Paydays, Stress and Model-Based Decision Making

How to cite

Halpern, David; Morton, Rebecca. 2019. "Paydays, Stress and Model-Based Decision Making." AEA RCT Registry. AEARCTR-0004636. https://doi.org/10.1257/rct.4636-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

David Halpern* , Rebecca Morton

* Primary investigator

Registry ID
AEARCTR-0004636
DOI
https://doi.org/10.1257/rct.4636-1.1
First registered
Last updated
Status
In development
Trial dates
2019-09-21 – 2022-06-21
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/4636

Abstract

A growing body of work has begun to investigate how a lack of financial resources can have an impact on cognition. For instance, using natural variation in resources due to a lack of consumption smoothing before harvest-time, Mani et al. (2013) provided evidence Indian farmers had greater working memory and cognitive control capacity after a harvest than before. However, some of these results have been difficult to replicate. Carvalho et al. (2016) found minimal differences in cognitive measurements (aside from present bias) when they surveyed a group of low-income Americans before and after a payday. In this study, we make psychological measurements of low-income workers in Abu Dhabi before and after a payday and hope to add to this literature in several ways. First, we measure cortisol, the gold standard biomarker for stress (Kirschbaum and Hellhammer 1994). By measuring changes in cortisol around a payday, we can make comparisons to the large existing literature on cognition and cortisol (e.g. Kandassamy et al. 2014) to understand the magnitude of changes we might expect around a payday. Second, we measure model-based decision-making (Daw et al. 2011), a computational formulation of dual system theories of decision making. Previous studies have linked variation in model-based or goal-based decision making to stress (e.g. Otto et al. 2013, Quaedflieg et al. 2019) which suggests that there might be variation around a payday. Understanding the relationship between poverty and "model-basedness" may provide for a new understanding of decision-making under poverty. Third, we measure changes in feelings of control and perceived constraints which have been theorized to underlie some changes in cognitive processes under poverty (Pepper and Nettle 2017, Krauss et al. 2011). Finally, we attempt to improve the precision and generalizability of our measurements over previous studies by making within-person measurements and including a true control group, as well as improving our control over the timing of measurements. We hope that these additions to the literature will help improve understanding of mechanisms of poverty’s potential effects on cognition.

Intervention

Subjects will come into the lab and conduct the psychological measurements twice, one week apart. Half of the subjects will come for their first session three days before their payday and half will come four days after. Workers will all be recruited from NYU Abu Dhabi contractors so we have precise information about when their payday occurs.

Experimental design

All subjects will come into the lab for two experimental sessions 1 week apart. For half of the subjects ("before" group), the first session will take place three days before their monthly payday. For the other half ("after" group), the first session will take place four days after (the same day as the second session for the "before" group). On the day of each lab session, subjects will collect four cortisol samples: 1) morning, 2) beginning of experimental session, 3) end of experimental session and 4) evening. The lab session will take place in the afternoon so that it occurs during the most stable part of the daily cortisol cycle. During the lab session, the subjects will complete four behavioral tasks in order: 1) Dot pattern expectancy task, 2) Digit span, 3) Two-Step Task, 4) Intertemporal Choice Task, 5) Cognitive Reflection Task. They will also complete a survey including the Perceived Stress Scale (Cohen et al. 1983), Personal Mastery Scale (Pearlin and Schooler 1978), Perceived Constraints Scale (Lachman and Weaver 1998), Survey measures of financial stress adapted from (Carvalho et al. 2016 and Mani et al. 2013)

Design details

Randomization method
Done by computer
Randomization unit
Individual, blocked by shift-timing
Clusters
400 subjects
Observations
800 subject-sessions
Arms
200 subjects in the "before" group and 200 in the "after" group
Primary outcomes
Salivary cortisol, Model-basedness, Working memory capacity, Cognitive control, Cognitive reflection, Intertemporal Choice
Secondary outcomes
Perceived Stress Scale (Cohen et al. 1983), Personal Mastery Scale (Pearlin and Schooler 1978), Perceived Constraints Scale (Lachman and Weaver 1998), Survey measures of financial stress adapted from (Carvalho et al. 2016 and Mani et al. 2013)

Cited by

No citation data available for this registration yet.

Keywords