I trace much of the past 30 years worth of discussions on how the philosophy of knowledge applies to social science research to the work of Lincoln and Guba (1985). They used a three-part framework that connected epistemology to both ontology at a more fundamental level and methodology at a more concrete level.

For Lincoln and Guba, methodology referred to issues such induction versus deduction, which they wanted to separate from the more mechanical applications of specific methods, such as open-ended interviewing or survey research etc. From this point of view, epistemology centered on the nature of knowledge and methodology centered on how one goes about knowing -- so long as the latter didn't get into anything as technical as collecting and analyzing data.

The main point of their work was ultimately to elevate qualitative research (under the heading of constructivism) to an equal plane with quantitative research (underling the heading of post-positivism), By making their argument primarily at the ontological and epistemological level (the nature of reality and what what could be known about it), they could claim that the qualitative approach to knowledge was just as well-justified as a quantitative one.

By emphasizing a very traditional version of the philosophy of knowledge, Lincoln and Guba moved social science into the age-old debate between realism (post-positivism) versus idealism (constructivism). The pragmatist philosopher John Dewey characterized that debate as the basis for an "epistemological industry," which in effect provided plenty of jobs for philosophers, despite the lack of any likelihood for a meaningful resolution of the dispute.

From my perspective, a generation of social science methodologists have now also had an opportunity to find employment in the "epistemological industry," with results that mirror the lack of resolution in philosophy itself.

This is the problem of the philosophy of science. As for the methods, we have Milla canons, and deductions and induction as you state. We deduce on the basis of inductive conclusions created on the basis of observation sentences. In addition, we describe the structure of the phenomenon with the help of mathematical statements, because they allow us to describe the dependencies of which phenomena are composed, after substituting the appropriate object. Then we use them in the technique of making a microwave, for example. On the one hand, we know the phenomena and describe it on the other, thanks to that we produce technical devices. So far it is enough. Some people, however, are looking for the essence of the phenomenon, some metaphysical basis, but what we are doing nowadays is enough for us to do a computer, a microwave, a television. On the one hand we have learned the second pure science which is the task of cognitive activity. It derives from epistemology, which specializes in minor areas. In terms of biology, Konrad Lorenz gave an interesting method. There is still a need to separate methodology from methodics. Methodology considers the appropriateness of methods, methodics "collects" methods.

A very likeable question because of the broad introduction. We know: In some respect the question can be better than any expected answer. To reflect the history of modern scientific thinking of the last 200 or 300 years is quite worthwhile, I think. One milestone is the philosophy of Kant, who separated the empirical world as  phenomenon and the non-empirically ascertainably world as being. This allowed to separate experimental mathematical-scientific research from reflecting religion, metaphysics and speculation. But it was also an opportunity to rediscover metaphysics, speculative thinking,, human self-reflecting, that touches qualitative research often. We may differentiate quantitative and qualitative research but in the social sciences it doesn't help  on the path to any reasonable end, if you try to give quantitative research a higher status of value than qualitative. The method must fit to the particular problem. However there is no pure epistemology, there are several positions and basic attitudes. I prefer the critical rationalism as epistemological guide. So it is my intention not to verify expected or unexpected results  - you have to formulate your expectations before you are starting to use a research method - , but to see wether there is a chance to falsify ruling views. If you are successful in your thoughts then the tested proposition or theory becomes a bit more sustainably. So if you ask whether epistemology enables the researcher to apply a special method  (or not), I would like to say: Yes, of course: Principally try to maintain the opposite (regarding the view of others) by testing the statement which is in discussion with good counterarguments.  

I trace much of the past 30 years worth of discussions on how the philosophy of knowledge applies to social science research to the work of Lincoln and Guba (1985). They used a three-part framework that connected epistemology to both ontology at a more fundamental level and methodology at a more concrete level.

For Lincoln and Guba, methodology referred to issues such induction versus deduction, which they wanted to separate from the more mechanical applications of specific methods, such as open-ended interviewing or survey research etc. From this point of view, epistemology centered on the nature of knowledge and methodology centered on how one goes about knowing -- so long as the latter didn't get into anything as technical as collecting and analyzing data.

The main point of their work was ultimately to elevate qualitative research (under the heading of constructivism) to an equal plane with quantitative research (underling the heading of post-positivism), By making their argument primarily at the ontological and epistemological level (the nature of reality and what what could be known about it), they could claim that the qualitative approach to knowledge was just as well-justified as a quantitative one.

By emphasizing a very traditional version of the philosophy of knowledge, Lincoln and Guba moved social science into the age-old debate between realism (post-positivism) versus idealism (constructivism). The pragmatist philosopher John Dewey characterized that debate as the basis for an "epistemological industry," which in effect provided plenty of jobs for philosophers, despite the lack of any likelihood for a meaningful resolution of the dispute.

From my perspective, a generation of social science methodologists have now also had an opportunity to find employment in the "epistemological industry," with results that mirror the lack of resolution in philosophy itself.

Please let me know if these reference/site is helpful to you:

Methods in epidemiology and public health: does ... - NCBI - NIH

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1731834/pdf/v055p00104.pdf

by D Weed - ‎2001 - ‎Cited by 21 - ‎Related articles

follows that epidemiological methods, whether quantitative ... methodological standards and guidelines are ... subsets of Hill's list and assign diVerent rules of.

Please see attached file below.

Dennis

Dennis Mazur

What do you have to say about the impact that the debates on research paradigms have on the career growth of early career and mid career academicians because the university gatekeepers decide and implement the unwritten growth norms structured on the basis of their own ontology? The debate on research paradigms and research methodologies have created an atmosphere of hostility among academicians and researchers, which can be quite disturbing and shocking for early career individuals. And I say this from personal experiences. I do not want to run into the generalization of my observation; but I have heard similar incidents in many academic establishments. 

Epistemology marks the basis of every form of generating or building knowledge. Depending on which epistemological cradle centers to do research is equal to the coherence and scope of it.

http://www.macrothink.org/journal/index.php/ijhrs/article/view/11644

The real question is how do YOU match the aspects into a Research Design that works?

Pragmatism (Patton, 2015) suggests that almost anything could work as long as it does work for YOU for YOUR research in reality.

Best

Roland

I agree with David that much of the epistemology mixed into our debates has the main purpose of legitimising qualitative research. This might have been necessary, and might even be necesary today in some quarters. Unfortunately, however, it somewhat stunted the grwoth of the methodology we actually need, namely a framework that tells us which methods are suitable for which kind of research problem. We have some of that for quantitative methods but almost nothing for qualitative methods. I would really like people who champion specific qualitative methods being more forthcoming about the type of research problem 'their' method cannot be applied to.

I sympathize with Subhanjan Sengupta that there was so much conflict stirred up what were known here in the U.S. as the "paradigm wars." Here, that conflict has decreased since the 1990s, but there definitely is a generational influence -- especially in some fields. In particular, those who were trained as graduate students to emphasize epistemological issues are now professors who demand the same thing of their students.

As for pragmatism, I am a strong supporter of this approach, but I disagree that it depends entirely on what works for you. That strikes me as a form of relativism that ignores pragmatism's emphasis on communities of shared belief (at least inn Dewey's version of pragmatism). Actually in any version of pragmatism, you are free to believe whatever you want. but the impacts of your beliefs will occur when you take action on them. So, beliefs have consequences, and we usually go to create effort to anticipate those consequences in every major aspect of our life, including our choices about how we do our research.

The answer to what kinds of problems are not well addressed with qualitative data is relatively straightforward, and hence may not have generated much literature.  If you want to know how extensive some issue is, how widespread among a population, and if the actual proportions matter, then you don't wnat to use qualitative approaches, which are better at telling you what you need to look at and what it means, and considerably less good at how many or what proportion.  If the issue is one that can meaningfully be answered as "yes" or "no" as in some legislative or other voting situations, then you probably do not want to use qualitative techniques.  If you want to understand what yes or no means, or what positions are in between, then you do want more qualitative approaches.  

If you know for certain what the problem is that you are looking at and how people interpret it, then you probably don't need qualitative methods, which are time consuming and sometimes come up with unique results.  if you need truly reproducible results, then you probably don't want to use qualitative techniques, which uncover that which may be unique, in contexts that may not be easily reproduced.  

Short answer, if you want to know how much, how many, what proportion--how many children eat free school lunches--quantiative designs work best.  If you want or need utter reproducibility, you probably don't want qualitative designs. 

there are other possible answers here, but this much is pretty clear.

Irene

 Thank you all very much for your interest in my question and for your valuable contribution to this discussion!

Regarding "research paradigms" I think that again a beautiful philosophical debate like "science wars" was translated poorly in a methodological fight like "paradigm wars".

Saunders, Lewis and Tornhill (2015) could you give an essential orientation in their book Management Research for Business Students, chapter 4.

A quewstion for you: what do you mean by 'more scientific'??

Epistemology by itself cannot help to determine the specific methodological rules. However,  epistemology as one of the significant attribute of components in philosophy of science, particularly in social science plays a vital  role  in the sequential development process of the methodological procedure. Before any decision that can be made about specific methodological rules, one’s should concern, how epistemology links to Ontology and how it feeds the Axiology. Furthermore, consideration should be care in the next sequence as how to write about methodological procedure as well (Rhetoric).

For Gillian Fletcher:

It was "more" scientific in my question.

I meant insisting more on methodological rules derived from epistemology: the rules that make a research to be scientific and a discipline to be a science.

For example, "little things" like making sure that the categories in an emergent coding scheme are exhaustive and exclusive. Or "big things" like stating clear what is the relation of the research outcomes with the existing theories: did the research tested an old theory, modified or completed one, developed a new theory etc.?

You can add an answer to my question now (like Bahram Shahedi did) or you just wanted to be mean?

Gillian was not just being mean, the question is important, because the term "scientific" makes it sound to many as if quantitavite measures were science,and hence truthful, while qualitative studies are anecdotal and hence, neither scientific nor valid in anymeaningful way.  Some have striven as a result to try to make qualitative work more scientific, such as counting the occurrence of some particular words or phrases, or measuring the duration and frequency of pauses.  Coding need not be exhaustive to be meaningful, in my experience, coding is not exhaustive, it depends on what you are looking for, there is always some meaning, some levels of meaning, left on the table.  Data can be coded and recoded many times, with different questions in mind.  The notion of science can be misleading here, hence Gillian's question.

Qualtiative work often provides its own source of validity, which is context, and different sources of similar information, all usable, looking at a phenomenon in three dimensions, walking around it, past present, and surroundings.  These checks are often missing in more quantitative work, which therefore has to struggle with more formal means of validation.  

 My answer to your question is yes, the way you view what knowledge is and how it can be gained, what meaning is, does influence what tools you are likely to use to do research.  Relatedly, personality matters, and the ability to tolerate complexity and uncertainty.  While philosophy of knowledge matters, it is not determinative of methods, but linked more loosely to sets of tecniques.

Irene

I'm not sure whether the original intent of the phrase "more scientific" was to equate being scientific with the standards of quantitative research. Going back to the paradigm wars, a major part of the reason for bringing philosophy of science into discussions about social science methodology was to open the question of what it means to be scientific. In particular, if there are various versions of how ontology and epistemology go together, then the approach to research we refer to as postpositivism could not have a monopoly on what it means to be scientific.

I personally consider that approach to be a smokescreen that primarily generates arguments about what it means to be scientific, and whether being scientific is somehow desirable. Instead, I think any given piece of research should stand on its own in terms of whether it is meaningful and appropriate. Plus, we need to recognize both the importance of community standards and the openness to change of such standards. So, what we judge to be a "good" (or "scientific") piece of research at one point may be judged as inappropriate in some later context.

Thank you Irene and thank you professor Morgan!

I think that the split of the social research methodology in quantitative and qualitative was a mistake. I think that according to our research questions we must be free to choose any methods and techniques to be employed without worrying about what "research paradigm" it will be, about not counting even it could be helpful, about mixing-methods in certain sequences etc.

I think all we have to worry about is how to make our research to be scientific. If we don't produce scientific knowledge we don't talk about science anymore but about art or phylosophy.

I don't think that "qualitative research" is not scientific. Interviews are as scientific as surveys, clasifying with a priori categories is as scientific as clasifying with emergent categories, discovering theories is as scientific as testing theories etc.

I found in Gauch (Scientific method in practice) some basic rules of the scientific method like applying deductive and inductive logic etc. I think that this is what we have to look for in epistemology: to make sure that in every step of our researches we are making science!

I'm still not convinced that calling our work "scientific" adds to its value. Yes, we want our work to be high quality, but that is up to colleagues to determine -- and if they disagree, then we need to find ways to assert our point of view. But the idea of being scientific strikes me as something that is highly valued in most of our cultures, so we can borrow from that "halo effect" to increase the prestige of our own approach to generating knowledge.

Professor Hein Retter

I liked reading your comment. It was good logic. Can you please elaborate a bit more on what you meant when you said: Principally try to maintain the opposite (regarding the view of others) by testing the statement which is in discussion with good counterarguments. 

Professor Roland Berberich

Pragmatism does talk about the acceptability of methodology as long as it works out for ME and My Research. But, that is where all the problem is. In academic fraternity, there is the need of acceptability of your research methodology by OTHERS; and that’s where the university evaluates the credibility and rigor of your research. What about that?

Professor Jochen Glaser

I do agree that there is need to specify what are the research problems where qualitative research cannot be applied. But is it clear that subjects which are already at an advanced stage of theory development, for example welfare economics, needs large scale quantitative inquiries for statistical generalizations; whereas nascent areas of inquiry, for example mental accounting, would need more in depth qualitative inquiry (such as case based or ethnographic).  

Professor David L Morgan

The paradigm wars have decreased in the US; and if I am not wrong, in Europe as well. But here in India, it is still a problem. There is still a thick wall between the quantitative and the qualitative. I hope this fades away here as well.

I also liked your comment on what means to be scientific. As a young researcher and academician, I have seen paradigm wars among senior academicians in doctoral committee meetings and doctoral defence sessions, which has left me pondering (almost every day) on how to deal with this. Why do we waste so much time on justifying what is ‘scientific’ (typically synonymous to being quantitative) and what not? My supervisor says, it does not matter whether you are a positivist or an interpretavist, whether your research is qualitative research or quantitative research; what matters is whether your research has ‘quality’. I am glad that he says this. But I do not know how to deal with this debate in the long run. I wonder how long will it take for research to move from being ‘tool-based’ to being ‘issue-based’. I agree to you when you say: “we want our work to be high quality, but that is up to colleagues to determine -- and if they disagree, then we need to find ways to assert our point of view.” But what to do when we are facing fascists who perceive statistical and econometric research as the God of all things.

Professor Irene S. Rubin

I agree to the fact that qualitative research is time consuming and sometimes come up with unique results. But that is the beauty and strength of qualitative research. I think this whole trend of generalization of results across populations, which started a few decades ago, has a basic fallacy of ignoring the fact that no two human is alike, no two community is alike, no two country is alike, no two issue is alike. Every day a context is full of unique incidents that go ignored, non-recorded, and non-researched. In fact I must say that unique results have the power to challenge the universality of research findings. The very approach of discovering and accepting the uniqueness of our lives, which are themselves subjective realities based on our social constructions, is a ‘naturalistic inquiry’ that should exist in parallel to the ‘rationalistic inquiry’ of positivist research. And I believe these two approaches are symbiotic, and not absolutely detached from each other. On this note, I found your perspective on validity interesting. I am interested to know more on what you mean by ‘formal’ means of validation. What makes the validation procedure of qualitative research ‘not formal’?

Professor Florentina,

Agree strongly to you when you say: I think that the split of the social research methodology in quantitative and qualitative was a mistake. I think that according to our research questions we must be free to choose any methods and techniques to be employed without worrying about what "research paradigm" it will be, about not counting even it could be helpful, about mixing-methods in certain sequences etc.

A few weeks ago, I had started writing on research paradigms; more spontaneously rather than in a planned and systematic manner that one would do in scholarly writing. This thread of discussions made me go back to that file today and give a shape to my thoughts on the issue of methodological dispute. I would be keen to know your criticism on this working paper. It is a work in progress. 

Working Paper THE FUTILE WAR BETWEEN QUANTITATIVE AND QUALITATIVE

By formal, I mean only that it is explicitly and usually quantitatively tested; in qualitative work, validity is not really a problem that has to be demonstrated, it is assumed, the data is transparent, its validity, its meaning, is the core of the research, it requires no further testing and certainly not quantitative testing.

For a change, I will give a simple answer to this question.

The answer is that you are conflating epistemology with a paradigm. Any paradigm consists of four elements: ontology (the understanding of what is; e.g., there is a world outside of our consciousness vs. there is only the world we are aware of), epistemology (the understanding of how we might come to know that which is; e.g., we know the world only through our consciousness vs. we know the world as it is), methodology (the practices to be employed in order to know what is; e.g., we may know the world only by observing it vs. we may learn of the world by reflecting on it), and axiology (the values that should guide our investigation of what is; e.g., knowledge should serve the better good vs. knowledge should be objective and indifferent). In a proper study, these elements are complementary. In many studies, unfortunately, they are not.

I recommend Guba and Lincoln's work (1985, but also the relevant chapters in the Sage handbook of qualitative studies). These are great for understanding what methodologies are compatible with what ontological and epistemological perspectives.

With regard to Axiology, it is not mentioned in Lincoln & Guba (1985), nor is it part of the philosophy of knowledge. Instead, it was a later import into the Lincoln & Guba system, to cover the fact that a pure version of the philosophy knowledge does not cover some of the most crucial issues in social science concerning ethics and values.

Most work on philosophy makes a strong distinction between epistemology ("the known") and ethics ("the good"), and the idea of axiology is almost unknown in that field.

To me, this is another sign of the overall weakness of the Lincoln & Guba system.

I took another look at Lincoln and Guba 1985 and they do mention axiology, but not as a defining characteristic of paradigms. That doesn't happen until (I believe) the third edition of the Denzin & Lincoln handbook.

I agree with David Morgan that Guba and Lincoln defined much of the debate about epistemology (and ontology) in the social sciences (see for example, Snape and Spencer (2004). However, I have serious issues with some of their assumptions, most importantly the relatively unchallenged acceptance that quantitative studies are 'post-positivist'.   

Post-positivism was defined in contrast to positivism. The fundamental premise of positivism is that the proper purpose of science is restricted to direct observation and to those things which are verifiable. B.F. Skinner was a positivist (Lindzey, Hall, and Manosvitz 1959). Post-positivism emerged when scientists began to question some of the rigidities of 'positivist' approaches. It  is a more nuanced approach in which it is recognised that any observation affects the phenomenon being observed and the observer brings inherent biases to the observation and its interpretation.  So all data is incomplete, contingent and understood in the context of the observer's interpretative frame (and theory).  No data can or should be treated as absolute and must be validated by multiple means before being accepted as an approximation of reality.  Post-positivism is still inherently a scientific approach,  premised on the scientific method of observation, experimentation and the development of theory. 

Post-positivism (and survey research)  is often classified as deductive (though that is more true of Popperian approaches) as opposed to inductive.  However, science is not necessarily deductive,  Darwin did not have a theory when he did his observations, Newton 'discovered' gravity from observation not the other way round. 

In the social sciences, some experimental and 'quasi- experimental designs do have some of the characteristics of  scientific method and have an underlying 'positivist' or 'post-positivist' epistemology.  From a formal perspective there are problems with such approaches in that they do not control for independent variables (such as differences between service deliverers, clients, social and economic context etc.) but that is another issue.

Most surveys, on the other hand, should not be classified as 'scientific' or post-positivist.  Science (and post-postivism) relies on direct observation, isolating dependent variables from context effects, repeatability, and understanding, or at least having a solid theory about, the mechanisms that produce the results. Most surveys do none of these things (nor indeed do most qualitative methods but they are not generally - pace Malinowski and others - presented as 'scientific'). 

Surveys do not directly observe the phenomena they are studying, especially when they attempt to measure internal mental states such as attitudes and emotions. Instead they measure answers to questions about the phenomena being studied.  Such answers are then treated as if the answers are inherently a more or less accurate reflection of behaviour, experience, understandings, etc.  Inconsistencies in the data and variations between data and behaviour are treated as 'bias'.

There is usually only limited testing of external validity except when reality forces a test (e.g. election results) and almost no understanding or testing of the cognitive mechanisms that people use to generate the answers (pace Schwarz, Tourangeau and others).  The few studies that look at cognitive mechanisms generally use outdated and discredited cognitive models; and are almost never referenced in the reporting of survey data.

Further, in contrast to psychological inventories and intelligence tests, most answers to survey questions are treated as if each answer occurs in isolation from the other answers given by the respondent. We are told that x% of people answered Y to question Z.  Each answer is  taken out of context of the context of the individual who provided it and reified as separate data in its own right. Such analysis ignores the evidence from psychology and cognitive science about how people develop responses (see Kahneman 2011).  It is not scientific, or even post-positivist, it is reductionist.

That was a rather long-winded way of saying that the debate about epistemology and ontology in the social sciences is based on some fundamental assumptions that should be scrutinized more carefully. A better understanding of how answers are generated is crucial to improving our understanding of the answers we are given and how those answers relate to the phenomena we are studying.

Great answer, David Glyn Roberts! Thank you!

I agree with you! I am not much of a diplomatic person. I think in many of the methodological writings we have a pretty big epistemological mess!

@David Morgan and Fiorentina Scameci-Dommisoru

I entirely agree with Lincoln and Goba (1985) when you say that, from their point of view, ‘epistemology centered on the nature of knowledge and methodology centered on how one goes about knowing’. However for me methodology is not a branch independent from epistemology. Epistemology is the equivalent of Philosophy of Science. The last term suggests that the sciences are the focus of the investigation, while epistemology is concerned with the nature of knowledge in general. In fact, philosophy of science is, more often than not, concerned – as is epistemology – with scientific knowledge in general, discussing for instance the nature of scientific theories, the ways of their validation, the role of experience in scientific investigation, the meaning of probability, etc. Many works in the philosophy of science refer to the sciences only in order to illustrate by examples the ideas put forward. Other philosophers of science try to give a new insight by a thorough examination of the scientific practices. Following this course, we emphasises the importance of methodology: a close examination of the methods in use allows us to bring nearer to each other the research practices and the philosophical concepts.

Robert Franck, as philosopher, and me, as social researcher, are the editors of the Methodos Series, published by Springer. This book series is devoted to examining and solving the major methodological problems social sciences are facing. Take for example the gap between empirical and theoretical research, the explanatory power of models, the relevance of multilevel analysis, the weakness of cumulative knowledge, the role of ordinary knowledge in the research process, or again the place which should be reserved to “time, change and history” when explaining social facts. These problems are well known and yet they are seldom treated in depth in scientific literature because of their general nature. However the epistemological scope of these methodological problems is obvious and resorting to Philosophy of Science becomes a necessity. The main objective of the Series remains however the methodological solutions that can be applied to the problems in hand. Therefore the books of the Series are closely connected to the research practices.

In the attached file you will find a short presentation of the volumes already published, which will give you an idea of the problems treated and their links to methodology and epistemology.

I have to disagree with David Glyn Roberts characterization of survey research. First, survey research is just as subject to replication (repeatability) as experimental research. Second, high quality survey research typically involves regression models and other controls for correlated independent variables (although opinion polls often report simple percentages).

As for studying mentalistic constructs, that would certainly violate the assumptions of classic positivism, but I am not clear on where post-poisitvism would stand on that. Plus, I assume that problem would apply to almost any kind of self-report data, so it is hardly limited to surveys. 

Also, operationalism is later development than the earlier forms of post-positivism, and it is a key element of survey research, which maintains a strong program of research on survey measurement. In particular, practices such as Structural Equation Modeling go to great effort to produce replicable measures of theoretical concepts. These efforts at objective measurement would hardly qualify as "constructivism."

I suppose a philosopher could argue that the work by survey researchers fails to meet the standards of postpositivism, but survey researchers themselves certainly do pursue those standards and make extensive efforts in that regard.

Thank you to illustrious colleagues for your contributions to this discussion which I'm finding very informative.

The thing that strikes me is that the label "science" is (paradoxically) emotive, even political, and not terribly helpful.  Reading contributions I've got the sense variously that it might (or might not) be: paramount in the creation of knowledge; synonymous with quality; that it is post-positivism; or that post-positivism is not science.  If nothing else this seems to provide evidence for the claim that the concept itself is socially constructed. 

For me an interaction with epistemological thinking is a powerful route to ensuring quality in research (whatever paradigm it might be considered to fall into). Sure we can quite legitimately shortcut that and fall into consensual 'packages' of ways of doing research - not all knowledge claims need to trawl the history of philosophy - but if we want, as it were to 'stress test' methodological choices, then epistemology seems a very useful sounding board.

Am I being naive?

I was earlier a physical science researcher. Post research, I served industry for long years and found that the positivist approach did not work. In fact qualitative research gave better result for sales forecasting. These methods included talking to every dealer and taking their sales expectation. This, to me, was a scientific method. I do not understand why Florentina is doubting qualitative research in social sciences. What is non-scientific in it?  i am further foxed with the terminology "more" scientific. Can she share her real problem so that I could help.

I obviously was not clear enough in my earlier post.

To take the last point first, I was not suggesting that survey research was constructivist.  If I gave that impression, it was poor expression on my part.

Nor did I mean to  suggest survey research could not be scientific, or even post-postivist, just that much of the survey research reported does not meet the standards I was taught for scientific research.

For example, I did not mean to suggest survey research could not be replicated, it certainly can be.  However, much of the survey data reported  does not treat the differences between survey results as an issue to be explored. The most obvious example is election polling, which generally does purport to be scientific.  The differences in results across surveys (and even from voting results) are assumed to be either the consequence of bias; or changes in voting intention.  There is very little effort expended on exploring why the results are not replicated perfectly; or the underlying mechanisms that generate such variations.

And yes, my comments do apply to all self-reported data (see for example Bertrand & Mullaninathan 2001 "Do People Mean What they Say …" Economics and Social Behavior, Vol.91., No.2) ). But it applies particularly to data that is analysed out of the context in which the responses are provided.  Isolating and reifying responses to each question as data that is somehow separate from the context in which it is elicited either assumes the context is irrelevant or that the context is controlled in ways that prevent that context from affecting the responses.  I contend that such an assumption is erroneous and not based on the available evidence.  For example,  recent research (Arnulf JK, Larsen KR, Martinsen ØL, Bong CH (2014) Predicting Survey Responses: How and Why Semantics Shape Survey Statistics on Organizational

Behaviour. PLoS ONE 9(9): e106361. doi:10.1371/journal.pone.0106361) shows that the pattern of an individual's responses to a questionnaire are based largely on the common meanings of the language used in the questions rather than reflecting differences in the questions.  Other research shows that chance is a better predictor of behaviour than self-reports (see Falk et al 2012 "From Neuronal responses to Population Behavior …"  Psychological Science 23(5) 439–445).

I am not familiar with operationalism so it may well address the concerns I raise, I leave that to others to discuss.

It seems I was not clear enough either.

Again: I think that we use epistemology in a wrong way. We give the impression that it provides specific methodological rules like using some techniques or not using some tools, like combining methods etc. For exemple when a researcher say he/she choose pragmatism or realism or other ism it means that his/her research will be mixed-method or that will be use questionnaires etc. But do you know any epistemological writing stating that in ethnography you cannot count or that in a case study you cannot employ questionnaires? I don't think so! We invented strange methodological rules using epistemology as an excuse.

We also use epistemology as a basis for a useless methodological fight between "research paradigms". But this is another discussion.

I think we have to use epistemology as a control agency for our research quality. If we want to make science then we have to do scientific research (after we decide what it is according to epistemological writings). I think social research (qualitative and quantitative) is or could be scientific when follow some rules. I think that a researcher' job is to make sure that he/she really produces scientific knowledge like applying logic, working with probabilities or like paying attention to biases, ensuring reliability etc.

I think the qualitative researcher has to be exigent with his/her work too. For exemple when doing thematic coding why do we say that 4-6 themes are enough? Is it scientific making such an arbitrary rule? How about making an exhaustive theme inventory by reaching saturation, by collecting more data, by checking the themes with participants etc. Isn't it "more" scientific?

Or when we employ semi-structured interview where do we get questions from? Do you think two different researchers having the same research question will ask the same interview questions? If not, is it scientific?

We can go on like that with many other research issues that need to be scrutinized and improved in order to make our researches "more" scientific.

With regard to social scientific use of surveys, I would not confuse that with election polling, which often does seek to produce a single, context-free percentage. And, as I mentioned, survey researchers have a long tradition of doing research on issues such as question wording. Beyond that, the issue of how attitudes relate to behavior also has a long research tradition, but that is a substantive question, rather anything related to survey research itself.

But my major point is that most post-positivists would have no problem identifying survey research as fitting within their assumptions, and most survey researchers would have no problem identifying with post-positivism. Whether survey researchers are successful in their aspirations (especially when working with attitudes or self-reported behavior) is another matter.

With regard to being "more scientific," I would interpret almost everything that Florentina said as falling into a call for more "objectivity" (or a least less "subjectivity"). I would say instead that we need to be as explicit as possible about our procedures and our justifications for using those procedures to meet our research goals. For me, this is a matter of doing good research, without regard to whether or not it is scientific. Indeed, I personally identify as a "social researcher" rather than a "social scientist."

of course there´s a (unshakeable) relation between an epistemic point of view and the method employed to prove its assumptions. A  epistemic position dictates ´´what to see´´ and the method ´´how to do it´´ . Both episteme and methodus are inseparable. We see this precisely in the process involved in proving a certain hipótesis (that has to emerge from a epistemic concern).

For example: if you are interested in a phenomenological point of view, regarding peoples personal opinion (doxa), it makes no sense to use a quantitative or numerical approach.

on the other hand, if you use a marxist or in a board sense structuralist episteme, regarding peoples unconscious relations (honor, prestige, domination, power positions, etc.), it would not be correct to use interviews as the principal interpretative approach (because of a structuralist-epistemic implication: people are not aware of their social determination).

Phenomenology is a philosophy. I don't think you will find any phenomenological writing mentioning interview, non-numerical approach etc. I also think that those who claim they are doing 'phenomenological research' hardly ever read any phenomenological (philosophical) paper.

We ask people about their experiences in interviews but also in questionnaires. We do not get the objective reality through questionnaires and I think that we do not make any mistake if we count in a phenomenological approach how many subjects experience the same thing.

I think the determinants could be find by the researcher in interviews or in documents too. For example in a visual study my subjects were asked to draw social classes. I found unconscious relations too like "the perceived level of inequality determines its acceptability". See: https://www.researchgate.net/publication/284912714_Acceptable_and_unacceptable_sources_of_inequality_in_Romania_A_visual_study

I think we developed too many restrictive methodological rules misinterpreting philosophical thoughts.

Article Acceptable and unacceptable sources of inequality in Romania...

Hi!

My research is mainly focused on palliative care and philosophy of science, specifically regarding the epistemology of qualitative research. I agree with David L Morgan, it is complicated. Consider that all the authors above responded well to the question of ontology and epistemology. Therefore, I do not need to go into that. I also think like you Florentina, that those who claim they are doing 'phenomenological research' hardly ever read any phenomenological in depth.

I also think we (or some of us) invent strange methodological rules using epistemology or philosophy of science, as an excuse and that  makes us create bad research quality. Some I have seen, are due to missreadings of philosophy of science!

However, Just to tell you, even in qualitative studies you can test theory, although it is unusual. I give you the reference to a study on the issue I published in 2015 in nursing science. The quantitative field has its limitations. With a quantitative method, you can establish a relationship between factors by verification, but you cannot create a "strong" scientific theory. So, what are aim? Is it to create a theory or is it to create a "relationship between factors?

Philosophy of science can provide you with tools to see how you can perform high quality scientific studies. They do not have "rules" that you should follow blindly, with their tools, your thinking can be developed, in depth.

Bergdahl, E., & Berterö, C. M. (2015). The myth of induction in qualitative nursing research. Nursing Philosophy, 16 (2), 110-120.

Best Regards

Elisabeth

With regard to Ricardo Silva's statement that "of course there´s a (unshakeable) relation between an epistemic point of view and the method employed to prove its assumptions." I hope that this only applies to the idea of starting with an epistemic point of view and then selecting a method to match it -- because it certainly doesn't apply in the opposite direction. In particular, you could be either a postpositivist or a constructivist and choose to use open-ended data analysis to collect the data and thematic analysis to analyze. it.

In that case, you would have used identical methods to collect and analyze your data, and all that would differ is your underlying philosophical assumptions. Which certainly leaves me with a big "so what"?

Dear RG Members

I Totally Agreed to David Morgan Message

Thank you

For example Bacon offere us, 

To study the quality of objects with their mutual change, such as: 1depth, 2depth, 3will, 4dest, 5destination, 6depth, 7depth, 8destination.

And so eg A) if it changes 1 it changes 6

  B) if it changes 8 it changes 8

etc. etc

Dear Florentina, in 2002 I was thinking about methodological issues and discourse and about which method would be consistent with research conditions and resources in developing countries.  I decided then to share ideas on this publishing the article I am sharing below.  I trust you and friends in the discussion can use it as food for thoughts.

Non-Traditional Research Methods and Regional Planning Needs in Developing Countries: Is there an Ideal Methodology?

https://www.researchgate.net/publication/26422725_Non-Traditional_Research_Methods_and_Regional_Planning_Needs_in_Developing_Countries_Is_there_an_Ideal_Methodology

Article Non-Traditional Research Methods and Regional Planning Needs...

Interesting statement. I think something is right that there are no definite rules other than the researcher's beliefs about the truth that is done in the research prisedur. Of course the truth comes from the researcher's honesty. Honest researchers in using any method (quantitative, qualitative, and merging both) can be ascertained that the research done becomes a truth. That is my opinion, thank you.

I understand Subhanjan's dilemma. There seems to be much 'ado' about almost nothing. Much conflict was stirred up by social science researchers which were known in the U.S. as the "paradigm wars." There, that conflict has decreased since the 1990s; it seems to be a generational influence -- especially in some fields. In particular, those who were trained as graduate students to emphasize epistemological issues are now professors who demand the same thing of their students. But in developing world where first generation of doctoral students, who were given in depth courses in epistemology are now coming up as supervisors. They find it difficult to make the old guards believe social constructivist approach. Tha qualitative-quantitative paradigm war is at its height but will surely lower down as more and more new generation supervisors come up.

Dear Florentina Scârneci-Domnișoru;

I've only just come back to this thread (for some reason I either wasn't alerted to, or didn't pay attention to, the responses generated). Please be assured I had no intention of being mean. Good luck with your work!

It depends how you look at the research and what is the purpose of the research. If it is a sponsored research, the sponsorer calls the cards. If it is purely an academic research, the researchers have certain degree of freedom. I pesonally believe that it is the research question and the researcher's beliefs about the truth that should prevail.

Epistemology determines the kind of knowledge to be had from taking a particular view of reality (Ontology) and thus is suggestive of the methods applicable to obtaining this particular kind of knowledge (Methodology).

I agree with Philips in stating that Epistemology determines the kind of knowledge the researcher wishes to acuire but which view he/she will take depends upon his/her view of reality (Ontology) which helps in deciding the methods applicable to obtaining this particular kind of knowledge (Methodology). Having said that, I repeat that the research question and the researcher's beliefs about the truth should prevail

Networks/discourses/practices are neither true or false. What are the conditions of possibility of epistemologies?

Kant separated the empirical world as  phenomenon and the world as being which allowed the distinction between experimental mathematical-scientific research and reflecting religion, metaphysics and speculation. Social science is a science though not an exact science like physics or chemistry.

Though, I believe in methodological rules, the determination of the methodology should be based on the research question rather than the epistemological belief of the researcher.

I have found this debate to be truly fascinating. I am a social researcher and not scientist and I feel that the debate comes about because in the main the positivist and phenomenological stances overlap. That is why the debate is whether the cart comes before the horse or the other way round? I am doing a qualitative study and I have stopped may times to ask the question- is this quantitative or qualitative? Some researchers have argued that these are the same and differ in whether your interpretation is in numbers or words? The answer seems to be in that the design based on the research question must fit the concepts and the depth requisite in making theory.

Essentially, your research design will be that which best leads to answering your research question.

Manuel,

perused your article; must confess that it seems a bit 'weighty' for my taste as a social researcher; however, to address the question, 'Who is telling the truth, nature or man?', I'd say that of course, only nature really tells the truth; nature is the 'real deal', while man searches for the 'truth' and works with interim perspectives / approximations, quantitative or qualitative; probably, that's all we can ever do.

I am of firm opinion that Research is research. Methodology, methods, tools and techniques are just help conducting the research - be it quantitative or qualitative that is decided by the Research Question.

Well, "men of society" are in search for truth within nature as Adams clearly pointed out. However, i should critical realism appears to be robust in being between the two methods; quantitative and qualitative if the truth is in nature and men are rotating around it to find the phenomenon of truth?

late response to a much earlier post...re survey research. i imagine there are few survey researchers who would acknowledge the subjectivity of what they do--in terms of framing both questions and answers that limit what respondents can say, as well as more technical decisions in data analysis. If this conjecture is true, then survey researchers miss one key aspect of post positivism.

In the survey research, the researcher practically dictates the answer to the respondents even though she gives options. Thus, truth may not come out in quantitative research.

Dear friends, have you ever thought about a reality that is based on codependent variables instead of independent ones? The shift to green markets is a shift from independent variables to codependent ones.....

Good day Manuel, I am talking about codependent vrs independent choice reality, independent choice reality allows you to maximize; codependent choice reality requires you to optimize.

Some strange turns in the debate ...

@ Arunaditya:

I don't think "dictate" is the right term here. Standardised questions limit answers to choices between investigators' understanding of the question, which of course is a severe limitation.

At the same time, interviewees in qualitative interviews have been known to answer the same question differently depending on the context of the interview. Where is the truth in that? Of course, it is just another methodological limitation.

I think the main question is not about what respondents or interviewees tell us but about what truth we are able to reconstruct from their answers through an analysis of what they tell us under specific circumstances.

@Lucio:

I totally agree about reality consisting of co-dependent, interdependent, whatsoever factors. The interesting point about research is the difference between ontological reality and analytical strategies based on epistemological choices. We can try to understand codepentent variables by analytically decomposing them into independent-dependent pairs, finding out about them, and then reconstruct co-dependence from what we have learned.

It is important to distinguish between reality and our strategies of investigating it.

OK. The can of worms has been tipped over. I think one of the fundamental issues are our ontological and epistemological assumptions about the nature of reality and what we can know about it. There are many different approaches with different combinations of assumptions and different variations of emphasis. I won't go into them here except to briefly describe two of the more common but contrasting positions.

A lot of research starts with the premise that there is one 'truth' or 'reality' and that it is ultimately knowable and understandable if we analyse and remove obfuscating factors. It often, though not always, assumes that complexity can be reduced to patterns, or even rules (see Parsons). There is a tendency to assume that such behaviour (including verbal behaviour) and patterns are consistent and stable across time and contexts. Such explanations are often treated as partial and contingent.

Some other research assumes that there are multiple realities depending on the perspective of individuals and that each reality is valid in its own right. Much of this kind of research also has an inherent assumption that such 'realities' are, at least to some extent knowable, or we would not be conducting research. Our knowledge of such 'realities' is considered to be partial and contingent but still worth doing. Why? Inherently, and often not explicitly, because it can help understand how individuals behave and interact with each other, mobilising conceptual resources in their interactions: but also because we believe can extrapolate from that knowledge to others. Even within this perspective, there is often, though not always, an assumption that there is one reality for each person; that such 'realities' are 'knowable' because they are relatively stable and consistent. Again a key task of such approaches is to reduce the complexity to develop a narrative that 'explains' such realities.

I think both sets of ontological and epistemological assumptions miss the point. Social reality is complex, changeable and context dependent. The fact that people give different answers to the same questions depending on context is not a methodological limitation that can be addressed by greater rigour; but a fundamental reflection of reality! People behave very differently in different contexts and at different times, in large part due to their perceptions of the situation. Individuals will even perceive the same situation differently at different times, observing some things at one time and other things at another. People even remember differently depending on context! Indeed politics is all about manipulating people's perceptions to generate changes in behaviour. We need to recognise that context, moderated by our perceptions of context, is a crucial factor in generating behaviour. A survey or interview context affects our verbal behaviour and such verbal behaviour may bear little relationship to our physical actions in the situation of interest.

There are research approaches that seek to explore and revel in the complexity of social reality; and they, too, ultimately try to understand the dynamics of behaviour but they do not assume that the answers to a necessarily limited number of questions in a specific interview or survey context provide sufficient data to draw conclusions about the complex and dynamic 'reality' of people.

I perfectly agree with David. Especially on the context, time links with people behaviour. However, i think the problem of ontological and epistemology in terms of limitation is the tracks in timing. Of course context and time phases are not to be treated in isolation.

Truth or Reality is not absolute. Some believe it to be one and absolute while others believe it to be multiple and different. Realities could be multiple depending on the perspective of researchers. Each reality perceived by the researcher may be valid in its own dimension. Quantitative research considers one reality while qualitative research may provide opportunity to consider multiple reality.

And I suppose reality in itself can varied in qualities? Perhaps people need to see these dimensions of truth then?

Arunaditya Sahay

david glyn roberts says well what I was going to say. I vote for his answer. The fact that two people answer qualitative interviews differently is not a shortcoming or limitation of the methodology, but actually a way of recognizing the complexity of what we consider truth--we do not demand that the vision in a kaleidescope be based on all the pieces being of the same color and shape. The fact that two people react to the same stimulus differently is a major part of our political world--two people react to Trump's tweets differently. That is reality...and we need to see it and understand it.

Irene

Manuel, I found your follow up comment to my second comment a little bit curious as you seem to understand exactly why I said, but I will add a little bit more here to see where you are going.....

1) Reality under independent assumptions allows you to maximize

M = X + Y

*You can maximize any component X or Y at the expense of the other

*Additive thinking based model

*example, It allows for dominant structures such as the traditional market where only the economy matters or socialist markets where only society matters

*exclusion is the rule

2) Reality under codependent assumptions requires you to optimize

M = XY

*You can not maximize X or Y, you must optimize

*System thinking based model

*example, it allows for codependent structures such as green markets(where both the economy and the environment matter) or sustainability markets where all the economy, society and the environment matter at the same time

*Partial or full inclusion is the way to go

That is what I meant Manuel.

Have a nice day!

Sounds like we are descending into the realm of philophical game playing -- like asking whether the statement "There is no such thing as truth" is true or false. (And I admit to playing the same game when when my qualitatively oriented students assert there is no such thing as objectivity, which they clearly believe is an objective truth.)

This sort thing is a reason why I prefer pragmatism, with its emphasis on the nature of human experience, rather than abstract arguments about metaphysics.

Yes, I think we are "..descending...". Thus, to curtail this descent, I address Florentina's question directly: epistemology offers methodological options (and see my previous post).

I do agree that abstract metaphysical arguments can ' go on forever', but this needn't be so; curtail them by taking an ontological/epistemological position---then leave the debating chamber and take action---'research 101'.

Let's be more practical: please give me a reference, an epistemologist (philosopher) who wrote something about social research, who mentioned something about surveys, qualitative analysis or any other method or technique.

I think we erroneously use epistemology to justify some questionable methodological options and rules.

All I am saying is that one takes a perspective on 'reality', and then further, devises a strategy to 'find things out' in this reality; for instance, taking the position that 'reality' is socially constructed, one takes this into account in one's research.

All researchers do this, whether they choose to acknowledge it or not---period.

I disagree with Phillip's claim that all researcher's take a position either implicit or explicit, on the nature of reality (and just to be ironic, since he obviously assumes that this statement is true, he has just labelled himself a implicit realist with regard to epistemology).

I myself think most researchers are thoroughly agonistic about the nature of reality -- certainly I am as a pragmatist. And as far as I'm concerned, those who claim that everyone has to have some kind of hidden views on the nature of reality are just making assumptions that justify their own favored approach. The pragmatist philosopher John Dewey referred to this as the "epistemological industry" -- a self-perpetuating justification for a limited approach to the philosophy of knowledge that has persisted for thousands of years without reaching any useful conclusions.

I think that we need a certain amount of philosophising (but not too much that it substitutes for taking action; just enough, as part of establishing a transparent audit-trail, such that our work can be subject to comprehensive peer critique) in order to understand the mechanics of doing research.

Must admit that up till now, I have never thought of my philosophical position regarding research, as either 'this' or 'that'; however, on consideration that I am an implicit realist----ok; I do however, also subscribe to a causal criterion of reality-------'things' are regarded as real if they can influence our behaviour----which allows me to be a constructionist as well.

And of course, one can argue ".. those who claim that everyone has to have some kind of hidden views on the nature of reality are just making assumptions that justify their own favoured approach"; but in making said claim, I wasn't actually thinking about justifying my approach; rather, I just cannot see how a researcher can do any piece of research at all without considering the nature of reality, the kind of knowledge to be had from this conception of reality, and thus, the actual methods to be used in the research.

Thus, I can only conclude that pragmatists unwittingly 'telescope' the practical philosophical underpinnings of research to the detriment of providing a transparent audit-trail of their work, such, that it can be properly critiqued by their peers.

Pragmatists simply deny the relevance of the traditional version of the philosophy of knowledge. Instead we wish to focus on human experience, while ignoring unansweable — and hence irrelevant— metaphysical questions

I agree with the pragmatist position of 'getting on with the research' rather than taking part in debates concerning ''...unanswerable---and hence irrelevant---metaphysical questions.. .''; but I think that we can descend into making these questions so, by obsessing over them rather than simply taking a philosophical position regarding them----then 'leaving the debating chamber' and taking action.

Manuel,

No; I am not suggesting at such thing; indeed, my post above---which you recommended---counsels exactly, and in agreement with you, that researchers must indeed take into account how they are able to conduct research in the first place.

However, as my last post indicates, at some point---preferably sooner than later----researchers must 'leave the debating chamber' with a practical philosophical position regarding research---and conduct research; not doing so is for 'eternal philosophers', not researchers.

Exactly.

In my opinion the integrity of the research is unrelated to epistemology. Instead our actions should be driven by our goals — which includes ethical goals

And so whilst we agree that there is treatment of philosophical positions and David argues that integrity and ethical considerations are key how do we then apply the standards? Aren't we back to the same argument of the original question?

Does epistemology offer specific methodological rules?

No doubt, a researcher must follow her ethical goals but what about her design goals?

Actually, I agree for the most part, with David's last post, above; not all-the-way, but mostly.

All I have tried to put across with previous posts on this question is that I think that research underpinned with a 'traditional' philosophical position (ontology>epistemology>methodology) offers the most transparent audit-trail for peer-critique of the work-----they can see 'where it's coming from' entirely.

Thus, as I answered in a previous post, epistemology does not offer specific methodological rules, but rather, the flexibility of options.

I also do not think that epistemology offers methodological rules. Instead, methods have to be matched to goals, which includes ethical goals

Dora

No doubt research follows certain conventions but it need not not follow standards as it is about delving into unknown. There should be right research questions; that is the pivot around which everything else like design, methodology, methods, tools and techniques revolve.

I agree with David that methods have to be matched to goals.

However, contra Philip, both the accepted goals and the methodological choices are most often limited by accepted practice within a discipline. Methods generally flow more from established practice within a discipline, or sub-discipline. Established practice is usually associated with a specific epistemological-ontological position; but the history of research (see Kuhn) suggests that ontology and epistemology are largely secondary and post hoc rationalisations for the preferences researchers have about what kind of research to do and what methods to choose. Epistemology and ontology form part of the paradigm but are not the principal drivers of methodological choice.

This does not mean people cannot come up with 'new' methods (for example 'grounded theory') but rather that such methods are usually consistent with the establish paradigm of practice. Nor does it mean that philosophical positions cannot influence methods (e.g. Skinner), of course they do. However, it takes a radical upheaval, or paradigm shift, within a discipline for methods within that discipline to change significantly; and such changes lead to subsequent rationalisations about philosophical positions.

Further, there are too many instances of 'peers' dismissing research because it does not fall within one of the research paradigms (or milieus); or because it adheres to another paradigm. Trying to categorise research methods into one or other of the accepted epistemological-ontological positions takes the emphasis away from the research, the chosen methods and what the research is trying to achieve. It might make peer-review easier but it is no substitute for focusing on the quality of the research, its goals and the fit between goals and methods.

By default, when we choose a method of research, this choice is underpinned by ontological/epistemological position-----it can't happen any other way, whether this is overtly recognised or not.

If human cognition followed Descartes principles, I would agree with Philip.