Table 1.

Sample benefits and risks of longform audio recordings by population

	Benefit	Risk
Individual Participants	More life experiences represented in scientific research Quantitative reports (e.g. daily language counts) could be made available to participants and clinicians	Loss of privacy and confidentiality as participants record their personal information, such as names and financial information Possible legal consequences if illicit activity such as drug use or domestic abuse is recorded, discovered, and reported
Society and Communities at Large	Studies of interventions carried out using the technology may lead to advances in public policy/health recommendations Comprehensive datasets document cultural and linguistic traditions Development of normative standards based on more representative samples	Generalizations of potentially unflattering findings Potential to reveal aspects of family/ community life that conflict with how communities would want to be portrayed Propagation of bias if algorithms are developed based on non-representative samples
Researcher/Research team	Efficient collection of large quantities of data Ability to reuse data for additional research projects	Snapshots of private moments or illegal activity may lead to discomfort or researcher ethical conflict of interest Potential for equipment damage as recordings are made outside of the lab
Scientific Community	Larger, more comprehensive datasets, leading to greater ecological validity and generalizability of findings Documentation of naturalistic behaviors from more diverse communities, leading to robust theories Processing algorithms that permit comparison across datasets and laboratories Greater opportunities for data re-use among researchers with varied interests from different fields	Processing algorithms that permit comparison across datasets and laboratories may not capture relevant behaviors within individual communities Over-reliance on audio could lead to theories that emphasize auditory and verbal information over other modalities Exploring the vast number of measures that can be derived could lead to spurious findings