Effects of Age on Attrition Rates: A Meta-Analysis on Infant fNIRS Studies

Abstract

Imaging the infant brain has always been challenging. Functional Near Infrared Spectroscopy (fNIRS) is rapidly becoming a predominant technique in infant research for its safety and practicality. However, an outstanding setback to infant research is high attrition rates. Understanding the trends and predictors of attrition rate is important to ensure generalizability, external validity, and research integrity. In this meta-analysis, we hypothesized that age is a large contributor to attrition rates in infant fNIRS studies. We examined trends and interactions between age and characteristics of infant fNIRS studies on two types of attrition rates: overall attrition and infant-driven attrition rate. One hundred and eighty-two studies published between August 1998 to April 3, 2020 provided information on 272 experimental groups for the current study. We found that age was a significant positive predictor of overall attrition rate and was driven mainly by infant-driven behavior. Then, we tested the interaction effect between age and varying study parameters. We found that age and study design were negative predictors of infant-drive attrition, and that stimuli type and age were also negative predictors of attrition. The present paper provides a benchmark for attrition, examines the influence of age-related parameters on attrition, and suggests methods that are beneficial for particular age groups.