That pet project you love at work? New research says it might be hurting your overall job performance.

A study published in the Academy of Management Journal says that tasks we find highly motivating seem to have a negative impact on other job tasks that are less interesting. Essentially, when we get to work on something we love, it draws attention away from boring tasks and makes them seem more intolerable by comparison.

That doesn’t mean that being interested in your job is a bad thing—far from it. The study also found that tasks with low intrinsic motivation (i.e. – boring) can also hurt overall performance. In what’s known in the statistics world as a “curvilinear effect,” being required to work on high motivation and low motivation tasks had an adverse effect on participants’ performance across multiple tasks and jobs. The best performance was observed when participants completed tasks with a moderate level of interest before moving on to the rest of their work.

Participants in the study were given two tasks to complete that were either highly engaging (like playing a game or making a list of their favorite YouTube videos) or boring (like copying names from a phonebook or watching math tutorial videos). The results showed that doing an engaging task followed by a boring task hurt performance on the boring task. The study also examined real-world performance of retail workers by comparing their self-reported interested in job tasks versus how their supervisors rank their performance on each of these tasks. Results from this survey showed that employees with high motivation in one task had lower performance in subsequent, less-interesting tasks.

The authors—Jihae Shin of the University of Wisconsin-Madison and Adam M. Grant of the University of Pennsylvania—write that the findings offer insight into how we should structure our workdays in order to maximize performance. Rather than starting the day with the most unpleasant task (the popular “eat the frog” method) or the most rewarding task, “starting the day with a moderately interesting task, followed by a less interesting task, and then a highly interesting task” might lead to the best performance on each individual task during a day.

Shin and Grant compare this to the way high-performance athletes train for competitions. As an important competition draws near, many athletes will “taper” their training to be less rigorous in order to reach peak strength and endurance during the competition itself. “By tapering interest levels down gradually, it may be possible to sustain energy and effectiveness even in tasks that lack intrinsic motivation,” they write.

So next time you’ve got hours of data-entry to do, stick it right in the middle of your day and reward yourself at the end with a task you enjoy. We can’t promise it will make your day any more enjoyable, but it may lead to fewer mistakes.