I was a relatively slow week in biotech land, but there was still a lot of good discussions on Twitter about individual companies and the industry as a whole.

Nice call, with a good analogy, Brad. Investors initially worried about issues over median overall survival, which wasn't reached, versus reduction in risk of death, which appears to be the best of the current offerings, in Medivation's (MDVN) clinical trial testing Xtandi. The Street eventually got it right and Medivation went on a three-day tear.

Investors who had the guts to buy when Medivation went negative for a bit on Tuesday are sitting on a substantial 20% gain.

I'm not sure what scientific paper this is from, and perhaps it's best to keep the guilty party anonymous because, as Matthew points out in his hashtag, "slight evidence" of significance is just another way of saying that it wasn't statistically significant.

P-values tell you the probability that an observed difference occurred because of chance. A p-value of 0.05, which means that there's only a 5% chance that the difference happened because of chance, is the gold standard. In general, anything below that is considered statistically significant.

Sometimes drug companies -- like the author in this paper -- will try to claim that a p-value larger than 0.05 is close and therefore acceptable. If it's a phase 3 trial, investors need to understand the FDA is unlikely to accept it no matter how much the company insists there's some evidence of statistical significance.

If it's a phase 2 trial, investors can be a little more forgiving if the difference is clinically meaningful. The p-value is determined by the magnitude of the difference and the size of the population. Larger phase 3 trials can be statistically significant with the same results as a smaller phase 2 trial.

Of course the gamble is that the phase 2 trial results were just a fluke. That's, of course, what not being statistically significant means.

It sure did seem like the biotech bubble was bursting at the end of last week, but the Nasdaq Biotechnology Index has roared back close to the all-time high it hit earlier this month. I still think we're in for a major correction when generalists leave the sector after realizing it really is as risky as they say.

We saw a little of that after ARIAD Pharmaceuticals (NASDAQ: ARIA) crashed and burned and then caught fire. ARIAD has made a bit of a comeback, up 35% off its lows, which I take as a sign that casual investors overreacted and sold for whatever they could get, allowing value investors to pick up the pieces.

How many more failures will it take before investors start selling their winners? Or will there be some other trigger? Timing the top is never easy.

This week, Depomed (ASRT 5.11%) sold future royalties and milestone on a few diabetes drugs its helped develop to PDL BioPharma (PDLI) for $240.5 million. The deal is considered nondilutive because the biotech was able to raise cash without increasing the share count.

There's no doubt that dilutive financing hurts investors in the long term, but in this market, as Nathan points out, investors are more tolerant of capital raises. With shares up 27% over the last year, it's tempting to argue that Depomed should have just sold shares rather than future revenue, but with a market cap around $425 million, it's unrealistic to think Depomed could have raised as much money as it was able to get PDL BioPharma to pay out.

Um, yeah, sorry about that.