Systematic errors are unlike random errors in that they change the observed value from the true value in a consistent direction. One example would be if you read a thermometer from the wrong angle; you might consistently read the level of the liquid inside to be a few degrees higher than it actually reads. This would systematically bias the observed value toward a higher temperature than the true value. Another example would be if a scale were calibrated incorrectly, such that it read 5 grams even when no weight was placed on it. This could result in an observed mass that is systematically higher than the actual mass.