There is no perfect solution, but requiring access to both the data and full methodology for experimental sciences should help. Even problem sof cherry picked

There is no perfect solution, but requiring access to both the data and full methodology for experimental sciences should help. Even problem sof cherry picked data would be partially exposed eventually; and eventual exposure is still a very effective deterrent in science. My 2c.