Too many investors -- not to mention analysts and financial journalists -- suffer from the same affliction: a short attention span. A bit of news here, a rumor there and suddenly we have a trend.

Case in point: the March quarter. No sooner had it ended than a flurry of warnings, particularly by software makers, threw many of us into a tizzy. And when

IBM

(IBM) - Get Report

sprang a surprise miss on the market on April 14, there was such consternation that it is probably a good thing windows in most offices can no longer be opened.

Remember what happened the next day? The

Dow Jones Industrial Average

closed down 191.24 points, or 1.86%, to 10,087.51, marking a three-day decline of more than 420 points, or 4%. The

S&P 500

fell 1.67% that day, while the

Nasdaq Composite

fell nearly 2%, bringing its three-day loss to 4%.

But five weeks later, the "dead" technology sector is back in favor. The tech-heavy Nasdaq is once again north of 2000, closing Monday at 2,056.65 for a gain of 4.5% over two weeks, while the Dow reached 10,523.56, a gain of 1.7% over the same time period.

And the big cigars of the investment banking world are once again urging investors to buy tech stocks; in the last week two major houses have upgraded the software sector, while a third boosted the entire technology group.

Meanwhile, some sell-side analysts are now calling for an end-of-year rally in technology, which implies that investors need to buy ahead of it.

How could the sector heal itself so quickly? Actually, it didn't. Technology companies aren't much different in May than they were in April. Simply put, the first quarter was never as bad as some people thought, and the next quarter or two may not be as good as some are thinking now.

"I'm generally favorable on the tech sector," says Loomis Sayles analyst Tony Ursillo. "Valuations are attractive despite recent appreciation, companies are being more responsible in capex and spending on acquisitions and share buybacks, and companies are managing more reasonable objectives," he adds.

While not stellar, tech's first quarter of earnings was more of an average one than a stinker. In a typical quarter, 59% of the tech stocks will beat expectations, 21% will match and 20% will disappoint, according to Thomson Financial analyst John Butters. With 72 of 80 companies reporting so far, 57% did better than expected, 26% equaled expectations and 17% were subpar.

Bottom-line growth averaged a modest 15% year over year, compared with 68% growth in the first quarter of 2004. Butters points out that last year's first quarter was exceptionally strong; tech stocks grew by 21% in the first quarter of 2003.

Besides misjudging the overall strength of the quarter, some investors either watched the wrong companies or drew the wrong conclusions from the right companies.

Many companies that warned early were relatively small, like

Borland

(BORL)

,

TheStreet Recommends

Ariba

(ARBA)

and

RSA Security

(RSAS)

.

IBM and

Siebel Systems

(SEBL)

are certainly large, but IBM's sales drop-off in March proved to be the exception to the rule, while Siebel's poor quarter had to do with a variety of internal factors, including execution so poor the company fired its CEO.

Similarly,

Linear Technology

(LLTC)

issued targets below Wall Street's expectations, citing "general concerns in the macroeconomic environment," but overall chips pulled out a reasonably good quarter.

Sound like a mixed bag? It is, and that worries Brent McGibbon, head of McGibbon Asset Management. "Lack of definition

consistent pattern is a problem. There's no clear path to see when techs will be strong," he says.

Still, there's plenty of talk that technology stocks will take off at the end of the year. A late surge has been seen in six of the last seven years, and with a basket of bellwether technology issues trading at or close to a market multiple, it may well happen again, Goldman Sachs analysts Laura Conigliaro and Rick Sherlund said in a recent note.

"In recent history, tech has rallied to end both good and bad years and has started earlier in the past two years, moving from October to mid-August," they wrote. The rally could start as early as July when the second-quarter preannouncements come to an end, "but investors must first slog through difficult seasonal business, including the overhang of a very weak Europe."

Ursillo called the June quarter a potential pothole on the way to a late rally. "The second quarter tends to have more disappointments than any other." That's because investors usually figure that the second quarter, sandwiched between two typically weak periods, will be strong. The problem, he says, is that expectations are sometimes too high, so companies that report in line, rather than with upside, are punished.

Interestingly, in that same note, the Goldman analysts also saw signs of weak IT spending. Growth in 2005 IT operating budgets -- and in IT capital budgets -- will slow to 3% and 2%, respectively, from earlier readings of 3.6% and 2.9%, according to a survey of 100 IT managers with purchasing authority at large multinational corporations.

Another rally stopper could be the

Federal Reserve

. Unless Chairman Alan Greenspan gives some indication that the Fed is nearly done raising interest rates, "there won't be much of a rally," says McGibbon.

Then there's the point of view of Gus Zinn, an analyst with Waddell & Reed Investment Management. Although he concedes that tech spending has gotten back-end loaded, he's not at all sure that implies the late rally. "If everyone expects it, there won't be one. We just try and find the best companies," he says.

If Zinn is correct, the lesson of the last quarter becomes a harsh one. Technology investors need to be stockpickers. The sector may simply be too uneven to make money any other way.