2014; Kienhues, Stadtler, & Bromine, 2011; Stromso & Brâten, 2014). A concurrent strategy refers to situations in which two or more data collection and analytical methods are simultaneously and interactively applied to understand a phenomenon (Leech & Onwuegbuzie, 2009). In this strategy, the different methods have equal priority. This kind of strategy may help strengthen the validity and reliability of the study. It is worth noting that mixed- and multimethod research is not limited to take either a sequential or concurrent strategy; the same study may employ both these strategies. Figure 11.1 illustrates the main differences between sequential and concurrent strategies.