A number of recent field studies have collected all the data necessary to calculate the predicted response from a heat-balance model. This can then be compared to the actual responses of real people. One such predictive model which is widely used in temperature standards is Fanger's PMV(Fanger 1970). PMV is supposed to predict people's response to the thermal environment.
Its predictions can be compared to predict people's actual thermal responses in real situations. The results have been disappointing, particularly where peopleare in buildings with a lot of temperature variation. As a result researchers are looking at the adaptive approach to see if it has any explanation.
The first result of the Adaptive Principle is that people are comfortable at the mean (over time) of the thermal conditions they experience. People learn how to regulate thermal conditions to suit themselves, or take actions which allow them to be comfortable in the environment they are experiencing. This means that if they can take adequate adaptive actions to make themselves comfortable and if these actions are not restricted by the circumstances, then they will make themselves comfortable.
The image below shows the model on which the heatbalance model is based (picture courtesy of Max Fordham)