With so few combined technology papers, you need to do some of the older papers to practice, including papers in the technology that you are least comfortable with.
In general:
- Look for papers that are "C-like" - requiring a more general technical knowledge
- Past papers relied on candidates knowing certain basic facts in that technology. Physics / chemistry / biology that you knew in school but have forgotten - e.g. chlorine is a halogen, thermoplastics can be softened by heat, moment = force x distance, chemical compounds can have several names. In the current papers, all those facts are provided, either by the client or in the prior art. For past papers, you may have to look these things up to determine genus/species relationships, differences (important for novelty), effects (important for inventive step).
- EM papers: can derive effect (function) from structure and vice-versa. A feature is usually used to distinguish from prior art - comparing drawings is fastest way to see this. Functional claims are often expected, often close to a "result to be achieved". Enablement issues are very rare. Skilled person can use some extrapolation => similar structures provide similar effects, similar effects imply similar structures. The result of a process can usually be predicted, so claiming of product-by-process is very rare. Skilled person can often fill in a lot of gaps - for example, claims are much broader than specific examples in description.
- CH papers: effects are supported by experimental data. Ranges and range values are used as distinguishing features, but just a different value may not be enough (a diiferent effect associated with this value is evidence of novelty). Effect (function) is not derivable from structure. Structure is not derivable from effect (function). Skilled person cannot use extrapolation based on structure. Some very narrow extrapolation based on effects, but teachings are followed very exactly. Enablement is often an issue. The result of a process cannot be predicted, so claiming of product-by-process is frequent where the product cannot be claimed structurally. Skilled person cannot fill in gaps - claims find exact support in description. It was expected to turn the whole letter from the client (in A) into a part of the description (no longer required). Unity was often an issue (has not been tested in combined papers, but a simple case could be covered). Markush claims were expected (not to be expected in the combined papers)
- Avoid past EM papers that are based around finding a claim that covers a lot of embodiments. Claim language was not given - you were expected to provide intermediate generalisations yourself - e.g. screw, nail, glue => generalise to "fixing means".
