Bloomberg says that the company modified the load weight of the cars during testing, which made them appear more efficient. Mitsubishi says that it used a different running resistance figure (a combination of resistance caused by tires and air when a car is moving) than mandated by Japan's testing body.
Whether this involved physically changing the load weight, or simply changing the relevant figure in a spreadsheet is unclear, but regardless, Mitsubishi accepts that the testing method "provided more advantageous fuel consumption rates than the actual rates." And it's this discrepancy between the real-world figures and Mitsubishi's claims that was picked up by Nissan during in-house testing.
This data manipulation is neither as high-tech nor as widespread as Volkswagen's emission lies -- the German company used software to change the way its diesel cars behaved during standard testing. Nonetheless, another car manufacturer being exposed for doctoring test results does not help engender faith in the auto industry.
Mitsubishi has halted production and sales of all the affected cars -- the eK Wagon, eK Space, Dayz, and Dayz Roox -- and Nissan has stopped selling the relevant Dayz models. In total, 157,000 Mitsubishi cars are affected, and 468,000 Nissan. The pair will hold talks regarding compensation. With Nissan seemingly the innocent party here, the size of that compensation is likely to be considerable.