Calspan Corp., the Cheektowaga research company that helped set safety standards for child car seats, performed the controversial tests that caused Consumer Reports to retract a report condemning all but two makes of car seats for infants.
Consumer Reports and Calspan on Friday both confirmed initial reports in the Detroit News that testing for the car seat study in the February issue of the magazine were conducted at Calspan.
The side-impact tests performed by Calspan, which Consumer Reports said occurred at 38 miles per hour, were actually in excess of 70 mph, the National Highway Traffic Safety Administration reported.
Video of the test shows the rear-facing car seat tossed around like a rag doll after the impact.
"Our initial review of the Consumer Reports testing procedures showed a significant error in the manner in which it conducted and reported on its side-impact tests," the agency said in a statement Thursday.
The testing mistake, still unexplained, caused a major embarrassment for Consumer Reports, whose brass-knuckle reporting relies on accurate testing, and for Calspan, which conducts hundreds of automobile crash tests for the government and private industry.
The government in September extended a $43.4 million contract to Calspan to study how cars and trucks fare in crashes and what causes injuries in those accidents.
Douglas Love, a spokesman for Consumers Union, the magazine's parent company, said the magazine and Calspan are conducting a full review to find out what happened.
He could offer no explanation on how the impact speeds were nearly doubled.
"We are, as the statement on our Web site says, doing a full internal review of this," Love said. "At this point, we're just two days into that process, so it's way too early for us to say."
Lissa Carroll, a spokeswoman for Calspan, also declined to offer any explanations, other than to say Calspan and Consumer Reports are working as quickly as possible to do a thorough review.
Although many cars now legally travel at speeds near 70 mph, industry experts say impacts at those speeds are rare and explain why lower speeds are used.
The Consumer Reports story carried three photos of an infant car seat knocked loose at impact with the headline: "What if this were your child?"
The article caused an immediate stir as it reported that nearly all the infant seats it tested had failed.
The story urged federal officials to order a recall for one of the seats, the Evenflo Discovery, saying it failed to meet minimum standards.
A company spokeswoman said Friday its infant seat had been tested more than 200 times in the last two years and had met or exceeded federal standards.
Although the Consumer Reports story and its findings were retracted, the February issue had already been mailed to subscribers, and it is on the magazine racks. Those issues, the company said, would not be recalled.
Lost in the controversy was the original premise for the story: that the federal government's impact standards for car seats are not as tough as they are for the cars themselves.
The magazine reported that car seats cannot be sold unless they can withstand a 30-mph frontal crash.
But most cars themselves, Consumer Reports said, are also tested for 35 mph frontal crashes and 38 mph side crashes. Car seats don't have to meet those standards.
The impact research done at Calspan tested the infant seats for both the higher speed frontal and side crashes, 35 mph and 38 mph respectively, the magazine reported.
"When we crash-tested infant car seats at the higher speeds vehicles routinely withstand, most failed disastrously," the magazine reported.
"The car seats twisted violently or flew off their bases," the article added, "in one case hurling a test dummy 30 feet across the lab."
The car manufacturers immediately challenged the results, and the National Highway Traffic Safety Administration said its initial finding showed the side-impact speeds were more than 70 mph, not 38 mph as reported.
The agency took those results to Consumer Reports on Jan. 2, which eventually led the magazine to retract its findings Thursday and launch an investigation.
Rae Tyson, a spokesman for the federal agency, said the testing controversy will not injure the long-standing relationship that Calspan has with the government for crash tests and highway safety.
"Calspan has done a great deal of work for the agency and will continue to do so," said Tyson.
The seeds of Calspan began in 1943 when it was founded as part of the Curtiss-Wright Airplane Division in Buffalo, and it was part of the Cornell Aeronautical Laboratory from 1946 until it became Calspan in 1972. A series of corporate owners gave way to a return to local control in 2005.
Calspan is credited with creating the first automobile crash-test dummies in 1948, and its work 40 years later led to the creation of more stringent standards for car seats and how they are secured.
Calspan says its tests over the years also have led to improved designs for seat belts, safer tires, studded snow tires and impact-absorbing highway guardrails. The company employs about 240 workers. Consumer Reports said it would disclose the results of its investigation once it is complete.
"We have a strong obligation to our subscribers and to others who heard this report," a company spokesman said, "to say what happened here."