Qualitative Research
Introduction
The field of social sciences is primarily concerned with making sense of the world around us. The question of how to study the social world, however, unleashes a number of intense philosophical debates. These debates chiefly revolve around ‘ontological’ issues, denoting ‘what is there to know about the world’. Ontological debates concern themselves with questions about whether the social world contains truths independent of human observation or whether it is human interpretation that gives meaning to phenomena. Hidden in this debate is the tension between qualitative and quantitative research methods and the tension over the use of numbers in qualitative research (Kura, 2012).
Couples with the ontological debate about the essence of truth in the social world, social scientists also grapple with epistemological concerns about what can be known about the world and consequently, how to study it. A related epistemological question is about whether there exist relationships between social phenomena and what are the best ways to study such phenomena (Kura, 2012).
Positivist Approach vs Interpretive Approach
Rooted in the ontological framework of social reality, positivists are concerned with variables. They assume that the social world can be studied in the same way as the natural world and that the methods of studying both worlds are complementary. As a result, positivists hold that scientific laws must withstand the rigor of ‘logical consistency, falsifiability, explanatory power and survival.’ Reality, to positivists, is stable. Therefore, there is a need for empirical evidence and statistical proof for theories. A systematic and methodological value free process based on rationality, objectivity, prediction and control is the way of arriving at the truth for positivists. Despite its popularity and appeal, positivists can be accused of trying to over-simplify the social world. Positivists are also unable to explain the nuances behind the causes and processes of research phenomena emanating from case studies restricted to single units of analysis (Kura, 2012).
The interpretive approach to study the social world emphasizes the need to rely of language rather than numbers. Social scientists practising interpretivism believe that social reality is based on the subjective interpretation of actions. The examination of language is considered a rich source of meaning to understand people’s actions and behaviors. A detailed understanding of the context of people’s lives, achieved through observation and immersion of the researcher into the lives of the subjects, lies at the root of the interpretive approach (Kura, 2012).
Chief Characteristics of Qualitative Research
Qualitative research flows out of the school of interpretivism. It downplays the role of statistics. It focuses on finding means to understand the context of social situations through means such as interviews, observation, participant observation and focus groups. The qualitative method, therefore, deals little in numbers. It seeks to determine the context and nuance of social situations by focusing on small representative samples meeting the criteria of ‘cause, consequence and dimension’. The data collection methods are characterized by close interaction between the researcher and the subject and the data generated is extremely rich in scale, relativity and absolutism. The data analysis yields emergent concepts and ideas and establishes patterns and typologies. As a result, qualitative research is useful to explore or to revisit old ones with a fresh perspective. The result of qualitative method is a holistic picture that provides an in-depth understanding of the phenomenon being studied. The chief method of a qualitative researcher, therefore, is observation, as opposed to measurement – the chief method of the quantitative researcher (Kura, 2012).
The Role of Numbers in Qualitative Research
Given the ideological and philosophical underpinning of qualitative research in the epistemological position of interpretivism, social scientists pursuing qualitative research are skeptical about the utility of numbers in meeting their requirements. The fact that qualitative researchers do not overly use statistics often puts the significance of their work into doubt, especially in the data driven modern age. Critics have labeled qualitative research as being imprecise. In response, qualitative researchers have begun using quantitative techniques to lend broader credibility to their work (Maxwell, 2010).
Quasi Statistics and the Likert Scale
Critics have attacked the use of statistics in the work of qualitative researchers as being a work of ‘quasi-statistics’. They charge qualitative researchers to ascribing numbers to relatively incomparable degrees of agreement, such as ‘many, often, typically and sometimes.’ In this context, of specific relevance is the quantitative treatment given to the Likert scale. Qualitative researchers obtain a number of real world inputs through surveys where participants comment on the acceptability and desirability of phenomenon in gradations of ‘definitely, most likely, likely, unlikely, most unlikely and certainly not’. Such gradations are called the Likert scale, which could be in the scale of five or seven distinctions or more. Essentially, the Likert scale represents ordinal data. It captures the subjective response of the respondent. The response depends upon what each respondent means by ‘most’ or ‘least’. Qualitative researchers have begun to treat this data as ratio data and have begun to determine measures of central tendency such as the mean from such data. This makes little sense. For example, if the average response of a five point Likert scale questionnaire were to be 2.7, the deduction that people were ‘more inclined than average’ for a certain course of action may not be apt. This is primarily because a ‘2’ on the Likert Scale is not essentially the double of a ‘1’; ‘least likely’ is not half of ‘less likely (Jameison, 2004).’ The mid point between Candidate A and Candidate B on a voting machine may be equally construed as ‘either’, ‘neither’ or ‘I don’t know’. Therefore, quantifying Likert scales is always fraught with contextual risk (Baka, Figgou, & Triga, 2012). However, qualitative researchers defend the trend of converting ordinal response to ratio data and treating the same quantitatively using the logic that the effort is useful to discern patterns in the responses of participants.
The Significance of Numbers for Qualitative Research
Notwithstanding the charge of resorting to ‘quasi statistics’, numbers have a role to play in qualitative research. Counting remains central to the analytic process. Data can yield patterns if subjected to statistical analysis. A pattern generates clues about the typicality and intensity of an event. Such counting is often unconscious as qualitative researchers attempt to arrive at the context of a social situation. While the quantitative researcher uses counting to get the facts right, the qualitative researcher would use counting to get the interpretation of the participant right (Sandelowski, 2001).
Therefore, based on the sample size, qualitative researchers may be able to elicit descriptive and inferential statistics from their studies. Because numerical description of information can make information stand out clearly, it can lead to qualitative researchers arriving at working hypotheses (Sandelowski, 2001).
Sometimes, verbal explanations of phenomena can become overly complex. In such situations, it is prudent to represent the data numerically for comprehension. This approach may also reveal a key finding that would have been otherwise hidden amidst words. Numbers are also useful in re-presenting events in the forms of means and frequencies for the ease of comprehension of readers (Sandelowski, 2001).
Complications in the Use of Numbers by Qualitative Researchers
While number play a role in qualitative research, there is a danger in the misrepresentation of rich data through numbers. Specifically, qualitative researchers need to be wary of ‘verbal counting, over counting, misleading counting and acontextual counting (Sandelowski, 2001).’
Verbal counting error occurs when researchers imply numbers without actually giving any. For instance, researchers may use terms like ‘a few’ without indicating to the reader the context of their findings. Over counting occurs when researchers overcompensate to avoid verbal counting; in this case, there is an overuse of numbers to count inessential data. Misleading counting occurs when researchers use percentages when representing small samples. Acontextual counting occurs when researchers provide a statistic and do not follow up on the implications (Sandelowski, 2001).
Towards a Happy Medium – Mixed Method Research
Qualitative research, while being rich in contextual appeal, remains stymied by criticism of low mathematical rigor. Using numbers strictly within the realms of qualitative research is not enough to bring rigor into qualitative research, primarily because the context and setting in which qualitative research is done is more amenable to verbal treatment. To address the twin requirements of rigor and optimal research, mixed method research is beginning to make its presence felt (Maxwell, 2010).
Mixed method research may involve an initial explorative research using qualitative means, a follow up with rigorous quantitative methods and a confirmation using qualitative means such as interviews. Mixed method research creates a dialogue between the qualitative and quantitative research methodologies of seeing and interpreting, not simply by applying concepts related to one method of research to another. In effect, mixed method research stresses on the strengths of qualitative and quantitative methods without compromising on the essence of both (Maxwell, 2010).
Conclusion
Qualitative research delves into the context of situations and emerges with rich insights. Qualitative research requires the immersion of the researcher into the situation and its results are best described verbally. Critics of qualitative research point out that it lacks mathematical rigor. In response, qualitative researchers have begun to use numbers. While numbers play a role in categorizing and organizing qualitative research, there remain risks of misrepresentation of qualitative responses through mathematical treatment. Likert scales are an apt example of the contradictions involved in the quantitative treatment of qualitative responses. The way forward is to combine the strengths of quantitative methods and qualitative methods through mixed method research.
References
Baka, A., Figgou, L., & Triga, V. (2012). ‘Neither agree nor disagree:’ A critical analysis of the middle answer category in voting advice applications. International Journal Electronic Governance 5/3-4: 244-261. Retrieved 12 Oct 2014, from https://www.academia.edu/2527539/Neither_agree_nor_disagree_a_critical_anlalysis_of_the_middle_answer_category_in_Voting_Advice_Applications
Sandelowski, M. (2001). Focus on research methods: Real qualitative researchers do not count – the use of numbers in qualitative research. Research in Nursing and Health 24: 230 – 240. Retrieved 12 Oct 2014, from http://onlinelibrary.wiley.com/store/10.1002/nur.1025/asset/1025_ftp.pdf?v=1&t=i15nqnnq&s=fff473920c84b3a3dcb14e5b829d3c31e4bb55a9
Jameison, S. (2004). Likert scales: How to (ab)use them. Medical Education 38: 1218-1218. retrieved 12 Oct 2014, from http://medicina.udd.cl/ode/files/2010/07/jamieson_ME_2486.pdf
Kura, S.Y.B. (2012). Qualitative and quantitative approaches to the study of poverty: Taming the tensions and appreciating the complementarities. The Qualitative report 17.34: 1-19. Retrieved 12 Oct 2014, from http://www.nova.edu.ssss/QR/QR17/kura.pdf
Maxwell, J.A. (2010). Using numbers in qualitative research. Qualitative Inquiry 16: 475-482. DOI: 10.1177/1077800410364740. Retrieved 12 Oct 2014, from http://qix.sagepub.com/content/16/6/475
