I'm hesitating between these two and I don't know which one is more correct: wider width or larger width?
Google Books evidence shows a preference of larger vs wider, but both are correct.
Width affected only the accuracy of the slope estimate (accuracy was better for the smaller width) and the bias of the width estimate (bias was smaller for the larger width).
There should be no decision on this measure until a full study is completed. Then, if the wider width is found to be safe — which we doubt at this time — action could be taken.