A paper from the School of Continuous Improvement works out what a rising tide does to a quadrat survey, and finds that its own remedy costs more than the problem on any single tide.
A simulated rocky shore, four ways of dividing a low tide, and a correction that pays for itself on about the ninth visit
A survey of a rocky shore is finished when the sea says so, and a paper released today by the School of Continuous Improvement sets out what the sea takes with it. Stopping on the Count, Not the Clock: Allocation and Censoring Bias in Quadrat Surveys Worked Ahead of a Rising Tide is by Dr Mirela Hanke, Postdoctoral Fellow and Deputy Convenor, Living Dashboard, Dr Joost Nwosu, Postdoctoral Fellow, and Dr Solveig Adeyemi, Lecturer.
We went down to the shore, in a manner of speaking, with a recommendation already drafted. The University has always left its people room to come back with a different one, and on this occasion we have used all of it.
— Dr Mirela Hanke, Deputy Convenor of the Living Dashboard
The paper brings the mathematics of sampling into alignment with the timetable of the flood tide, anchoring both in a simulated platform of five bands surveyed 10,000 times at each of seven tidal windows. It reports that a team working each band until the water arrives will tend to undercount, by a few per cent, and that a rule which removes most of that shortfall leaves the estimate less precise on any one tide than it was before. The authors put the point at which the rule begins to repay its cost at about nine tides.
The School regards the paper as a fair instance of the patience with which its researchers are prepared to follow a method to the water’s edge. “A good deal of care went into deciding when to stop,” Dr Nwosu said, “and we would not wish to hurry anyone who is still deciding.”
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.i9mrf6.