William Schiavone, PhD

Nonlinear perceptions of goal behaviors

Background and Goals:

I developed and conducted these studies at the University of Georgia to expand on existing knowledge of goals and motivation.

Goal pursuit often occurs over a long period of time and people interpret goal success and failure based on smaller subgoals. These studies examined whether people feel differently about success and failure when they are close to or far from a standard.

Based on several lines of existing literature, I hypothesized people’s perception of goal behaviors follow a non-linear pattern, with smaller changes in perceptions of behaviors further from a standard.

Method:

2 experimental surveys

  • Study 1
  • 237 UGA students saw 10 dieting goal scenarios using a fictitious unit called Trexels.
    • Participants were told 100/day was average, and told they had a goal to eat less than 80/day.
  • Study 2
  • 404 MTURK workers saw 8 step goal scenarios differing by 1,300 steps each.
    • Goal of 10,000 steps/day

Both studies – Across 3 randomized blocks, assessed perceived failure, guilt, and success for all randomized scenarios

At the end of each study, participants completed individual difference measures and demographic items.

Key Findings:

  • I used mixed models to test whether ratings of behavior changed as the amount of behavior became more extreme; cubic models had the best fit.
  • Across two analytic methods, we observed converging evidence that people’s perceptions did not match objective increases in success, even when that standard was framed in the context of a larger goal.
  • Near a standard, changes in behavior had larger effects on perceptions of success compared to changes in behavior further from a standard.
Figure 1. Cubic trend of success across trexels (Study 1) and steps (Study 2). Average slope in black; individual participants’ slopes in grey.

Impact:

  • This work empirically tested predictions from several existing theories and provided evidence that people may experience changes in behavior as more impactful near a standard (vs far).
  • These studies provided a platform to identify other variables that may be associated with the ways people think about their goal progress/behaviors.

Personal takeaways:

  • Often, simple designs work best for initial tests of complex ideas
  • Breaking down processes into components can be useful for understanding some mechanics, but including other variables in tests helps to generalize findings
  • Real life goal pursuit is often messy; simplified goal scenarios can provide insight into how we might understand how evaluations of behavior initially influence someone’s thoughts about a goal