According to Wikipedia, a "frame in social theory consists of a schema of interpretation—that is, a collection of stereotypes—that individuals rely on to understand and respond to events." In other words, framing is a kind of cognitive tunnel vision in which a schema of interpretation makes relevant data appear irrelevant.
Like many other cognitive biases, this one seems all-but-universal. The Wikipedia entry on framing tends to take a point-counterpoint approach that contrasts frames, instead of illustrating how to think beyond them. It could be that framing is simply the incorrect use of focusing, which we need to do to think at all.
There's no question that we like narratives, and tend not to like them interrupted with a counterexample unless explained as not impugning them. Look at these two examples to see what you prefer:
Example 1: Dividends should be reinvested because they boost the portfolio return significantly. A look at the growth of two portfolios, which only differ on full reinvestment or lack of, shows it. Although this approach means refraining from drawing income from the portfolio, the benefit of reinvestment sheds light on the true opportunity cost of doing so.
Example 2: Dividends should be reinvested because they boost the portfolio return significantly. A look at the difference in growth between a portfolio that reinvests all dividends and one that reinvests none illustrates this point. The point, however, is irrelevant to someone who would rather take out the dividend income from his/her portfolio after taking into account the opportunity cost of doing so.
The first example uses a frame. The second does so too, but calls attention to it through adding a counterexample. Which one's easier to assimilate?
Example 2 might remind you of an academic paper. Academics tend to de-frame their papers by limiting the scope and/or pointing out relevant exceptions. The most famous example of de-framing was Albert Einstein's triple test of his own General Theory of Relativity. He was careful enough to say that it would be invalidated if any one of those three predictions didn't pan out.
In the less recondite area of investment, spotting frames is as easy as getting your hands on a research report. Perhaps inevitably, for the sake of reader interest, the financial press is full of framing. Wise consumers of financial news just take it in stride. Some bloggers poke fun at the process.
In the more serious area of investment decisions, though, framing does cost. Typically, framing nurtures an already-existing cognitive bias. Frames go well with, say, overoptimism bias or confirmation bias. When it comes down to brass tacks, the only way to avoid framing is to tease out assumptions from an investment strategy. I've found that doing so is an acquired skill, and is often hard to learn.
Example 1 above assumes that every rational dividend-collecting investor wants to boost his/her return without thought of any extra-portfolio monetary needs. This entire blog assumes that non-ETF stocks with market caps of 500M or more, P/Es in the lowest quintile as screened by Google Stock Screener, and yields greater than the inferred yield of S&P SPDRs, constitute a set of stocks that's worthwhile to track. Consequently, it's a value-investing niche blog. Some people, although convinced of the statistical superiority of low P/E stock returns, think that the dividend requirement is irrelevant. Others think that the low P/E effect is but a chimera, or an anomaly that won't last. Investors in cyclical stocks know that cyclical stocks tend to exhibit low P/Es near their tops rather than near their bottoms. A deep-turnaround specialist knows that some turnaround stocks show a low P/E long before they're buyable.
As I hope I've shown, I'm aware of (some of) the limitations of this blog's underlying investment philosophy. Had I been more subject to framing, though, I would have brushed aside those exceptions...or the fact that the low P/E anomoly is statistical in nature.
In the everyday world, framing is known as "assuming." There's a '70s-era witticism that says, "if you assume, you make an ass out of u and me." It's not very academic, but it gets the point across.