But as police investigated, they eliminated the co-worker as the attacker and focused on Sheehan, in part because he had lived with the victim for eight months in 2005. When his house was searched, officers found a dress and two pairs of underwear that belonged to her.
When police asked if Sheehan was her attacker, the victim said no.
At trial, prosecutors relied "almost entirely" on the State Crime Lab's analysis that the partial palm print belonged to Sheehan, according to the ruling. The victim did not testify, but prosecutors did call the two State Crime Lab scientists who worked on the case, one of whom testified that to his knowledge, he had "never made an error."
But the judge did not allow another expert to testify who could have shed doubt on whether the palm print actually belonged to Sheehan: Simon Cole, who has written extensively about inaccuracies in fingerprinting.
The judge decided that because the print was admitted as evidence, testimony questioning the science wouldn't be allowed. But the appeals court disagreed, ruling that just because the state's forensic scientists are reliable doesn't mean a contradictory expert is unreliable. It's up to the jury to decide, especially in a case where the prosecution relied so heavily on a particular piece of evidence, the appeals court ruled.
"If Dr. Cole had been allowed to testify and challenge the reliability and validity of the State's expert witnesses, the jury could have easily found that reasonable doubt existed ... especially given [the victim's] earlier identification of a different perpetrator," Judge Michele Christiansen wrote in Thursday's ruling.
Sheehan was sentenced to five years to life in prison on the aggravated burglary and aggravated assault in 2009. Thursday's ruling will give him a new trial; no date has yet been set.