In the vast majority of social interactions, "there is an exception to every rule" is an understatement. Most of the time the context of the interaction is far more important than the social universals that we could back up with evidence.

As such, IPS has a challenge. If I start a question with "I have a 30 year long relationship with X" and then summarize the relationship in two paragraphs, there's literally years of context that's missing. And, most likely, the context that's missing is the important part. If the OP knew to include a particularly important piece of context, there's a good chance they could also arrive at the answer without help. The mere fact that someone is asking the question suggests they are not in a good position to provide the information.

Thus, most answers will by necessity be suggestions that must then be re-introduced into the context by the OP as they decide whether to implement or not implement the solution. The most promising suggestions will be up-voted and rise to the top.

I'd point to Worldbuilding as an example of a StackExchange that needed to walk this balance. Their solution is to have tags such as "hard-science" which state the level of rigor and evidence expected in the answers. There have been more than one question on WB because the OP tagged it as "hard-science" and the community agreed that the question was not phrased for that level of rigor. Many of these OPs then removed the "hard-science" tag and got the question re-opened.

Tl/Dr: A back-up rule for the site would be detrimental. It's negative reinforcement, rather than positive reinforcement, and other stack exchanges have show alternatives.

We dealt with the subjectivity problem during the beta for WorldBuilding.SE. What makes the subjectivity problem tricky is that everyone has a different opinion about how subjective a question or how justified an answer should be. Some people want fully backed up and justified answers with links. Others just want common sense or logic. Others simply want the answers. For an example of how extreme these differences are, look at Mathematics.SE vs MathOverflow. A good question on Mathematics.SE could get brutally downvoted and closed on MathOverflow because their target demographics are very different, leading them come to a different line between what is a good question/answer and what is not.

On WorldBuilding, this was a rather unique problem because every question is asking for a part of an answer. You can't ask WB to write your novel for you, so you always have to ask for pieces, and then incorporate those pieces into your novel. I think IPS operates the same way. Everyone is writing questions containing a piece of their life, and must re-integrate those answers into their life before determining if they truly have merit or not.

Because these questions are asking for pieces, there really is no right or wrong answer. This isn't Physics.SE, where you can cite wikipedia's answer, or do some math. The answers aren't complete until they are integrated into the story. IPS is the same way. An answer isn't an answer until you've integrated it into your life and found that it fits correctly.

This downplays the value of backing up the answers. In many cases, there is no official answer to these questions, so all we're doing is providing subjective evidence that the answer is good.

What constitutes backing up information anyways? When it comes to Physics.SE, it's easy because there is a well understood curriculum. For IPS, the schoolroom is life. We're learning every day, and we're learning in different ways. If I had to justify many of my answers, they would be using the language of Rational Actor Theory and Chinese internal martial arts, because those are the most prominent ways I view interpersonal relationships. My justifications may have very little value to you, unless also think in these terms (which I assume you do not). Whatever justification scheme you use, I guarantee you the OP uses a different one! Indeed, those who want a "suggestions" site have a justification scheme of their own -- may the most upvoted question win! These justifications clearly have some value, but its hard to come to a consensus on how valuable they are compared to the answers themselves.

On WorldBuilding, we found that people don't all agree on how much backing up is "good." Sometimes, people are just looking for quick answers which they can then integrate. Sometimes, people are looking for detailed proofs that the answer is good. I think IPS is similar. Some people are looking to use this site more as a "suggestions site" some are looking to use it as a "expert system" which links up expert information.

The solution on WorldBuilding was to rely on tags. An OP could tag a question to indicate what level of backing up of the answer they wanted to see. Tags which have had an effect on this are:

• Science-based - Everything should be backed up to science, but it can be a loose connection
• Hard-science - An acceptable answer must reference their sources, and typically rely on mathematical calculations.

Then we let the community go wild with them, and collected statistics. What we found is that the community really liked the rules of Science-based questions. It's the most popular tag on the site. The community of people who ask and answer science-based questions have evolved towards a balance of their own. If you try to be too loose with a science-based answer, you get called on it. Meanwhile, those who want hard-science can get it.

As a general rule in interpersonal skills, it's more helpful in the long run to use positive reinforcement rather than negative reinforcement.

"Our research found that negative reinforcement is actually far more effective for sparking initial habit change.... But here’s where things get interesting: If you were continually penalized ..., the negative reinforcement would eventually stop working. You’d become resentful of the constant punishment." - Pavlok

Rules are universally backed up with negative reinforcement (I can't think of a counter example. Can anyone think of one?). What we want is to raise the health of IPS in the long run, and that's best done with positive reinforcement. Tagging the level of backup expected gives us handholds with which to provide that positive reinforcement while simultaneously listening to the community and their desires.