Hannah Marie Kyllo - Academia.edu (original) (raw)

Uploads

Papers by Hannah Marie Kyllo

Journal of Individual Differences, 2020

Many psychology researches are performed through "psychology subject pools" which give participan... more Many psychology researches are performed through "psychology subject pools" which give participants considerable flexibility in when they participate. This "participant degree-of-freedom" has led to concern that the characteristics of subject pool participants may change with time, with the most engaged students signing up at the start of the semester and the least engaged students leaving it all to the end. In this paper we performed an exploratory analysis to look for evidence of this "good student effect". Consistent with previous work, we find support for the good student effect with earlier participants scoring higher on the Big-5 subscales of Achievement-Striving and Cooperation, as well as Grit and Empathic-Concern. In addition, we found a non-linear effect of time-of-semester on Sensation-Seeking, with this measure peaking in the middle of the semester as well as the end. However, the vast majority of the measures we tested, including measures of personality, cognition, decisionmaking and social interaction, did not correlate with time-of-semester or timeof-day at all. Thus, we conclude that, while some studies directly related to measures of Grit and Sensation-Seeking would do well to recruit throughout the semester, in most cases any bias introduced by the good student effect is likely to be small.

Scientific reports, 2020

The explore-exploit dilemma describes the trade off that occurs any time we must choose between e... more The explore-exploit dilemma describes the trade off that occurs any time we must choose between exploring unknown options and exploiting options we know well. Implicit in this trade off is how we value future rewards-exploiting is usually better in the short term, but in the longer term the benefits of exploration can be huge. Thus, in theory there should be a tight connection between how much people value future rewards, i.e. how much they discount future rewards relative to immediate rewards, and how likely they are to explore, with less 'temporal discounting' associated with more exploration. By measuring individual differences in temporal discounting and correlating them with explore-exploit behavior, we tested whether this theoretical prediction holds in practice. We used the 27-item Delay-Discounting Questionnaire to estimate temporal discounting and the Horizon Task to quantify two strategies of explore-exploit behavior: directed exploration, where information drives exploration by choice, and random exploration, where behavioral variability drives exploration by chance. We find a clear correlation between temporal discounting and directed exploration, with more temporal discounting leading to less directed exploration. Conversely, we find no relationship between temporal discounting and random exploration. Unexpectedly, we find that the relationship with directed exploration appears to be driven by a correlation between temporal discounting and uncertainty seeking at short time horizons, rather than information seeking at long horizons. Taken together our results suggest a nuanced relationship between temporal discounting and explore-exploit behavior that may be mediated by multiple factors. The explore-exploit dilemma refers to a ubiquitous problem in reinforcement learning in which an agent has to decide between exploiting options it knows to be good and exploring options whose rewards are unknown 1. For example, when ordering sushi at a favorite restaurant, should we exploit our usual favorite (the Rainbow Roll), which is guaranteed to be good, or explore the Burrito Roll, which could be delicious, disgusting or somewhere in between. As anyone who has agonized over a dining decision will know, making explore-exploit choices can be hard, and there is considerable interest in how these decisions are made by humans and other animals 2. Recently, a number of studies have shown that people make explore-exploit decisions using a mixture of two strategies: directed exploration and random exploration 3-8. In directed exploration, choices are biased towards more informative options by an 'information bonus, ' that increases the relative value of unknown options 9. In random exploration, behavioral variability, perhaps driven by random noise processes in the brain, causes exploratory options to be chosen by chance 1,10. Further work suggests these two types of exploration have different computational properties 4 , age dependence 11 , and may be controlled by different systems in the brain 12-15. Regardless of the type of exploration, the benefits of exploring over exploiting lie in the possibility of earning larger rewards in the future. For example, in our restaurant example, if the Rainbow Roll is an above-average item on the menu, then, in the short term, exploiting it will usually be best. In the longer term, however, if the Burrito Roll turns out to be sublime, then we could order this roll again and again for years to come. Thus, how much we care about future rewards, that is how we discount them relative to immediate rewards, should play a critical role in how we make our explore-exploit choice.

Journal of Individual Differences, 2020

Many psychology researches are performed through "psychology subject pools" which give participan... more Many psychology researches are performed through "psychology subject pools" which give participants considerable flexibility in when they participate. This "participant degree-of-freedom" has led to concern that the characteristics of subject pool participants may change with time, with the most engaged students signing up at the start of the semester and the least engaged students leaving it all to the end. In this paper we performed an exploratory analysis to look for evidence of this "good student effect". Consistent with previous work, we find support for the good student effect with earlier participants scoring higher on the Big-5 subscales of Achievement-Striving and Cooperation, as well as Grit and Empathic-Concern. In addition, we found a non-linear effect of time-of-semester on Sensation-Seeking, with this measure peaking in the middle of the semester as well as the end. However, the vast majority of the measures we tested, including measures of personality, cognition, decisionmaking and social interaction, did not correlate with time-of-semester or timeof-day at all. Thus, we conclude that, while some studies directly related to measures of Grit and Sensation-Seeking would do well to recruit throughout the semester, in most cases any bias introduced by the good student effect is likely to be small.

Scientific reports, 2020

The explore-exploit dilemma describes the trade off that occurs any time we must choose between e... more The explore-exploit dilemma describes the trade off that occurs any time we must choose between exploring unknown options and exploiting options we know well. Implicit in this trade off is how we value future rewards-exploiting is usually better in the short term, but in the longer term the benefits of exploration can be huge. Thus, in theory there should be a tight connection between how much people value future rewards, i.e. how much they discount future rewards relative to immediate rewards, and how likely they are to explore, with less 'temporal discounting' associated with more exploration. By measuring individual differences in temporal discounting and correlating them with explore-exploit behavior, we tested whether this theoretical prediction holds in practice. We used the 27-item Delay-Discounting Questionnaire to estimate temporal discounting and the Horizon Task to quantify two strategies of explore-exploit behavior: directed exploration, where information drives exploration by choice, and random exploration, where behavioral variability drives exploration by chance. We find a clear correlation between temporal discounting and directed exploration, with more temporal discounting leading to less directed exploration. Conversely, we find no relationship between temporal discounting and random exploration. Unexpectedly, we find that the relationship with directed exploration appears to be driven by a correlation between temporal discounting and uncertainty seeking at short time horizons, rather than information seeking at long horizons. Taken together our results suggest a nuanced relationship between temporal discounting and explore-exploit behavior that may be mediated by multiple factors. The explore-exploit dilemma refers to a ubiquitous problem in reinforcement learning in which an agent has to decide between exploiting options it knows to be good and exploring options whose rewards are unknown 1. For example, when ordering sushi at a favorite restaurant, should we exploit our usual favorite (the Rainbow Roll), which is guaranteed to be good, or explore the Burrito Roll, which could be delicious, disgusting or somewhere in between. As anyone who has agonized over a dining decision will know, making explore-exploit choices can be hard, and there is considerable interest in how these decisions are made by humans and other animals 2. Recently, a number of studies have shown that people make explore-exploit decisions using a mixture of two strategies: directed exploration and random exploration 3-8. In directed exploration, choices are biased towards more informative options by an 'information bonus, ' that increases the relative value of unknown options 9. In random exploration, behavioral variability, perhaps driven by random noise processes in the brain, causes exploratory options to be chosen by chance 1,10. Further work suggests these two types of exploration have different computational properties 4 , age dependence 11 , and may be controlled by different systems in the brain 12-15. Regardless of the type of exploration, the benefits of exploring over exploiting lie in the possibility of earning larger rewards in the future. For example, in our restaurant example, if the Rainbow Roll is an above-average item on the menu, then, in the short term, exploiting it will usually be best. In the longer term, however, if the Burrito Roll turns out to be sublime, then we could order this roll again and again for years to come. Thus, how much we care about future rewards, that is how we discount them relative to immediate rewards, should play a critical role in how we make our explore-exploit choice.