Excellent, we have requests for an educational dissection of the content at hand. This will provide an opportunity to improve the knowledge and critical thinking of all readers, including myself, as very few people inspect their thought processes at this level of detail. I welcome this exercise.
Let's begin with my initial assertion.
Conclusions reached by data inferred from anecdote are fallacious, regardless of the conclusion drawn.
This statement refers to the logical error of using (typically a small number of) self-reported stories as data, from which conclusions are made.
In retrospect, I would rephrase this slightly:
Conclusions reached by data inferred from anecdotal evidence are based on fallacious argumentation.
There may have been some confusion about "anecdote" versus "anecdotal evidence" and "fallacious conclusion" versus "fallacious argumentation." I want to be clear about my language, and if my initial statement was unclear, that is my mistake.
Anecdotal evidence is an informal report of personal experience. These reports are typically (but not necessarily) much less rigorous in detail and structure than scientific data gathering, as they are typically not designed in advance. They can also be subject to the bias of the reporter, and are limited to the perspective of the reporter. The value of anecdotal evidence is in using it to develop questions to more rigorously answer, but it is only that...a starting point, not a finish line.
Fallacious argumentation is the process of putting together a set of ideas to reach a conclusion, but the process is invalid due to one or more failures in reasoning. A failure in the reasoning is a
fallacy, and there are many kinds of fallacies (which are
very common in almost every discussion/debate). We will examine two types of fallacies.
In my original post, another forum user had commented about having no vehicle problems. I contrasted this with my abnormally high number and magnitude of problems. Both of those reports are anecdotal evidence. Each of us is just one owner with one vehicle. To make a conclusion based on either of our reports is called a
Hasty Generalization fallacy, which is common when people start from anecdotal evidence. Neither of our reports is sufficient to conclude quality of Tesla vehicles in general, as Tesla has sold hundreds of thousands of vehicles. 2 vehicles out of 200,000 is less than a percent of a percent, which isn't a meaningful statistic.
Also note that we are talking about evidence in a statistical/scientific context and putting together reasoned arguments (a philosophical context). Understanding the context of the terms and concepts used is important, and we'll get to why this is important later in this exposition.
Let's evaluate some of the content that attempts to refute the position that anecdotal evidence is invalid for drawing conclusions.
I don't agree that a person's first hand experience is, in this case, an "anecdote" in the first place. However, assuming it is, if I conclude, based on no problems with my vehicle over a number of months and miles (hence it is no longer anecdotal evidence)
We've established what anecdotal evidence means, which is necessarily first-hand experience. There may be some confusion in separating the idea of an "anecdote" (a short story) from "anecdotal evidence," but these sentences seem to equate them, given the use of both terms. Let's focus on the use of the term "anecdotal evidence" at the end here, since that is the common term.
The assertion is that some period of time (in months) and distance traveled creates a situation of extended usage that no longer makes the informal report "anecdotal." This might seem reasonable, but it is incorrect. The state of evidence being anecdotal is not limited by time. Anecdotal evidence is characterized by the way the information is gathered and reported. A single person reporting experiences over a week, month, or year is still a single person, with potential bias, reporting experiences about a single vehicle with whatever level of detail was collected at the time and retained to report. It is important to note that the focus of the anecdotal evidence here is
the vehicle, not the person owning/using it nor the time frame in which they owned it. Ten years of reporting experiences about one vehicle is still one vehicle.
if I conclude, based on no problems with my vehicle over a number of months and miles (hence it is no longer anecdotal evidence), that I want to buy another, that does not make my conclusion fallacious at all. In fact, it's quite a prudent conclusion. However, if I try to say that will be your experience as well, then yes, it is fallacious. In order for me to say that I need data that only comes from polling a sufficient amount of owners to have a low margin of error, or directly by way of service records.
This is interesting. The assertion here is that there are two different arguments and conclusions: one first-person ("my" next Tesla) and one second-person ("your" next Tesla).
Let's break down the first-person argument, which is supposedly valid:
1. I (singular Tesla owner) have experienced no (or minor) issues with my Tesla vehicle over the span of XX months and/or XXXXX miles (where these are non-trivial quantities, choose whatever values make this carry reasonable weight ).
2. Therefore, another Tesla vehicle I purchase will likely have a similar pattern of reliability.
3. I want to buy a vehicle with a similar pattern of reliability to my current vehicle.
4. Therefore, I want to buy another Tesla.
And the second-person argument, which is supposedly invalid (fallacious):
1. I (singular Tesla owner) have experienced no (or minor) issues with my Tesla vehicle over the span of XX months and/or XXXXX miles (where these are non-trivial quantities, choose whatever values make this carry reasonable weight ).
2. Therefore, another Tesla vehicle you purchase will likely have a similar pattern of reliability.
Carefully note the conclusion in #2 in both arguments:
the owner of the "another" vehicle is irrelevant. This is crucial to understanding the fallacy, as it does not matter if it is first-person ("My vehicle is fine, so the next one I buy will be fine") or second-person ("My vehicle is fine, so the next one you buy will be fine."). The conclusion is based on the anecdotal evidence regarding
a single vehicle, which is then necessarily applied to
all future vehicles. "My" next Tesla vehicle and "your" next Tesla vehicle are both in the undifferentiated pool of all future Tesla vehicles. If it is valid to conclude that "my" vehicle out of all those will likely be "good," then it must be likely that all of them are "good," which would mean "your" vehicle is equally likely to be "good." However, #2 is a hasty generalization in both arguments, so both of them are fallacious.
Now, let's engage the concept of equivocation.
Equivocation is a change in context that changes the meaning of words/terms as part of constructing an argument, which is a fallacy.
The assertion here is that anecdotal evidence is valid/acceptable as part of argumentation, based on the above reputable source.
Note the first line of this text. The equivocation is expressly stated. Scientific evidence and legal evidence
do not have the same definition. Evidence has a different meaning when used
in a legal context, so constructing an argument based the meaning of anecdotal evidence
in a legal context is a fallacy of equivocation. We are in a
statistical/scientific context, so we must consistently use the definition of evidence for this context.
When further considering how the context matters and changes the meanings of terms, consider what the concept of "proof" means in a scientific context, in a philosophical context, and in a legal context.
Aside from the equivocation, there's something else interesting in this source material. Let's assume the definitions of "evidence" and "anecdotal evidence" are the same across contexts. The last sentence states that anecdotal evidence
may have value
when applied to other forms of evidence. My initial statement was in regards to the fallacy when anecdotal evidence is the
only source of data. Note the similarities in those concepts.
The final topic I will address is one of the implications regarding a conclusion being the result of a fallacious argument.
It is important to understand that a conclusion can be
true and still be
fallacious (in that it resulted from a fallacious argument). Fallacious is not the same as false when constructing arguments.
For example:
1. My neighbor's dog is a dalmatian.
2. Therefore, all dogs are dalmatians.
3. Therefore, my dog is a dalmatian.
#2 is a hasty generalization, since the instance of a single dog does not define the entirety of dogs, so this
argument is
fallacious. However, #3 could be true. I could own a dalmatian, in which case #1 and #3 are true, but #2 is a fallacy. Evaluating arguments determines if the thought process is
valid (without fallacy), but does not confirm/refute the
truth of the statements and/or conclusions.
Alright, I think that's enough content regarding argumentation and fallacies. I've evaluated all of the relevant content, so there's nothing else for me to add. I appreciate anyone who took the time to read this post and, hopefully, learned something in the process. If anyone has any further questions about fallacies (the ones explained here and/or other types) or argumentation, feel free to privately contact me so we can let this thread resume its initial intent.
