More evidence that different devices fuel news consumption at different times