Question 1118331
With or without a bonus has at least two types of bias.  One is those who get the bonus may have more incentive to do more, and those who do not will not have that incentive and may even perform less.  If the bonus is randomized, the possibility for abuse in both directions exists.  Also, bonuses are usually given for good performance which biases such a study a priori.

Same subjects as their own control.  Compare sunscreen by putting it on one arm of a subject but not the other.  Or compare shoe wear by having a person wear a pair with one shoe treated differently from the other.

As for an example the one below with the bonus where one would not want to use a subject as his or her own control.  Nor can we do it with some interventions like vaccination, where the time period is important and we need to offer or not offer the treatment to two different groups at the same time.