'Compare-and-contrast' questions demand the identification of similarities and differences.
One method of tackling such an essay would be to distinguish five or six areas of similarity and contrast, and to devote a section of the essay to each area - a section in which you would assess the degree of similarity and reach a sub-conclusion.
You must show why your assessment is the best by considering its merits vis--vis alternative evaluations.
It might be useful to define and defend the criteria on which your judgement depends.
That is, to explain why they are the best criteria for judging the historical phenomenon at issue.
'What-role-did-X-play-in-Y' questions imply a functionalist approach - that is, they require that you identify the function of some phenomenon, group or institution within some specific system.A good answer will need to harmonise with all of this evidence, or explain why particular items have been dismissed as having no bearing on the problem.It follows from all of this that — that is, answers which fall outside the field of possible solutions or which fail to take account of received evidence — even though there is no 'absolutely right' answer.The evidence almost always permits a variety of solutions, and different approaches generate divergent conclusions.There are, however, limits to the field of possible solutions, since they must fit in with 'the evidence'.We want you to show us that not only have you acquired a knowledge of the topic but also that you fully understand the topic and the issues raised by it.Essays test understanding by asking you to select and re-organise relevant material in order to produce your own answer to the set question.The conclusion would then require a summation of the various 'sub-conclusions'.It needs to be stressed that none of these types of question calls for a narrative approach.History essays are less about finding the correct answer to the set question than they are about demonstrating that you understand the issues which it raises (and the texts which discuss these issues).With most historical problems (certainly the most interesting ones) it is seldom possible to arrive at a definitive answer.