This is an easy one! Just the largest value of a variable in any set of data. Clearly only applies to data that can be ordered, i.e. has at least ordinal scaling.
Distinguish between the observed maximum and theoretically possible maximum, e.g. for item scores the there will usually only be a limited number of possible scores and the theoretically possible maximum and minimum are obvious. The maximum possible age in a dataset from an dataset from a service for adolescents might be 18 years.
There may be many observations with the same maximum value in any dataset not just one so these limits can be of limited value, for instance, in a large dataset the minimum and maximum observed values for a particular five level item are likely to be the extreme possible values, e.g. 0 and 4, or 1 and 5 and there may be many people with the minimum value and the maximum. However, the minimum and maximum observed values for the whole score for a scale of say 40 such items may be much more informative and even in quite a big dataset there may be only one person with the minimum observed score and only one (a different person obviously!) with the maximum observed score.
Try also … #
Mainly chapter 5 but also in chapters 7 and 8.