The average knowledge worker enjoys a measly five minutes of uninterrupted time and, once interrupted, half won’t even get back to what they were doing in the first place. Yet organizational expectations or social pressures make it hard to resist the urge to check incoming emails or text messages — pressing tasks be damned.

Research suggests that such interruptions are not necessarily bad, and can even be productive.

“I know from personal experience that some interruptions are actually good,” says Shamel Addas, an assistant professor of information systems at Smith School of Business in Ontario, Canada.

“It depends on the content and timing. You can get some critical information that will help, like completing your task. So that’s one of the assumptions I feel needs to be challenged and tested.”

Addas has conducted several studies to learn more about the relationship between technology-related interruptions and performance.

He found that interruptions that did not relate to primary activities undermined workers’ performance — they led to higher error rates, poorer memory, and lower output quality. It also took longer for workers to return to their primary work and complete their tasks. Such interruptions also had an indirect effect on performance by increasing workers’ stress levels.

Email interruptions that related to workers’ primary activities increased stress levels as well but also boosted workers’ performance. Such interruptions were tied to mindful processing of task activities, which led to better performance both in terms of efficiency and effectiveness.

Addas also discovered that the very features of the interrupting technology can influence the outcomes of the interruptions positively or negatively. Individuals who engaged in several email threads of conversations at the same time or those who kept getting interrupted by email but let the messages pile up in their in-box experienced higher stress levels and lower performance.

And those who reprocessed their received messages during interruptions episodes or rehearsed their message responses before sending saw some benefits. Doing this enabled them to process their tasks more mindfully, Addas says, which boosted their performance.

What does this mean for managers? For one thing, just recognizing that there are different types of interruptions, each with its own trade off, can help managers mitigate the negative impacts on performance and stress.

Addas suggests that managers develop email management programs and interventions, such as specifying a time-response window for emails based on their urgency or relevance to primary activities. They can also establish periods of quiet time for uninterrupted work. And they can encourage work groups to develop effective coordination strategies to ensure one person’s interruptions do not adversely affect colleagues.

As for individuals, they can start handling interruptions in batch rather than in real time to reduce the costs of switching back and forth between tasks and interruptions. To reduce stress from overload, Addas suggests, people should limit parallel exchanges during interruptions and delete or folder messages that are of limited use for their core work.

“People might well consider thinking about the messages they construct and examining carefully their previously received messages as needed to ensure that they process their tasks more mindfully, which is beneficial for performance,” Addas says.

Addas believes there are design implications to consider as well, particularly relating to context aware systems and email clients. Context aware systems know what kinds of tasks people are working on and can detect high and low periods of workload. “Email clients can be programmed to screen messages for task-relevant content and distinguish between incongruent and congruent interruptions,” he says. “They can then manipulate the timing at which each type of interruption is displayed to users, such as masking incongruent interruptions until a later time.”