2. Verifiability: voters have some means to convince themselves rationally speaking, that the election outcome is correct without having to just blindly trust that the technology is functioning correctly, or that the election authorities are honest. An example of a voting system that has the verifiability property might be paper ballots where people are free to go and observe the counting process because then the voter can directly convince themselves with their senses that the outcome is correct. On the other hand, a system that is not verifiable is paperless DRE's*. Because in a paperless DRE, the only count of the votes is happening by black box software in a way that, that people can't observe and where people just have to have faith that the software is correct.
3. Auditability: the system has some way that it can be manually checked (audited) after the election to ensure that the votes have been counted properly. Optical scan voting has thisproperty. Because we have a set of paper ballots that we can spot check or recount to make sure that the totals are correct.
4. Software independence: the idea was coined by Ron Rivest, an American cryptographer who is the R in the RSA cryptosystem. A voting system is software independent if an undetected change or error in the software cannot possibly cause an undetectable change or error in an election outcome. This is a really powerful idea, because it allows us to distinguish between systems where we have no choice but to blindly trust that the software is secure and correct, and systems where we have some means of catching any, any problems or cheating attempts that the software might be making. Examples of systems that provide this property of software independence are optical scan systems and DREs with paper trails. So long as we are catching any problems by auditing those paper ballots or paper trails after the election.
Note:
* Using Direct Recording Voting machines
Source:
Securing Digital Democracy
