Data from 267 of the 1502 children were excluded from the analysis: 28 because of ocular abnormalities other than high refractive error and 238 because scorable IK4 images were not obtained. Of the latter 121 were excluded because the child did not cooperate during the measurement or because the video data quality was too poor to score (e.g., off center, out of focus), 107 due to technical/experimenter error (e.g., the tester failed to initiate recording before measurement), and 10 because the instrument was not functioning on the test day). In addition, the measurement from one child was an extreme outlier (−13 D) and was therefore excluded. Failure to obtain scorable IK4 images (n = 238) varied significantly across age (χ2 7 =107.19, P < 0.001), and occurred in 85 (23.5%) of 361 6-month to <1-year-olds, 75 (28.5%) of 263 1- to <2-year-olds, 34 (21.0%) of 162 2- to <3-year-olds, 19 (11.2%) of 170 3- to <4-year-olds, 15 (6.9%) of 217 4- to <5-year-olds, 3 (2.6%) of 115 5- to <6-year-olds, 5 (3.4%) of 149 6- to <7-year-olds, and 2 (3.1%) of 65 7- to <8-year-olds. After Bonferroni correction for multiple comparisons, the measurement success rate was significantly better in each of the oldest four age groups (ranging from 4 to <8 years) than in any of the youngest three age groups (ranging from 6 months to <3 years), and the success rate was significantly better in the 3- to < 4-year-old age group than in the 6-month to <1-year-old and 1- to <2-year-old groups. The mean age of the final sample of 1235 children was 3.35 years (SD 2.19; range, 0.50–7.97 years).