The role of internet communities in ex-christian deconversion
What do you think are the most common reasons/factors explaining conversion from atheism to Christianity? Have you known many who converted after identifying as atheist?
Mathematic and Genomic classifications, yes. Before the use of such quantitative techniques, it was done through observation alone which is useful and is part of the scientific method, but not sufficient to qualify early taxonomy as a science.
Have you known many who converted after identifying as atheist?
Why do you qualify it with "after identifying as atheist" does that mean you only consider vocal atheists as a viable member of the set of conversions?
For example: consider a hypothesis in this form: "members of animal population X are able to inter-breed with members of animal population Y, therefore they belong the same taxonomic class."
It seems clear to me that this hypothesis is falsifiable, and therefore it satisfies the criteria you've laid out. The analysis is also clearly based on a qualitative relation posited to exist between entities, and not a strictly quantitative one. This contradicts your assertion that only quantitative analysis can be falsified.
Note that I'm not saying that the above method of classification is the best possible method, only that it satisfies your criteria for science.
As a matter of fact though, genomic classification isn't really that different, conceptually. It just uses a richer source of data to identify comparative differences for classification. For example, genomic classification as I understand it generally requires that members of the class share a gene pool, which obviously involves the fact that they can inter-breed (leaving aside asexual reproduction for the sake of simplicity).
I also suspect that if we probed further into your understanding of the differences between qualitative and quantitative analysis that we'd find some conceptual problems, particularly in cases where underlying qualitative data is aggregated and subjected to statistical analysis, but the above seems simpler.
If you're interested, you should start a thread and present some of the evidence that you think ought to be persuasive.
No, sorry I should have been clearer. I was just trying to isolate the idea of atheists converting to Christianity, rather than conversion in general, since you specified atheists. By "identified as atheist" I don't mean necessarily in any particularly strong or public sense.
No, sorry I should have been clearer. I was just trying to isolate the idea of atheists converting to Christianity, rather than conversion in general, since you specified atheists. By "identified as atheist" I don't mean necessarily in any particularly strong or public sense.
OK. But I think this is at odds with your statement that the relevant criteria is falsifiability, and that qualitative analysis does not allow for falsifiability.
For example: consider a hypothesis in this form: "members of animal population X are able to inter-breed with members of animal population Y, therefore they belong the same taxonomic class."
It seems clear to me that this hypothesis is falsifiable, and therefore it satisfies the criteria you've laid out. The analysis is also clearly based on a qualitative relation posited to exist between entities, and not a strictly quantitative one. This contradicts your assertion that only quantitative analysis can be falsified.
Note that I'm not saying that the above method of classification is the best possible method, only that it satisfies your criteria for science.
As a matter of fact though, genomic classification isn't really that different, conceptually. It just uses a richer source of data to identify comparative differences for classification. For example, genomic classification as I understand it generally requires that members of the class share a gene pool, which obviously involves the fact that they can inter-breed (leaving aside asexual reproduction for the sake of simplicity).
For example: consider a hypothesis in this form: "members of animal population X are able to inter-breed with members of animal population Y, therefore they belong the same taxonomic class."
It seems clear to me that this hypothesis is falsifiable, and therefore it satisfies the criteria you've laid out. The analysis is also clearly based on a qualitative relation posited to exist between entities, and not a strictly quantitative one. This contradicts your assertion that only quantitative analysis can be falsified.
Note that I'm not saying that the above method of classification is the best possible method, only that it satisfies your criteria for science.
As a matter of fact though, genomic classification isn't really that different, conceptually. It just uses a richer source of data to identify comparative differences for classification. For example, genomic classification as I understand it generally requires that members of the class share a gene pool, which obviously involves the fact that they can inter-breed (leaving aside asexual reproduction for the sake of simplicity).
Using genomes to classify is basically gene counting. It's very numerical and hence quantitative.
I also suspect that if we probed further into your understanding of the differences between qualitative and quantitative analysis that we'd find some conceptual problems, particularly in cases where underlying qualitative data is aggregated and subjected to statistical analysis, but the above seems simpler.
Would I be wrong in assuming this forum has a large number of such threads already? I would not be wrong in saying there are thousands of such threads all over the Internet. What do you think starting a new one would do those have not?
I don't think it's very useful to distinguish quantitative from qualitative based on the ability to code cases in the way you've described, because literally every observation is quantifiable in this way. For example, I can read a deconversion narrative to see whether it mentions an intellectual concern with the veracity of the Bible. I code that as either 1 or 0. We did this sort of thing with our data, and as part of our coding of narratives. More simply, I can literally just count how often various terms and classes of terms appear in the texts. I also have that data.
Previously I mentioned that quantifying measures in order to do statistical analysis across large samples is the best way to assess the representativeness of some phenomena in the social sciences, something that purely qualitative analysis struggles with for multiple reasons. However, external validity in that sense isn't the only kind of validity that matters. Assessments of interval validity involve arguments about the logical/qualitative relationship between different entities or processes. In a sense, this is the other side of the coin, e.g. that correlation is not causation. Without a correlation some causal inference will not be generalizable, and correlation requires statistics. But without an analysis of the actual processes which create the relationship in question you can never say much about what the correlation means.
Good qualitative research in the social sciences is about trying to understand what those processes are really like, and about trying to construct richer conceptualizations of them. Good quantitative research in the social sciences is about showing that some conceptual theory applies in some general way across a wide variety of cases. Another analogy might be to medicine. Statistical data can tell you that behavior X is associated with higher rates of cancer, but you need other data and methods to understand the mechanism by which that relationship occurs.
It would enable us to talk about it with you, specifically. It might be enjoyable. But like I said, only if you're interested.
I don't think it's very useful to distinguish quantitative from qualitative based on the ability to code cases in the way you've described, because literally every observation is quantifiable in this way. For example, I can read a deconversion narrative to see whether it mentions an intellectual concern with the veracity of the Bible. I code that as either 1 or 0. We did this sort of thing with our data, and as part of our coding of narratives. More simply, I can literally just count how often various terms and classes of terms appear in the texts. I also have that data.
Previously I mentioned that quantifying measures in order to do statistical analysis across large samples is the best way to assess the representativeness of some phenomena in the social sciences, something that purely qualitative analysis struggles with for multiple reasons. However, external validity in that sense isn't the only kind of validity that matters. Assessments of interval validity involve arguments about the logical/qualitative relationship between different entities or processes. In a sense, this is the other side of the coin, e.g. that correlation is not causation. Without a correlation some causal inference will not be generalizable, and correlation requires statistics. But without an analysis of the actual processes which create the relationship in question you can never say much about what the correlation means.
Good qualitative research in the social sciences is about trying to understand what those processes are really like, and about trying to construct richer conceptualizations of them. Good quantitative research in the social sciences is about showing that some conceptual theory applies in some general way across a wide variety of cases. Another analogy might be to medicine. Statistical data can tell you that behavior X is associated with higher rates of cancer, but you need other data and methods to understand the mechanism by which that relationship occurs.
Good qualitative research in the social sciences is about trying to understand what those processes are really like, and about trying to construct richer conceptualizations of them. Good quantitative research in the social sciences is about showing that some conceptual theory applies in some general way across a wide variety of cases. Another analogy might be to medicine. Statistical data can tell you that behavior X is associated with higher rates of cancer, but you need other data and methods to understand the mechanism by which that relationship occurs.
Huh? Okay, that was weird, your wanderings are a bit hard to follow, but ...
1) I never said observation was not valuable. In fact, I said explicitly that observation and data gathering was valuable.
2) Nothing you said refutes the fact that falsifiability is a necessary part of the scientific method, and that qualitative analysis does not allow falsifiability.
Okay, that said, maybe you are trying to point out the problems with Statistical Inference? Yes, i agree Statistical Inference has weaknesses when applied to the Social Sciences, but hey, I never said it didn't, but that's all the Social Sciences have to make them respectable. I guess it sucks to be a Sociologist.
1) I never said observation was not valuable. In fact, I said explicitly that observation and data gathering was valuable.
2) Nothing you said refutes the fact that falsifiability is a necessary part of the scientific method, and that qualitative analysis does not allow falsifiability.
Okay, that said, maybe you are trying to point out the problems with Statistical Inference? Yes, i agree Statistical Inference has weaknesses when applied to the Social Sciences, but hey, I never said it didn't, but that's all the Social Sciences have to make them respectable. I guess it sucks to be a Sociologist.
Your claim was that quantitative analysis was the only thing that could make social science a science.
So according to you, even using using the most highly regarded of all scientific method, the empirical experiment, would not be scientific. There are numerous other scientific approaches that would not qualify per your arbitrary demands, but this one is such a striking example that I think it speaks for itself.
Now, I don't think you meant to say this. You merely don't know or understand the subject matter.
As for the given example that you had problems understanding: It illustrates that using analysis to falsifying hypotheses with a statistical model is not the same as falsifying that the model is an accurate representation of the world.
It's interesting that you on one hand argues that such models is the only thing that can make something scientific, while on the other hand flippantly discount use of such models and question the integrity of entire fields. There is a palpable irony to your chain of argument. "We need A to make something scientific. Oh btw, A isn't very scientific".
A debate on science is easier to grasp once you understand that the tools of science do not on their own make something scientific, in the same sense that waving a hammer around doesn't necessarily build a wall.
Fwiw; there are professional researchers who do not grasp that distinction and for whom the appearance of science has taken front-seat over actually doing science. So you are hardly alone.
It's interesting that you on one hand argues that such models is the only thing that can make something scientific, while on the other hand flippantly discount use of such models and question the integrity of entire fields. There is a palpable irony to your chain of argument. "We need A to make something scientific. Oh btw, A isn't very scientific".
A debate on science is easier to grasp once you understand that the tools of science do not on their own make something scientific, in the same sense that waving a hammer around doesn't necessarily build a wall.
Fwiw; there are professional researchers who do not grasp that distinction and for whom the appearance of science has taken front-seat over actually doing science. So you are hardly alone.
I'm sure you would find the conversation easier to follow if you stop moving your goalposts.
Your claim was that quantitative analysis was the only thing that could make social science a science.
So according to you, even using using the most highly regarded of all scientific method, the empirical experiment, would not be scientific. There are numerous other scientific approaches that would not qualify per your arbitrary demands, but this one is such a striking example that I think it speaks for itself.
Now, I don't think you meant to say this. You merely don't know or understand the subject matter.
Your claim was that quantitative analysis was the only thing that could make social science a science.
So according to you, even using using the most highly regarded of all scientific method, the empirical experiment, would not be scientific. There are numerous other scientific approaches that would not qualify per your arbitrary demands, but this one is such a striking example that I think it speaks for itself.
Now, I don't think you meant to say this. You merely don't know or understand the subject matter.
Originally Posted by pulvis
Falsifiable conclusions bring it within the realm of science. Statistical analysis exposes falsifiability. Quantitative methods allow statistical analysis to be brought to bear.
Originally Posted by tame_deuces
It doesn't really matter, because the problem is still that the model is s**t and the hypothesis ridiculous. And numbers aren't what you need to dispel that, but observation, inquiry and actually bothering to explore the issue.
OK. But I think this is at odds with your statement that the relevant criteria is falsifiability, and that qualitative analysis does not allow for falsifiability.
For example: consider a hypothesis in this form: "members of animal population X are able to inter-breed with members of animal population Y, therefore they belong the same taxonomic class."
It seems clear to me that this hypothesis is falsifiable, and therefore it satisfies the criteria you've laid out. The analysis is also clearly based on a qualitative relation posited to exist between entities, and not a strictly quantitative one. This contradicts your assertion that only quantitative analysis can be falsified.
Note that I'm not saying that the above method of classification is the best possible method, only that it satisfies your criteria for science.
As a matter of fact though, genomic classification isn't really that different, conceptually. It just uses a richer source of data to identify comparative differences for classification. For example, genomic classification as I understand it generally requires that members of the class share a gene pool, which obviously involves the fact that they can inter-breed (leaving aside asexual reproduction for the sake of simplicity).
I also suspect that if we probed further into your understanding of the differences between qualitative and quantitative analysis that we'd find some conceptual problems, particularly in cases where underlying qualitative data is aggregated and subjected to statistical analysis, but the above seems simpler.
For example: consider a hypothesis in this form: "members of animal population X are able to inter-breed with members of animal population Y, therefore they belong the same taxonomic class."
It seems clear to me that this hypothesis is falsifiable, and therefore it satisfies the criteria you've laid out. The analysis is also clearly based on a qualitative relation posited to exist between entities, and not a strictly quantitative one. This contradicts your assertion that only quantitative analysis can be falsified.
Note that I'm not saying that the above method of classification is the best possible method, only that it satisfies your criteria for science.
As a matter of fact though, genomic classification isn't really that different, conceptually. It just uses a richer source of data to identify comparative differences for classification. For example, genomic classification as I understand it generally requires that members of the class share a gene pool, which obviously involves the fact that they can inter-breed (leaving aside asexual reproduction for the sake of simplicity).
I also suspect that if we probed further into your understanding of the differences between qualitative and quantitative analysis that we'd find some conceptual problems, particularly in cases where underlying qualitative data is aggregated and subjected to statistical analysis, but the above seems simpler.
Things that can't be scientific, per Pulvis' criteria:
1. Experiments
2. Simulations
3. Observation
All of which are employed in social science, and none of which require quantitative analysis (aka "the only way to make the social sciences scientific").
And if we assume the criteria for what is scientific also applies outside social science, someone needs to call up the STEM departments and tell them they're out of business. Well, or get some surveys sent out and do proper science.
I'm not going to go point by point because it's repetitious. It's not worth my time repeating myself for someone who has not bothered to take the time to understand the things that he's talking about.
1)
I'll also add:
https://en.oxforddictionaries.com/definition/etymology
New words are constantly being introduced into our language, and there is not a single "official" English. Such a claim would go far beyond that which any linguist would claim. There was once a time when "email" was a technical term that only a few people knew, and many of them understood it first as "electronic mail" which was eventually shortened into a new word. If you google "etymology email" you will get that the etyomology of "email" is precisely that. It started as electronic mail, then was shortened to email. That's how language evolves over time.
2)
----
Please notice how I was able to quote myself in both cases. You can click the links to see where they came from.
You can try to claim all you want about science and etymologies and even the "status" of the word exitimony. But at this point, it's clear that you are not interested in truth. You have not taken any time to really try to test or verify your level of knowledge. You are simply asserting out of ignorance and a desire to try to express emotional disdain in the form of an intellectual argument. That's not how any of this stuff works.
https://en.oxforddictionaries.com/definition/etymology
1) The study of the origin of words and the way in which their meanings have changed throughout history.
1.1) The origin of a word and the historical development of its meaning.’
1.1) The origin of a word and the historical development of its meaning.’
2)
The paper is a "data gathering" effort followed with analysis. This is what people are actually saying about themselves and their experiences. It fits into a larger structure of what we know about human society, and the things that have an impact on human behavior. The scientific method does not boil down to simply running experiments in controlled settings all the time. There are times when you simply make observations and try to fit those observations into existing intellectual frameworks to try to understand them.
You have never countered any argument, exactly in which post did you counter either of the two positions I listed just above?
You can try to claim all you want about science and etymologies and even the "status" of the word exitimony. But at this point, it's clear that you are not interested in truth. You have not taken any time to really try to test or verify your level of knowledge. You are simply asserting out of ignorance and a desire to try to express emotional disdain in the form of an intellectual argument. That's not how any of this stuff works.
You're far too polite.
Things that can't be scientific, per Pulvis' criteria:
1. Experiments
2. Simulations
3. Observation
All of which are employed in social science, and none of which require quantitative analysis (aka "the only way to make the social sciences scientific").
And if we assume the criteria for what is scientific also applies outside social science, someone needs to call up the STEM departments and tell them they're out of business. Well, or get some surveys sent out and do proper science.
Things that can't be scientific, per Pulvis' criteria:
1. Experiments
2. Simulations
3. Observation
All of which are employed in social science, and none of which require quantitative analysis (aka "the only way to make the social sciences scientific").
And if we assume the criteria for what is scientific also applies outside social science, someone needs to call up the STEM departments and tell them they're out of business. Well, or get some surveys sent out and do proper science.
Data gathering may be a precursor to making a falsifiable claim, but it is not sufficient in and of itself.
I'm not going to go point by point because it's repetitious. It's not worth my time repeating myself for someone who has not bothered to take the time to understand the things that he's talking about.
1)
I'll also add:
https://en.oxforddictionaries.com/definition/etymology
New words are constantly being introduced into our language, and there is not a single "official" English. Such a claim would go far beyond that which any linguist would claim. There was once a time when "email" was a technical term that only a few people knew, and many of them understood it first as "electronic mail" which was eventually shortened into a new word. If you google "etymology email" you will get that the etyomology of "email" is precisely that. It started as electronic mail, then was shortened to email. That's how language evolves over time.
1)
I'll also add:
https://en.oxforddictionaries.com/definition/etymology
New words are constantly being introduced into our language, and there is not a single "official" English. Such a claim would go far beyond that which any linguist would claim. There was once a time when "email" was a technical term that only a few people knew, and many of them understood it first as "electronic mail" which was eventually shortened into a new word. If you google "etymology email" you will get that the etyomology of "email" is precisely that. It started as electronic mail, then was shortened to email. That's how language evolves over time.
2)
Originally Posted by Aaron W
The paper is a "data gathering" effort followed with analysis. This is what people are actually saying about themselves and their experiences. It fits into a larger structure of what we know about human society, and the things that have an impact on human behavior. The scientific method does not boil down to simply running experiments in controlled settings all the time. There are times when you simply make observations and try to fit those observations into existing intellectual frameworks to try to understand them.
1) The paper itself claims it is a qualitative analysis. This in no way allows for a falsifiable claim to be made.
2) It does gather some data, but makes no scientific claims based no that data. Data gathering alone is not sufficient to rise to the level of satisfying the scientific method. For example, simply doing word counts does not make the act of counting an employment of the scientific method.
3) I have said multiple times that qualitative methods can be useful and interesting as a precursor to doing the actual science.
You can try to claim all you want about science and etymologies and even the "status" of the word exitimony. But at this point, it's clear that you are not interested in truth. You have not taken any time to really try to test or verify your level of knowledge. You are simply asserting out of ignorance and a desire to try to express emotional disdain in the form of an intellectual argument. That's not how any of this stuff works.
You overestimate your answers. They fall short as I have shown.
I'm not really sure what the issue is here. Regardless of whether or not qualitative sociology is *Science*, it is obviously still potentially useful and can add to the sum total of human knowledge. For instance, most history books would not qualify as science under pulvis's criteria, but they still seem both worthwhile and useful for learning about the world. Qualititative methods in the social sciences seem to me a clear improvement over just using anecdotes (they are usually paired with quantitative elements and address statistical concerns with representativeness and various biases). My view is that most sociology isn't really a science in the same way that physics or chemistry is because there is no consensus on a paradigm, but so what? Are we supposed to just not use rational and empirical methods to study societies? What are the better alternatives?
Notice with email there is meaning and derivation to all it's parts. The definition requires the collection of syllables to be a word. The word "extimony" is the prefix ex, which has meaning, and then a meaningless suffix "timony". I could just as easily say "exdorami" meant something just because it has an ex prefix. We could arbitrarily force it into the language over time and pretend it had an etymology, but that does not mean the initial contrivance had an etymology.
Secondly, the idea that the existence of "ex" as an affix means that everything that starts with ex must therefore be interpreted as an affix is just dumb. That shows no understanding of the English language.
As to the bolded, it's literally just an accounting of the word. Some words come from places, while others are just spontaneously created. "Google" is a word that has no etymology. Same with "Kleenex." They are mere utterances that people used to name companies or products, and then that became part of the lexicon.
Just for fun, I looked around the internet for a bit:
https://www.fanfiction.net/s/13141872/1/One-Life
The ex[-]Dorami member was telling all the girls to go home.
If you want to deny that "extimony" is derived from "testimony" you can continue to do so and be considered ignorant for it.
The scientific method absolutely requires falsifiability.
1) The paper itself claims it is a qualitative analysis. This in no way allows for a falsifiable claim to be made.
2) It does gather some data, but makes no scientific claims based no that data. Data gathering alone is not sufficient to rise to the level of satisfying the scientific method. For example, simply doing word counts does not make the act of counting an employment of the scientific method.
3) I have said multiple times that qualitative methods can be useful and interesting as a precursor to doing the actual science.
1) The paper itself claims it is a qualitative analysis. This in no way allows for a falsifiable claim to be made.
2) It does gather some data, but makes no scientific claims based no that data. Data gathering alone is not sufficient to rise to the level of satisfying the scientific method. For example, simply doing word counts does not make the act of counting an employment of the scientific method.
3) I have said multiple times that qualitative methods can be useful and interesting as a precursor to doing the actual science.
Formally, you are correct that the concept of falsifiability is a requirement of a scientific hypothesis in the scientific method. But the scientific method does not define the entirety of science.
Newton's Third Law of Motion is essentially non-falsifiable. It's a model. If we were to ever to find something that does not appear to conform to Newton's Third Law, we would assert that there's some other yet-to-be-discovered force that's not being accounted for. That's because Newton's laws function axiomatically as a tool for analyzing situations, which is what models do. (What exactly is a "force"? It's the thing we think we're measuring even when we have no idea what it is.) And we teach Newton's Laws of Motion as part of the scientific curriculum, and it's most certainly considered science.
The lines you're trying to draw here are pretty arbitrary, and if you understood more about what you were talking about, this would be more obvious. But again, you can continue to yell at the clouds in utter ignorance if you choose. It's your life. Spend it how you want.
-I don't find it surprising, but it is depressing how fringe ideas (like mythicism) get taken up so readily by online communities with a common identity.
-My pet theory is that Christians that post a lot in religion forums like ours are less likely to be significantly involved in a physical church community. Did you find any indication of this in your research?
-There's a selection bias in focusing on a forum titled "Ex-Christian." In focusing on people who either use this as an identifier or hang out with other people who do, you are selecting for people for whom their specific history as a former Christian continues to be relevant. I know many people who don't identify themselves as "atheist" to avoid this implication - they acknowledge that they are an atheist if pressed on it, but view that as a mostly negative description, a way of saying I'm not religious, rather than as a way of identifying what they are: I am a humanist, or socialist, existentialist, effective altruist, feminist etc. How broadly do you think the description of the ex-Christians in your study apply to other former Christians who hang out on the internet, but don't focus on their past religion so consciously?
It's a discreet value that can be either 0 or 1 and is based on the objective observation that it resulted from breeding. It does not have to be a real number or allow for values in a greater range.
Your "coding" is subjective because it's based on the subjective use and understanding of language or even whether the sampled item is telling the truth ( This is why double blind trials are done). Whether a baby is born is not subjective.
Your "coding" is subjective because it's based on the subjective use and understanding of language or even whether the sampled item is telling the truth ( This is why double blind trials are done). Whether a baby is born is not subjective.
So it's clear from the fact that you accept the one and reject the other that it's not quantifiability which is important to you; rather it's your assessment of the objectivity of the data/methods. Here's another example: I make a prediction that it will rain tomorrow <at a given set of coordinates>. That prediction is again entirely falsifiable, objective, and qualitative. It requires no quantitative analysis whatsoever. Quantifiability is not the sine qua non of objectivity or science.
If you make a claim based on a qualitative analysis and I disagree all you can resort to is an argument from authority (you cite some claimed expert) or an argument from numbers (you cite a bunch of people who agree with you). Both are logical fallacies. That is not science.
And while there may be cases where there is ambiguity in a given text, this is no less so for many other types of data you consider to be objective. For example, in the case of interbreeding, if you're going to use quantitative methods than presumably you'll measure how often individuals of each species successfully mate. But how will you determine the threshold of success needed to say that the two species are the same? What will you do if they succeed ~50% of the time? That is no less difficult a problem than the problem of an ambiguous text.
If I was going to sum up my disagreement with the way you are using the terms "objective" and "subjective", I would say that the problem is you aren't paying sufficient attention to inter-subjectivity in relation to actual scientific methods as implemented by real human beings, particularly when the domain under investigation is how people think, feel, act, and interpret their own experiences. This would become clearer probably if we got deeper into explanations about why "objectivity" and "falsifiability" are important in science. That is, they are important epistemologically, in relation to the reliability and validity of knowledge and the replicability of the methods used to obtain knowledge. It's not possible to deal with the epistemology in detail without acknowledging that "objectivity" is not as simple as you would like it to be, even in natural sciences, because it is always humans assessing objectivity. See for example Kuhn's point about the theory-ladenness of observation.
Intersubjectivity is also relevant, in a different way, in the social sciences simply because so much of what we want to understand involves the meaning people attach to ideas, events, values, and so on. Meanings are inescapably intersubjective as well.
Beyond those concerns though, I think you are just artificially limiting "science" to those conclusions which I would suggest are the intended result of a larger scientific process. What I mean is this: the paper linked in this thread is definitely not trying to assert a concrete theory or hypothesis related to deconversion. It's not making a falsifiable claim analogous to "it will rain tomorrow" or "all Christian deconversion involves such and such a theoretical process" We don't have enough data to do so confidently. The purpose, as I said before, is mainly descriptive. It's essentially a case study. But collecting and describing data is also a part of the scientific process. Before you can test hypothesis you have to form hypotheses, and that requires a lower level of theoretical analysis of observations. The purpose of the data-gathering and analytic methods used in the study are to try to ensure the internal validity of the analysis presented, and those methods are decades old (cf. grounded theory) and widely accepted within the discipline.
It's still way too much work to try to attempt to convince you of the validity of those methods, but like I said you can always buy some books on it if you want. I'm not actually that concerned with whether or not you think this particular paper is "scientific" or not. I do think the more modest task of convincing you that your qualitative/quantitative distinction is insufficient, or that qualitative methods/data can be objective, should be feasible. Perhaps the above helps.
One thing I find interesting (just my impression/opinion) is how leaving their religion doesn't necessarily change everything about how people think about the world, or even their own prior religion. For example in a lot of my own reading on that site I think people who used to hold to fundamentalist readings of the Bible tend to still hold to those same readings as ex-Christians, in a sense. Their understanding of hermeneutics hasn't really changed. What changed is that they used to think the Bible was true and now think it's false. But they are often skeptical that other readings or approaches to the text could even be legitimate. Many who deconverted from more conservative sects continue to have very negative opinions of more liberal sects; they think they are disingenuous in their approach to the Bible.
-There's a selection bias in focusing on a forum titled "Ex-Christian." In focusing on people who either use this as an identifier or hang out with other people who do, you are selecting for people for whom their specific history as a former Christian continues to be relevant. I know many people who don't identify themselves as "atheist" to avoid this implication - they acknowledge that they are an atheist if pressed on it, but view that as a mostly negative description, a way of saying I'm not religious, rather than as a way of identifying what they are: I am a humanist, or socialist, existentialist, effective altruist, feminist etc. How broadly do you think the description of the ex-Christians in your study apply to other former Christians who hang out on the internet, but don't focus on their past religion so consciously?
One potential point of comparison might be the site agnostic.com (or humanist.com. It's the same site). It's a bit of a weird forum, which also acts like a dating site for people who identify as agnostic/humanist. I haven't made any rigorous effort to see what's going on there but it does seem like people use that site less for support and more just for general social interaction, like the way they might attend a church for social interaction.
Selection bias is completely unproblematic in a qualitative study though. That's one of the thing that makes them brilliant.
The quantitative fixation with conclusions generalized to bigger populations is often not that interesting. It can be, but localized phenomena can be just intriguing and often far more so.
Let's say I want to study how culture affects running. Now, a sample that should allow me to generalize to the population of Europe isn't useless, but if I found a very small village that had an abnormally high amount of fast runners... that would be far, far more interesting.
Now obviously we can use quantitative measures on small populations without generalizing to a large regional population, but they do grow far less useful as the populations grow under 500... and honestly they are largely useless if you are below or hovering around a 100.
The quantitative fixation with conclusions generalized to bigger populations is often not that interesting. It can be, but localized phenomena can be just intriguing and often far more so.
Let's say I want to study how culture affects running. Now, a sample that should allow me to generalize to the population of Europe isn't useless, but if I found a very small village that had an abnormally high amount of fast runners... that would be far, far more interesting.
Now obviously we can use quantitative measures on small populations without generalizing to a large regional population, but they do grow far less useful as the populations grow under 500... and honestly they are largely useless if you are below or hovering around a 100.
Selection bias is completely unproblematic in a qualitative study though. That's one of the thing that makes them brilliant.
The quantitative fixation with conclusions generalized to bigger populations is often not that interesting. It can be, but localized phenomena can be just intriguing and often far more so.
Let's say I want to study how culture affects running. Now, a sample that should allow me to generalize to the population of Europe isn't useless, but if I found a very small village that had an abnormally high amount of fast runners... that would be far, far more interesting.
Now obviously we can use quantitative measures on small populations without generalizing to a large regional population, but they do grow far less useful as the populations grow under 500... and honestly they are largely useless if you are below or hovering around a 100.
The quantitative fixation with conclusions generalized to bigger populations is often not that interesting. It can be, but localized phenomena can be just intriguing and often far more so.
Let's say I want to study how culture affects running. Now, a sample that should allow me to generalize to the population of Europe isn't useless, but if I found a very small village that had an abnormally high amount of fast runners... that would be far, far more interesting.
Now obviously we can use quantitative measures on small populations without generalizing to a large regional population, but they do grow far less useful as the populations grow under 500... and honestly they are largely useless if you are below or hovering around a 100.
Feedback is used for internal purposes. LEARN MORE