This paper compares the values used for the Griffiths constant (G=0.5) and the running mean constant (α=0.8) in adaptive comfort algorithms with the values calculated from thermal comfort field surveys in two naturally ventilated junior schools in Southampton, UK. The surveys were conducted outside the heating season in 2011 and 2012 respectively, including both questionnaire surveys and environmental monitoring. A total of 2693 pupil responses were used for this analysis. The data was examined in two steps: first, each survey set; obtained over a 1-day visit to the school; was examined in order to derive the relationship between indoor temperature change and comfort vote with minimum impact of adaptation. Second, the dataset was investigated for the prolonged periods of the surveys, in relation to weather experienced by the pupils in order to estimate their time for adaptation to outdoor temperature changes. The paper gives an insight into the response of pupils to internal and external temperature changes, immediate and over prolonged periods, in comparison to adults.