Investigating sources in conjunction with each other requires the employment of many of the skills used in analysing a single source, be that primary or secondary, but with a few key differences. 'Cross-referencing' is the most important thing here, not only regarding how the context of the sources differ (i.e. what they say, the facts they present), but concerning the authors (their affinities, prejudices, experiences etc.), the date of the source in relation to the date of the event concerned, and perhaps most importantly, the purpose (tone) and nature of the sources. Obviously, the uniqueness of the sources will require a more in depth analysis, but as a general overview there are certainly some things to avoid. It's poor analysis to disregard a source from a partisan author, for fear of it being "biased" - Similarly, two sources that agree on the significance/course/legacy on an event do not necessarily make them more 'reliable' when used together. The trick is, not only to cross-reference the two sources, but within those sources cross-reference the 'elements' (purpose, audience, nature, date, authorship). It's sounds tricky, but soon becomes second nature.