This topic comes up from time to time. Being in the auto industry, I work first hand with car companies and testing publictions so I am in a position to elaborate on the subject.
We sometimes see rather large discrepancies between acceleration test times when comparing data from different magazines. Sometimes the manufacturer given acceleration specifications are a few tenths slower than what the major car magazines publish.
Why is that?
Most of the time, it's a difference in the measurement practices. The biggest factor comes down to a phenomena called "rollout".
Rollout defined: The concept relates directly to NHRA dragstrips and the measurement methods used in quarter mile racing. For the unfamiliar, this is how it works. When you pull into the staging area or starting line, there are two light beams which are aimed perpendicular to the track. The first light beam is the "pre-staging" light. It has no real function except for telling you when you're getting close to the starting line. The 2nd light beam, called the "stage" beam, is the actual starting line. Ignoring staging strategies (shallow stage, deep stage, etc), we'll assume the car pulls up to the staging line until the beam is just barely broken by the front tires...as this red Civic shows below (the black wheel).
When the light turns green, the driver mashes the gas. However, the timing clock doesn't know that the car is moving until the front tire moves far enough away from the light beam to allow it to fully shine across the track. This is demonstrated by the grayed out tire in the below photo. In reality, the car moves a distance of about 12" taking 0.3 seconds for free...the clock hasn't started yet. Once the clock starts, the car is already moving 3 mph. Here Car and Driver discusses the importance of rollout.
-CLICK HERE TO SEE A DIAGRAM ILLUSTRATING ROLLOUT-
That brings us back to modern day test measurements. Car magazines and car manufactures don't test on dragstrips very often. They use sophisticated computerized GPS or "5th wheel" type measurement systems. A commonly used system comes from a system called a Racelogic VBOX.
How does this all tie together? Well, this fancy measurement system eliminated the need to have an optical start/stop line like a dragstrip does. However, magazines want to publish times that relate to what the average Joe can accomplish if he takes his car to the local NHRA dragstrip...so all the major US car magazines still test with a 12" simulated rollout. This also makes acceleration times look faster on paper, which of course sells too.
Car and Driver, Road and Track, and Motortrend all use this simulated 12" rollout. That means when you read any acceleration statistics in those magazines, it's not a true 0-60, 0-100, or 0-150 mph time. It's actually measuring 3 mph to 60, or 3 mph to 100, or 3 mph to 150. The car starts from a stand still, but the clock doesn't begin to run until the car has moved 12", gained 3 mph, and traveled for 0.3 seconds.
GM and Ford also use rollout when claiming their factory times.
Road & Track admits to using rollout.
Here Car and Driver states how they use rollout.
Quote from Car and Driver:
"Before you take out your car to try to equal our times, remember that our results are adjusted for weather conditions [see “Correcting for Weather,” page 152]. We also average the best runs in two directions to cancel out the effects of wind, and we use a 3-mph rollout. And of course there is car-to-car variability."
Motortrend calls rollout "US Traditional" as they use it too.
Automobile mag does NOT use rollout.
Edmunds does NOT use rollout, but GM does...
Because NHRA and dragstrips are basically non-existant outside the US...ONLY US BASED MAGAZINES AND MANUFACTURERS TEST WITH ROLLOUT.
This means that British, German, or Japanese magazines will clock times that are 0.3 seconds slower than US magazines for the same car as they time true 0-60 and true quarter mile times. They use no rollout.
That partially explains why German cars often perform much better in US magazine testing than the manufacturer specifications have you believe. People often wonder why German companies publish so conservative acceleration times, especially in 0-60 mph. The lack of rollout is one major reason.
Lastly, some magazine "test cars" aren't exactly as bone stock as you'd like. I've personally been in catless, semi tuned magazine test cars by major car companies. They all do it, don't fool yourself.