14^th December 2025 (Published here on 23^rd January 2026)

Starting Point: My Current Understanding.

I always enjoy thinking through writing. Each time I do this, it feels as if I am leaving the most distilled part of my brain here, rather than the dross or the ineffable stuff. In this respect, I have been inspired by one of my previous supervisors who once strongly encouraged me to write some “publishable” text. Doing so would greatly help train the rigor of my (academic) writing and thinking. I also realize that these so-called insights also unfold gradually during the writing process. This subtle influence may not have a positive effect on one’s thinking ability quality in the short term, but over time, as thoughts take shape in words, it becomes as natural as breathing. This is what I enjoy most about doing a PhD – to think, to learn, and to grow.

What I want to write in today’s post is about the logical rigor within four research-philosophical worldviews (post-positivism, constructivism, transformative worldview, and pragmatism). The differentiation of research worldviews also implies the use of different types of research methods (broadly speaking, qualitative and quantitative). As I view qualitative and quantitative methods as tools to advance my intended research ideas, I am more interested in the logical “truth” behind these methods.

This article will begin with the comparison between qualitative and quantitative research, and will expand the discussions to think more deeply about the logical implications of the four philosophical research worldviews.

First, I will now state what I currently agree with: I agree that, judged by academic research standards, qualitative research is not as rigorous, and this is also a kind of consensus in academia: its logic tends more toward deductive-inductive reasoning. Moreover, a major problem is that when different people study the same topic, even using the same method, they may end up with very different results. Meanwhile, the logic of quantitative research, relatively speaking, tends to lean more toward deductive reasoning.

This is my starting point. Now, I much acknowledge that I will use generative AI (particularly ChatGPT) as my thinking collaborator (drawing inspiration from my supervisor’s words) Such collaborative approach aims to enhance my thinking process and foster intellectual growth.

Three Fundamental Paths of Reasoning.

1. Deduction (Deductive reasoning): from general to specific.
2. Induction (Inductive reasoning): from specific to general.
3. Abduction (Abductive reasoning): inferring the best explanation from an outcome.

Deduction

• Logical flow: theory → hypothesis → observation → confirmation/falsification.
• Deduction corresponds to quantitative research. Its rigor lies in the correctness of formal logic: as long as the premises are correct, the conclusion must be correct.

Induction

• Logical flow: observation → pattern → provisional hypothesis → theory.
• Induction corresponds to qualitative research. When I say, “qualitative is not rigorous enough”, what I mean is that induction can never derive “absolute truth” (for example, the black swan paradox). It can only provide “high-probability explanations” or “deep understanding”.

Abduction

• In many cases, qualitative research and pragmatism are actually doing abduction: that is, when facing a phenomenon, they seek the most reasonable explanatory framework.

Through dialogue with generative AI (namely ChatGPT 5.2 Thinking and Gemini 3 Pro Thinking), I began to push this thought further: if qualitative research is regarded as “less rigorous,” what exactly does “rigor” mean then? Is rigor equivalent to logical certainty? Or is it something else?

In post-positivist logic, rigor is largely equated with internal validity, control of confounding variables, replicability, and the strict application of formal logic. From this perspective, rigor means that if the same study is repeated under the same conditions, the results should be the same. This definition of rigor aligns very closely with deductive reasoning: if the premises are fixed and the logic is correct, the conclusion must follow.

However, once this definition of rigor is applied beyond post-positivism, problems arise. In constructivist research, the assumption is that reality is socially constructed and context-dependent. Under this research worldview, expecting identical results from different researchers or different contexts is not only unrealistic but conceptually misguided. If meanings are co-constructed between researcher and participants, then variation is not a flaw in rigor, but a reflection of reality itself.

The discussion above leads to a crucial realization: rigor is not a single, universal standard. Instead, rigor is a worldview-dependent attribute. What counts as rigorous reasoning under one worldview may be inappropriate, or even meaningless, under another. At this point, I began to realize that the common critique of qualitative research as “not rigorous enough” often stems from an implicit post-positivist yardstick being applied where it does not belong: Qualitative research is not failing to meet the standards of post-positivism; rather, it is operating under a different logical contract altogether.

This raises a deeper question: if rigor is not absolute, then what anchors the credibility of research across different worldviews? What prevents research from sliding into arbitrariness or relativism? To address this question more explicitly, Table 1 below compares how rigor is defined, justified, and constrained across four research-philosophical worldviews.

Table 1: Criteria for evaluating research rigor across research worldviews

Moreover, the statement that “quantitative research tends more toward deduction” is not entirely accurate. Quantitative research does indeed often organize its argumentation around a deductive chain of theory → hypothesis → operationalization → testing, but it also makes extensive use of: (1) exploratory analyses (with strong inductive tendencies); (2) data-driven modeling (sometimes not starting from theory at all); and (3) abductive revisions following anomalous findings revealed through model comparison. A more precise formulation would be that quantitative research is more likely to explicitly write its reasoning structure in a deductive form, rather than that it is purely deductive in nature.

From this perspective, my earlier lines of thought somewhat reduced “rigor” to a one-dimensional indicator. If one truly wants to think seriously about logical rigor, rigor needs to be unpacked into at least two distinct layers. The first layer is inferential validity – whether the conclusions hold within the adopted epistemological assumptions and methodological rules (internal coherence); The second layer is testability and strength of constraint – by what means a conclusion can be challenged, falsified, or revised (external constraint). Otherwise, the discussion will remain stuck at the superficial qualitative–quantitative divide.

Therefore, rigor is not an attribute of any single method. Rather, it is the result of an integrated configuration of worldview + reasoning rules + forms of evidence + constraint mechanisms. When the research topic concerns areas such as my PhD focus on changes in user experience, curiosity, boredom, or meaning-making, certain implications are unavoidable:

• Some forms of “truth” resemble mechanistic explanations (why/how), which need to grow out of qualitative inquiry and abductive reasoning.
• Some forms of “truth” resemble testable differences (does it work), which require quantitative methods and deductive reasoning to be consolidated.
• Still other forms of “truth” are normative (should we), which push evaluation toward the standards of the transformative worldview / pragmatism (values, consequences, beneficiaries).

What follows is therefore a more rigorous line of argumentation:

Now, I am more inclined to understand rigor as a form of stronger external constraint (e.g., reproducible procedures, a testable chain of reasoning, and explicit governance of researcher degrees of freedom). In this sense, many qualitative studies do not treat statistical replicability as their primary goal. More often, they establish trustworthiness through reflexivity, triangulation, and an audit trail.

Therefore, qualitative research is not necessarily non-rigorous; rather, it relies on different rigor mechanisms and evaluative standards. At the same time, qualitative research is indeed more easily influenced by differences in researchers’ interpretive frameworks: even when the research questions and methods are similar, different researchers may still generate very different conceptual structures and narrative conclusions. By contrast, quantitative research more often organizes its argument through a deductive chain of “theory → hypothesis → operationalization → testing,” and strengthens comparability and auditability through standardized measurement and statistical inference (but it also faces the problem of researcher degrees of freedom arising from choices in operationalization and analytic pathways).

This is precisely the starting point of what I want to pursue next through four research-philosophical worldviews: how different worldviews define what counts as rigorous, and by what means each constrains reasoning and approaches the “truth” it recognizes.

How Rigor is Defined Differently across the Four Research Worldviews.

Qualitative and quantitative research should be understood primarily as tools for research rather than as ends in themselves. The deeper intellectual significance of methodological choices lies in the worldviews that underpin them. When we move beyond surface-level methodological debates and focus on how various research worldviews attain logical coherence or conflict, the conversation shifts from specific techniques to underlying assumptions about reality, knowledge, and justification. Knowing these logical implications across worldviews helps reveal why disputes about rigor often persist and why they cannot be resolved at the level of methods alone.

From a post-positivist perspective, rigor is grounded in a revised objectivist logic [1] that combines determinism with reductionism. This worldview assumes that truth exists independently of the researcher, while acknowledging that human cognition is inherently imperfect. Thus, absolute truth cannot be fully attained; it can only be approached asymptotically through successive refinement. Within this framework, quantitative research plays a central role, precisely because falsification is necessary when direct access to truth is impossible. The notion of rigor invoked earlier in the text—emphasizing reliability, validity, and generalizability—is in fact rigor as defined by post-positivism. Within this closed logical loop, quantitative research is indeed more rigorous than qualitative research. However, once one steps outside this loop, its limitations become apparent. In this sense, post-positivist rigor requires complex social phenomena to be translated into variables, thereby sacrificing wholeness and contextual richness in exchange for precision and control.

In contrast, constructivism or interpretivism operates according to a logic of intersubjectivity rooted in relativism and hermeneutics. This worldview rejects the assumption of a single, objective truth and instead understands truth as something constructed through interactions between subjects and objects. Within this logic, the observation that different researchers studying the same topic may arrive at different conclusions is not a methodological weakness but a necessary consequence of the ontology itself. Rigor in constructivist research does not rest on reproducibility in the statistical sense, but on credibility. This requires recognizing that the researcher is not external to the research process but is part of the research instrument. In this respect, rigor entails reflexivity, transparency, and careful engagement with meaning-making processes, resonating with Prof. Xiang Biao’s (anthropologist at Oxford University) characterization of “using oneself as method.” At the same time, constructivism carries a fundamental logical vulnerability: if all knowledge is constructed, a central challenge is to prevent inquiry from collapsing into self-referential narratives with insufficient constraint.

The transformative worldview introduces a distinct logic centered on power, drawing heavily on critical theory and political orientation. Here, research is not primarily concerned with explaining the world, but with changing it. From this perspective, both post-positivist and constructivist approaches may inadvertently obscure or marginalize the voices of disadvantaged groups. Rigor within a transformative framework is therefore no longer defined by methodological impeccability alone, but by ethical and political justice. A study may be internally consistent and logically sound yet still be considered insufficiently rigorous if it ignores structures of oppression or reproduces existing inequalities. In this sense, research rigor is inseparable from normative commitments and accountability to those affected by the research.

Finally, pragmatism is organized around action and consequences, shaped by a utilitarian and problem-oriented logic. From a pragmatic perspective, disputes over whether reality is objective or socially constructed are often treated as metaphysical questions with limited payoff for inquiry. Truth is defined instrumentally as what proves workable in practice. This stance loosens the qualitative-quantitative divide and, by extension, legitimizes mixed-methods insofar as they effectively address the problem at hand. Accordingly, rigor is judged by the utility, adaptability, and real-world implications of knowledge claims. However, this flexibility also introduces a risk: pragmatism can slip into logical opportunism, where methods are assembled solely for problem-solving convenience while deeper epistemological tensions are ignored.

From “Qualitative vs. Quantitative Rigor” to Pragmatic Justification.

Returning to the initial thought, the belief that “qualitative research is not rigorous, whereas quantitative research is rigorous” is, in itself, an ontological judgment [2]. If the world is assumed to operate like a machine (positivism), then the deductive logic of quantitative research does appear more stringent. If the world is assumed to be like a flowing river, constituted by human minds (constructivism), then quantitative logic can appear rather “leaky”, because it filters out too many contextual factors. From this perspective, rigor should not be treated as a single-dimensional ruler (e.g., quantitative research is more rigorous; qualitative research is less rigorous). Instead, it is more appropriate to view rigor as a matter of logical alignment:

• Quantitative rigor: do the measurement instruments precisely correspond to the objective variables? (construct validity)
• Qualitative rigor: does the interpretive framework faithfully reflect participants deeper lived experience? (authenticity)

Genuine academic training and logical rigor are not about proving that quantitative research is superior to qualitative research (or the other way around). They rest on whether a researcher can clearly state the logical framework within which the study operates and then build an argument that is internally coherent and fully justified.

Positioning My Research Worldview

Going one step further, it is now clear that the most central research worldview of mine is neither (post-)positivism, nor, strictly speaking, constructivism. It is closer to a pragmatist worldview or perhaps moving toward critical realism.

Reflecting on my thinking over the past few days, I have come to realize that the core logic guiding my position is pragmatism. Traditional scholars may engage endlessly in debates over whether the world should be understood as a machine or as a river (the so-called paradigm wars). The pragmatist response, however, is straightforward: it depends on the research problem. This stance refuses to be constrained by a single metaphysical commitment. Ontology, from this perspective, is not a belief system that must be defended at all costs, but an analytical lens that can be adjusted to fit the problem at hand. Truth, therefore, is not defined as absolute correspondence with reality, but as instrumental usefulness: if treating the world as a machine better predicts outcomes for a particular problem, that model is adopted; if treating the world as a river better explains meaning in another context, that model becomes more appropriate. More strongly, if this position is motivated not only by problem-dependence but also by a deep desire for an integrated and rigorous logic, then it may align more closely with critical realism.

• Logic: The world contains both machine-like aspects (deep causal mechanisms and structures) and river-like aspects (events at the level of experience and human interpretation).
• This worldview treats reality as stratified. The aim is to excavate the structural mechanisms (machine) beneath appearances (river).

From this perspective, a new set of critical insights emerges. Once ontology is treated as problem-dependent, the central concern is no longer the relative rigor of quantitative versus qualitative research. Rather, attention shifts to two major methodological issues that require careful consideration.

The first issue lies in logical coherence: how to avoid “methodological opportunism”. This challenges the rigor of the logic of justification: whether the reasoning for selecting a worldview, ontology, and methodology is internally coherent. If you treat worldviews as switchable, you must be able to argue why this particular research problem, at this point in time, requires this specific ontological commitment and methodological approach. It is not enough to say, “I want to study this, so I chose qualitative methods”, just as it is not enough to say, “I study leadership, so I will use a survey”. A rigorous justification would instead show why alternative approaches are logically premature or invalid. For example: because we do not yet understand what “leadership” means, and how its meaning is constructed in an emerging industry, imposing a pre-specified scale risks measuring the wrong construct; therefore, a constructivist stance and qualitative inquiry are necessary to elicit and theorize participants’ meanings before measurement becomes defensible. Here, rigor is demonstrated through transparency and argumentative force: reviewers should see these choices as deliberate philosophical decisions, not convenience-driven selections.

The second issue lies in logical integration: how to connect layered realities (mechanisms, events, and experiences), especially if your position aligns with critical realism. If you use mixed methods, you must avoid producing two parallel outputs that never meet: “machine data” and “river data”. Quantitative analysis may report an event-level regularity (for example: X and Y correlate at r = 0.40), while qualitative analysis may offer an experience-level account (for example: participants interpret X as leading to Y because of Z). The higher-order challenge is to explain how these two kinds of findings relate: what deeper causal mechanism could plausibly generate both the observed correlation and the lived reason Z. In this dimension, rigor lies in explanatory power. You cannot stop at reporting results; you must use abductive reasoning to move from surface appearances toward generative structures and deliver a logically complete theoretical model that integrates multiple layers of reality rather than a list of disconnected findings.

[1] Revised objectivist logic means that an objective reality is assumed to exist, but it is acknowledged that researchers can never know it perfectly and can only approach it through imperfect, testable, and revisable models.

[2] Further Inquiry: Why is this an “ontological judgment”?

To explain this thoroughly, it is necessary to strip away the methodological outer layer and examine the underlying worldview core. In simple terms, the concept of ontology investigates the nature of being. It asks: what is “reality” actually like? When a researcher instinctively feels that “quantitative is more rigorous and qualitative is not rigorous”, the person is not merely evaluating two tools. Rather, the person has already, at an implicit level, presupposed a particular form of reality.

From this foundational question, the ontological assumption at work is that reality is “a single block of machine” (realism). When quantitative research is felt to be rigorous, the implicit ontological assumptions typically include — (1) objective independence: the world (truth) exists “out there”, independent of consciousness. Whether or not it is observed, the Earth still turns, and temperature is still there; (2) singularity: there is only one truth; (3) measurability: since truth is an objectively existing entity, if the ruler is precise enough (a quantitative instrument), it can be measured. Under this ontology, quantitative research appears rigorous because it attempts to describe the objective entity precisely through numbers. By contrast, qualitative research then appears “not rigorous” because it relies on subjective experience, and humans change and are biased. It is, in effect, using something soft and subjective to measure something hard and objective, which seems unreliable. This is why, as a researcher, quantitative research can feel more “logical”: the logical starting point is an (post-)positivist ontology (the world as a machine).

Another ontology, however, holds that reality is a flowing river (relativism). If the ontology shifts (constructivism), the worldview changes accordingly — (1) social constructedness: reality is not a stone from a physics textbook, but meanings produced through human interaction. For example, happiness, anxiety, and organizational culture would not exist without people; (2) plurality: truth is not singular. What counts as high efficiency for a boss may feel like exploitation for an employee. Both are “real”, and neither is simply right or wrong. Under this ontology, quantitative research can instead appear not rigorous, but omission prone. It becomes like forcing cold numbers to define complex and fluid phenomena. When anxiety is defined as “4 points on a scale”, the concrete pain and background of a particular person’s anxiety are effectively amputated. From a constructivist perspective, this is a distortion of reality and an epistemic dishonesty. From this angle, qualitative research appears rigorous, because it acknowledges complexity and ambiguity, and enters each unique case to capture the truth of that moment. Seen this way, the earlier judgment can be understood as an ontological mismatch: a precise steel ruler (quantitative logic) is being used to measure a cloud (socially constructed reality). The problem is not the cloud (qualitative research), but the misalignment produced by the chosen measurement standard (ontological stance).