Using 1,600 common searches, the test found that Siri accurately acted upon the user's request 68 per cent of the time, in a quiet environment, and 62 per cent of the time on a loud street.
It performed much better in terms of actually recognising the user's speech, understanding words and sentences 89 and 83 per cent of the time, respectively.
Piper Jaffray analyst Gene Munster, who led the study, said Apple will be able to improve the performance of Siri in years to come, but described using the current service as a lottery.
A plan to be more competitive
"You're playing the lottery when you're using Siri," he said. "They have a plan to be more competitive, but it's going to take a couple of years.
"Apple right now gets a 'B' in comprehension and a 'D' in accuracy. There's a big difference between comprehension and her actually doing what you want her to do."
The test showcases what many users already know; that Siri is a work in progress and not the all-singing-all-dancing reason to buy an iPhone 4S, as Apple's marketing material often suggests.
Via: Washington Post
Are you a pro? Subscribe to our newsletter
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
A technology journalist, writer and videographer of many magazines and websites including T3, Gadget Magazine and TechRadar.com. He specializes in applications for smartphones, tablets and handheld devices, with bylines also at The Guardian, WIRED, Trusted Reviews and Wareable. Chris is also the podcast host for The Liverpool Way. As well as tech and football, Chris is a pop-punk fan and enjoys the art of wrasslin'.