Interesting. Can you give an example of this?
A possible example might be the "quants" (
http://en.wikipedia.org/wiki/Quantitative_analyst) as an example of Ne/Si, vs a more "hands-on" approach to investing (Ni/Se). If you've not heard of "quants", they're the kind of people who came up with the infamous "derivatives" on mortgages to manage risk. (This isn't a knock on quantitative analysis, but it will highlight some of the flaws that can arise.)
What quants essentially do is mathematically analyze everything that "has happened" in markets (Si), and then extrapolate patterns from that to make sound investment decisions. Such analysis is inherently statistical, and for the most part it works well. What's the missing piece? Cause and effect. The math is derived what historical examples, so, on average, they'll be right and tend to make money. Such models necessarily model
trends, but they cannot model the
changes in trends, because, well, the changes haven't happened yet.
This is where Ni comes in: Ni looks at the same historical data that Ne/Si does, but instead of figuring out trends and handling special cases, Ni tries to internalize a "story" of how the changes take place. You can hear a simplistic version of this in stock market reports, where the newscaster says, "Stocks are up on news of <good market news>" or "Stocks are down on news of <bad market news>". (There is always good market news and bad market news, the story writers just insert the appropriate version of the "news" to "explain" why the market did well or poorly. And yes, this is a typical Ni (and Se) mistake, but Ni doms tend to make this mistake in less obvious ways.)
So in this case, for example, Ni has a "story" model of how market bubbles work, and knows what market bubbles "look like" (Se). So analysts such as Peter Schiff (
http://en.wikipedia.org/wiki/Peter_Schiff#Economic_and_Public_Policy_Views) see a housing bubble coming, even as most analysts do not. There are youtube videos where you can watch Schiff explain the coming crash to a bunch of skeptical fellow analysts, who scoff at his analysis. Why? There's nothing in the
market data (Si) that says a crash is coming: everything is positive, people are making money, and there's plenty of room for growth.
But Schiff looks at the same market data in an Ni way, and the data to him is a retelling of the "bubble story". He sees real estate prices going up not because people need and want more housing for themselves, but investors are buying housing only to resell it at a higher price. On top of that, he sees the highly-leveraged zero-down-payment, no-interest loans (basically, you "buy" a house by "renting" it) as a typical example of the kind of too-easy credit that fuels bubbles. In short, the "real price" of housing is how much a real person would pay to live in or otherwise use it for a productive purpose (renting apartments, as a business office, etc.), but turning housing into a commodity investment had raised the price of housing to levels far out of the range a typical buyer would be willing or able to pay. Eventually, we'd have people who'd paid too much for housing entirely unable to sell it without losing money. The core insight is that housing has a purpose, and the price of housing should reflect that purpose. If housing is costing far more than its purpose would indicate, and people are having to borrow more than they can realistically afford to pay for it, the bubble will burst and prices will go back to what people can afford.
Keep in mind, a lot of this is very obvious in retrospect: the story has been told many times and has become part of the narrative of the crash four years ago. But in 2005, it was not obvious to most people or most analysts. And this is where Ni comes in: it takes these kinds of narratives and sees how they apply to other situations. For example, an Ni analyst might say that we can expect a higher education bubble, as tuition prices rise to levels that no one can afford and don't justify the employment one might expect to find with the degree achieved. Education has a purpose: if it starts costing so much that people have to borrow more than they can realistically afford to pay for it, the bubble will burst and prices will go back to what people can afford.
I'm giving you this example showing Ni in a positive light because you requested it. The negative version of Ni would be using anecdotal evidence to arrive at incorrect conclusions, usually because the anecdote really doesn't contain the details necessary to apply it in general. In this positive case, it's still "anecdotal evidence" in that the reasoning is based off of "the asset bubble story", which isn't simply anecdotal evidence, but a fairly sophisticated cause-and-effect analysis.
The Ne/Ni crosstalk comes from Ne habitually rejecting the Ni story-based reasoning as lacking supporting data, while Ni rejects the Ne analysis as overly reliant upon statistical correlation and trends. Both can be very sophisticated and intelligent - and both can even be right. But even when they're both right, Ne and Ni
believe that they're right
for different reasons. To Ne, the statistical analysis with lots of data is convincing. For Ni, the story, the understanding of the "how" is what is convincing.