If Michael Roy were to estimate how much time he has spent researching time estimation, he would likely underestimate it.
Years after writing his dissertation on how memory biases lead to prediction biases, Roy, associate professor of psychology at Elizabethtown College, studied time estimation with Tatem Burns ’16. They examined if music training increased time estimation accuracy. To conduct this research, Roy and Burns created a large, online study, in which participants estimated the time it would take them to listen to a song, fill out a survey and watch a video. Roy and Burns also researched the subjects’ total time estimates of the three activities. Roy hypothesized that musicians would have more accurate time estimations than those without musical training. Much to his surprise, there was no correlation. However, one phenomenon did surprise him—“Individual estimates were completely independent of overall estimates,” Roy said.
The basic thing is, people are just bad at estimation.”
Roy and Burns found that participants often overestimated time taken for individual activities–the song or survey–but underestimated the total amount of time. The researchers also found that individuals’ estimates were influenced by the activity completed first. For example, if a short song came first, followed by a longer survey, participants overestimated the survey time.
Desiring more control over variables, Roy conducted a second study. Three science videos of varying lengths were played in random order. Despite conducting the study differently from the first, the results were comparable.
“The basic thing is, people are just bad at estimation,” Roy said. He acknowledges that gut estimates, the ones we make daily, we don’t think through. “It doesn’t necessarily have to do with optimism or attitude toward the activity,” Roy said, “but with memory.”
In addition to Burns, the E-town psychology professor has worked with several students to conduct this research; he wants them to gain experience for graduate psychology schools, which prefer past research experience.
“I put a lot of this on them,” Roy said. They found the videos, put them together and conducted the data entry. Roy said he hopes to continue this research, potentially applying it to criminal justice and other psychology-related fields.
So, how can we be more accurate in our own time estimations? “You just have to track your time,” Roy said with a smile.