The human subjects research (HSR) studies under ReSCIND have amassed a wealth of cyber and psychological data. This data was collected under differing conditions to measure cognitive vulnerabilities and human cyber behavior. ReSCIND consisted of five performer teams: CIRCE, CASPAR, PSYCCDEF, GAMBIT, and ASCEND.
CIRCE
PRIME
SUBS
Loss Aversion—Endowment Effect
Cognitive and cyber skills surveys, plus two 1-hour cyber tasks
34 participants on cyber task with threatened loss of resources
Surveys and skills data; cyber activity, Splunk logs and session data
1-hour series of funfair-themed games to measure cognitive effects such as risky choice and near miss effect that align with cyber tasks in “Gold” exercise
Naturalistic Cyber Attack Behavior with Host/Network Manipulations (EXP1)
2-day cyber exercise attacking a single network with multiple key objectives and intermittent intelligence about the task
Host/Network Manipulations include realistic but potentially erroneous information such as mislabeled host names, applications, accounts, or files containing outdated credentials
19 red team participants who passed a cyber screening questionnaire
Screening and demographics, self-reports and psychometric data, PCAP, Suricata, keylog, and terminal histories
Naturalistic Cyber Attack Behavior with Host/Network Manipulations (EXP2)
2-day cyber exercise attacking a single network with multiple key objectives and intermittent intelligence about the task
Host/Network Manipulations include realistic but potentially erroneous information such as mislabeled host names, applications, accounts, or files containing outdated credentials
20 red team participants who passed a cyber screening questionnaire
Screening and demographics, self-reports and psychometric data, PCAP, Suricata, keylog, and terminal histories
Naturalistic Cyber Attack Behavior with Host/Network Network Manipulations (EXP3)
2-day cyber exercise attacking a single network with multiple key objectives and intermittent intelligence about the task
Host/Network Manipulations include realistic but potentially erroneous information such as mislabeled host names, applications, accounts, or files containing outdated credentials
22 red team participants who passed a cyber screening questionnaire
Screening and demographics, self-reports and psychometric data, PCAP, Suricata, keylog, and terminal histories
Online survey-based measaures of cognitive bias across geographically distributed populations
Big Five, Portrait Values Questionnaire; C-Test (English language proficiency); CRT-3, GRiPS, ADMC, overclaiming, Assessment of Biases in Cognition (ABC)
Over 2,100 crowdsourced participants from 78 countries
PsyberQuest CTF: Virtual Capture The Flag (CTF) Event
Multi-day CTF event with four challenges designed to simulate real-world cyber-attack scenarios with multi-step attack paths against small network topologies
Collected in four batches with cyber log data from 26 participants
Dataset includes: cyber log data, self-report data, flag submission data, session data, and psychometric data