The most likely value: in an empirical distribution the data value with the highest frequency (the hump!), or in the theoretical distribution the value with maximum probability. Sometimes a distribution has more than one mode, for example in university courses the more mathematical courses often have one group of students towards the higher end (those that 'get it'), then a gap and then another group with lower marks. Distributions with more than one mode sometimes happen because there is a mixed group (e.g. expert and novice users), or because of positive feedback effects, for example when students who understand some concepts learn further concepts, whereas those who misunderstand one thing are confused by everything that follows.