How do you decide what to buy, whether computer related or otherwise, and whether on the Web or at a brick-and-mortar store?

No matter if you're buying for yourself or your family or for a small business or multinational corporation, there are savvy ways of buying and not so savvy ways.

Despite manufacturers' or vendors' warrantees, there's no guarantee that you'll make a smart choice, even with your best efforts. But you can put the odds in your favor.

Many people when buying rely on their own experiences or those of family, friends, or colleagues. Others rely on whether or not the brand, manufacturer, or service provider is widely known. These methods shouldn't be dismissed and can be incorporated into the buying decision.

But anecdotal evidence from a small number of users or from a reputation that may have been generated by past rather than current performance or by a multimillion dollar advertising campaign can be far from foolproof.

Surveys of large numbers of people can go a long way to balancing the biases of anecdotes and reputation. Consumer Reports magazine, computer magazines, trade and specialty publications, industry Web sites, nonprofit organizations, and market research firms are among those that try to scientifically gauge people's experiences with products and services.

"Lies, damned lies, and statistics" is a phrase, popularized in the 19th century by Mark Twain, that accurately describes the limitations of surveys. Statistics can be manipulated in intentional ways and be less than accurate in unintentional ways, depending on the wording of the questions, the types of people surveyed, and how the data is processed.

A new and improved Internet Age cousin of surveying goes by the name of "crowdsourcing." Web sites such as Amazon.com, Yahoo Shopping, Netflix, Google, and many others try to tap into the collective wisdom of users through various innovative means. Amazon.com, for instance, lets users write reviews of the myriad products it sells. Other users can then rate the usefulness of the reviews written. Reviews deemed most useful are displayed first.

Reading reviews, in general, can be an excellent way to buttress the information you glean from other sources, but here as well there are limitations and precautions you should take.

Some reviews are written by professional reviewers, writers, or journalists and may appear in print, on the Web, or both. Most reputable publications and Web sites have strict rules that prohibit reviews from being influenced by whether or not the company advertises with it, but not all do. Most reputable journalists prevent any free access they may have to a product or service from influencing their evaluations, but not all do.

Amateur reviewers, writing without being paid for their work, can also be a good but not perfect source of buying information. Regular users of products or services, whether writing at their own blogs or other Web sites, often have in-depth experience with the category and are passionate about delving into the various nuances.

Some companies, however, have tried to "manage" the amateur review process. Reports have periodically surfaced in the media of companies planting fake reviews or secretly reimbursing users to write favorable reviews. The Federal Trade Commission has gotten involved with this, issuing guidelines in 2009 against fake online reviews and fake blog posts and ruling in 2010 against a public relations firm that was planting good reviews online on behalf of its clients.

Whether written by professionals or amateurs, the best reviews of any product or service aren't "first look" impressions but are written after the reviewer has had enough time to really get to know it. The best reviews also place the product or service in context, comparing it with similar offerings rather than just describing its features and discussing its benefits.

Various ways exist to help you determine the validity of any given review you come across.

Be wary of reviews that are overly positive or negative. Be careful also of clever reviews that include only a minor negative or a minor feature that's missing.

Discount reviews that describe the reviewer in too much detail, which may indicate that a PR firm is targeting the particular demographic represented by the reviewer's description.

Don't give much weight to reviews that just talk about impressive-sounding features you won't use or that spew out too much jargon, which may be written by somebody just showing off.