A calculator reports the solution to 2 to the x power = 10 as x = 3.3, rounded to the nearest tenth. What is the maximum possible absolute error in x due to this rounding?