Best Online Credit Check

FICO is an acronym for Fair Isaac Corporation (traded publicly under the symbol FIC) and refers to the best-known credit score model in the United States. The FICO score is calculated using statistical methods, developed by Fair Isaac, to information in one’s credit file. The FICO score is primarily used in the consumer banking and credit industry. Banks and other institutions that use scores as a factor in their lending decisions may deny credit, charge higher interest rates or require more extensive income and asset verification if the applicants credit score is low.

FICO scores are designed to indicate the likelihood of a borrower being delinquent within the next 24 months. No public information is available to determine what the scores mean in terms of statistics. A separate score, BNI, is used to indicate likelihood of bankruptcy.

The three major credit reporting agencies (also often, but inaccurately referred to as credit bureaus) in the United States, (Equifax, Experian and TransUnion) calculate their own credit scores, which go by different trademark names as well as many different versions of the score (often differing because of what they are meant to predict and when they were written): Beacon, Beacon 5.0, Beacon 96, and Pinnacle are all available only from Equifax; Empirica, Empirica Auto 95, Precision Score, and Precision 03 at TransUnion; and Fair Isaac Risk Score at Experian. These versions, while all developed for the agencies by Fair Isaac, differ and are periodically updated to reflect current consumer repayment behavior. The NextGen Scores is the newest scoring model, but many creditors still use the Classic FICO score, developed in 1989, because it has been extensively tested, and shown to be accurate.

The scores use a multiple scorecard design. Each version uses 10 or more individual scorecards, and an individual is typically compared with similar others. (For example, a borrower with two 30-day late payments will be scored against a population with some minor delinquencies.) An individual is then graded according to what variables seem to indicate a repayment risk in that group. This feature may cause a borrower with delinquencies to score in the same range as a borrower without delinquencies.

Nearly all large banks also build and use their own proprietary statistical models for credit scoring purposes, often in conjunction with the FICO score or other outside scores.

The statistical models that generate credit scores are subject to federal regulations. The Federal Reserve Board’s Regulation B, which implements the Equal Credit Opportunity Act, expressly prohibits a credit scoring model from considering any prohibited basis such as race, color, religion, national origin, sex, or marital status. Regulation B also stipulates that credit scoring models must be empirically derived and statistically sound. Furthermore, if an adverse action is taken as a result of the credit score (e.g. an individual’s application for credit is denied) then specific reasons for the denial must be provided to the individual. A statement that the individual "failed to score high enough" is insufficient; the reasons must be specific.

There exist several generally accepted algorithms for extracting the primary contributing factors to a low credit score. One or more of these algorithms is typically used to supply a list of reasons when a loan applicant has been denied credit, in order to satisfy the Regulation B requirement that specific reasons are disclosed. Some consumers feel these adverse action reasons are somewhat disingenuous, as the only determining factor for credit denials is a numeric score — the "reasons" are summed up only for the consumer.

Each of the credit reporting agencies has developed its own version of the credit score intended to compete with Fair Isaac’s score. Although not as widely used, these scores (for example Trans Union’s "TransRisk" score or Experian’s "ScoreX" and "PLUS" scores) are less expensive than the FICO score. These scores are often derisively referred to by consumers and lenders as "FAKO" scores, for they do not use official Fair Isaac methodologies. The cost savings of a non-FICO score are tempting to some banks and credit card companies, who need an accurate risk assessment on millions of accounts every year. For ease of use, these scores tend to be mathematically scaled so that they fall in the same general range as the FICO score. Fair Isaac offers scoring models for the U.S., Canada, and South Africa. It also offers a "Global FICO" for many other countries.
References :
http://en.wikipedia.org/wiki/FICO_score

The corrolation is that people with bad credit tend to be "risky" in other areas of their life as well. I am not saying I agree with that, I am only telling you the rationale of the insurance companies.
References :

The specific correlation would depend on which risk is being measured, and which credit score (there are several besides FICO, and FICO comes in several flavors). Presumably, whatever correlation is being measured is real, but I don’t know that any definitive studies have been done on the underlying causes. I can tell you that the theory usually floated is that people who exhibit risky behaviors in one area of their lives (poor credit history) will often exhibit risk in other areas (smoking in bed or whatever).
References :