Easterby-Smith et al. (2012) state that establishing the most suitable methodology and methods is still under discussion among researchers as implementing different methods will provide different perspectives on what is being studied. Many authors such as Jankowicz (2005) and Robson (2011) emphasise that there is no straightforward rule that obliges researchers to choose one method for one investigation and another for another investigation. Thus, mixed methods research should not be considered as being universally applicable or being the most appropriate approach for any research context (Creswell, 2014; Bryman and Bell, 2011). Some arguments emphasise the possible dangers of believing that the use of mixed methods is the ultimate solution in deciding upon the best choice of research method. Bryman (2012) stresses that mixed methods research is not superior to research that uses a single method.
However, Guthrie (2010) argues that the findings achieved from using mixed methods may gain credibility and be acceptable to a large audience,
125
especially when similar findings come from different sources. The mixed methods approach uses the strengths of both quantitative and qualitative research (Kumar, 2014). Mixing methods strengthens the validity and reliability of research. Guthrie (2010) argues that adopting the positivist paradigm can increase the reliability and generalisation of research, but it lowers its validity and relevance whereas following the interpretivist research improves validity and relevance but at the expense of reliability and generalisation. Nunan (2006) points out that “a mixed methods
approach can offer additional benefits for the understanding of the phenomenon in question”. Creswell and Clark (2011:13) also support the
above that the mixed methods approach is “a preferred mode to
understand the world” and it is practical as it gives the opportunity to the
researcher to be free to use all methods possible to address a research problem. Johnson and Christensen (2012) favourably argue that the mixed methods approach allows us to understand both the subjective (individual) and the objective (casual) realities in our world.
According to Lou and Dappen (2004), a mixed methods approach is required in educational settings because educational problems are complex and inflexible and need multiple ways of understanding. Therefore, conducting mixed methods research to investigate language- related issues can enhance knowledge in this field as it focuses on explaining and understanding the topic being searched (Riazi and Candlin, 2014). This study uses questionnaires to measure the students’ views about the challenges of reading comprehension. According to Bryman and Bell (2011) and Gray (2014), questionnaires are a well-known and widely
126
used primary data gathering technique for collecting quantitative data. On the other hand, the tool of semi-structured interviews is used in order to support the findings for quantitative data.
Although Creswell (2012 and 2014) indicates that the procedure of using both qualitative and quantitative methods is time-consuming as it requires collecting and analysing extensive data, he indicates that this combination of methods provides a broader perspective to the study and a better understanding of the research problem and question than using one method by itself. Greener (2011) presents some justifications for mixing methods in a research project which are: to achieve breadth (through using quantitative methods) and depth or closeness (through using qualitative methods); to gain greater validity for research, to bring the methods themselves into focus to check whether different methods produce different findings; and to see if using mixed methods generate new findings in fields that tend to be in a particular way. Similarly, Patton (2002) strongly supports the use of mixed methods as it significantly enhances the accuracy of the data, which is the basis of any research. Furthermore, by including both numeric trends and specific details, a better understanding of a phenomenon might be achieved. In addition, using mixed methods may improve the validity of the research and allow for making generalisations, which is normally not easily done using a qualitative or quantitative research method alone.
Saunders et al. (2012) highlight some reasons for using a mixed methods design as follows: the use of mixed methods may help findings to be
127
elaborated, clarified, illustrated and confirmed; it might also allow the generalisability of a study; it solves the problem of getting insufficient data from one method; it may help reflect a greater diversity of views in the study; combining data helps to make sure that the findings from the two methods corroborate with each other; and it might lead to greater confidence in the conclusions of a study.
Table 4.6: Quantitative, mixed methods and qualitative research
Quantitative Research Mixed Methods Qualitative Research
Scientific method
Deductive or “top-down”. The researcher tests hypotheses and theory with data
Deductive and inductive
Inductive or “bottom-up”. The researcher generates new hypotheses from data collected during fieldwork
View of human behaviour
Behaviour is regular and predictable
Behaviour is some- what predictable
Behaviour is fluid, dynamic, situational, social, contextual, and personal Most common research objectives Description, explanation, and prediction
Multiple objectives Description, exploration, and discovery
Focus Narrow-angle lens, testing specific hypotheses
Multi-lens focus Wide-angle and “deep-angle” lens, examining the breadth and depth of phenomena to learn more about them Nature of observation Attempt to study behaviour under controlled conditions Study behaviour in more than one context or condition
Study behaviour in natural environments. Study the context in which behaviour occurs
Nature of reality
Objective (different observers agree on what is observed)
Common sense realism and pragmatic view of world (i.e. what works is what is “real” or true)
Subjective, personal, and socially constructed
Form of data collected
Collect quantitative data based on precise measurement using structured and validated data collection
instruments (e.g. closed- ended items, rating scales,)
Multiple forms Collect qualitative data (e.g. in- depth interviews, participant observation, field notes, and open-ended questions). The researcher is the primary data collection instrument
Nature of data Variables Mixture of variables, words, and images
Words, images, categories
Data analysis Identify statistical relationships
Quantitative and qualitative
Search for patterns, themes, and holistic features Results Generalisable findings Corroborated findings
may generalise
Particularistic findings. Representation of insider (i.e. “emic”) viewpoint. Present multiple perspectives Form of final
report
Statistical report (e.g. with correlations, comparisons of means, and reporting of
statistical significance of findings)
Eclectic and pragmatic Narrative report with contextual description and direct
quotations from research participants
128