For example, some people believe they are excellent human lie detectors. But, research shows that the average person is only able to detect deception about 54% of the time. What techniques could be used to influence perceived expertise?
Expertise – defined by Hovland, Janis and Kelley (1953) as "the extent to which a communicator is percieved to be the source of valid assertions" – is the second dimension of source credibility besides trustworthiness and attractiveness in their source-credibility model. The perceived expertise is, of course, not independent of these other dimensions: attractive people are considered more knowledgeable, and trustworthiness has a similar effect.
Leaving aside obvious objective cues for expertise (e.g. proven success with identical problems, the university degree of a medical doctor, or years of practical experience in a relevant field) I want to focus on irrelevant and subjective cues.
What constitutes expertise would most certainly depend on the field and the expertise of the person judging the expertise of another. For example, Golbeck and Fleischmann (2010) found that in online Q&A sites such as cogsci.stackexchange.com, photo cues such as smiles or showing the person egaged in a relevant activity only improve trust in a poster's knowledgeability among non-experts. And while the expertise of a person producing tangible goods can be easily discerned even by a layman user ("Do these shoes fit?"), intangible services such as child care or a physician's diagnosis and the relevant expertise are very difficult to measure, sometimes even for experts. (Zeithaml, 1981)
In a study on the social communication of political expertise, Huckfeldt (2001) found the most important factors to influence perceived political knowledge to be interest (influencing perceived expertise with a probability of .74), objectively defined knowledge (.36), agreement of discussion partner (.31), agreement between discussion partners (.29), partisan extremity (.14) and education (.11).
As you can see, apart from objective knowledge, perceived expertise is in part influenced by wether other people agree with a speaker, how strongly he favors one standpoint (neutrality and scepticism, both signs of healthy scientific reasoning, seem to substract from apparent expertise), and how educated he appears in general (i.e. a well-groomed person with a PhD will appear more of an expert, no matter if he indeed has any knowledge of the field he expands about or not).
Following a dual processing model such as the elaboration likelihood model, source factors such as expertise, likability, attractiveness and consensus serve as heuristic cues when recipients lack the ability or motivation to scrutinize a message effortfully. But not all recipients use the same cues. Low self-monitors (whose attitudes are based on values) elaborate an expert source's message more effortfully than the same message delivered by an attractive source, while high self-monitors (whose attitudes are based on sochial adjustment concerns) process messages from an attractive source more effortfully. (Ziegler, von Schwichow & Diehl, 2005)
In other words: When a message is unambiguous, strong arguments lead to more agreement than weak arguments (not source characteristics such as expertise). But when the message is ambiguous, it leads to more favorable reactions (aggreement, attitude change) when it is delivered by a source whose characteristics match the personality of the message's recipient.
I am sure there is more, but this is what I can come up with at the moment. If you want to find more, look at research on persuasion.
A note on lie detectors:
In a study on "Police officers' ability to detect suspects' lies" (Mann, Vrij & Bull, 2004), the ability to detect lies strongly correlated with (self-)percieved experience in interviewing suspects. Participants where UK police officers of differing ages (22 to 52), length of service (1 to 30 years) and work areas (e.g. criminal investigation, trainers, traffic officers etc.) Average self-reported experience in interviewing suspects was 3.75 (on a scale from 1 = totally inexperienced to 5 = highly experienced).
For the whole sample accuracy in detecting lies was 66%, which is significantly higher than the level of chance. Experience was somewhat correlated (around .20), while length of service was uncorrelated (-.07) to accuracy.
It seems like your question is tapping into a number of different questions. First, you might ask about what causes people to perceive expertise in others verus what causes people to perceive expertise in themselves. Second, there is the question of when are expertise perceptions inflated and if so what individuals are prone to this and what tasks are more likely to induce this.
What makes someone be perceived as an expert?
I imagine the question can be answered somewhat by casual observation. Having appropriate qualifications, having a known track record of behaviour indicative of expertise, dressing and behaving like other experts, presenting ideas and acting with confidence, an so on are presumably related to perceptions of expertise. Of course, some people exploit these signs to demonstrate expertise that they don't possess (e.g., using technical jargon to intimidate; exaggerating stories of past successes).
What tasks lead to inflated sense of expertise?
You use the example of detecting deception. This suggests you are interested in why people would think that they are "experts" at detecting deception, even when they are actually relatively poor at the task.
There are many similar tasks which are renowned for involving inflated perceptions of expertise. These include driving a car, speculating on the stock market, lie detection, evaluating employees from job interviews, and so on. For many of these tasks, there is a literature on why people tend to have an inflated opinion of their own expertise (e.g., see this previous question on self-perceptions of driving ability.
As for causes of inflated expertise, you might look at as interaction between task and person features.
Task features include:
There are wide range of psychological processes that can lead people to have inflated perceptions of their own expertise (for instance see this list). Take two examples, first self-serving bias reflects one general bias. People may wish to see themselves as experts and thus they may rationalise their experience so as to justify inflated self perception. Second, people often don't know what they don't know.
A lot more could be said about all this, but I imagine more interesting answers could be provided by a more focused question. If interest lies in a particular domain of inflated expertise such as lie detection, then there is a dedicated literature on that. Or if interest is focused on the most effective way to demonstrate expertise in some context such as leadership or advertising or ongoing relationships, there will also be a specific literature on that.
Generally, the more you know, the less you call yourself an expert. You could teach me the capitals of each state, for example. Say I knew 13 capitals and you asked if I knew a lot about state capitals. "Sure! Yeah I'm an expert."
"Well, how about Oregon?"
"Oh, I don't know that state's capital."
"Oh, I don't know that one either."
"I thought you said you were an expert!"
Now ask the guy who memorozed 47 state capitals. "Well... yeah, I guess I know a few. There's still Florida that I don't know... or New York... There's a lot of stuff I don't know."
This post does not cite any references or sources. Please help improve this article by adding citations to reliable sources. Unsourced material may be challenged and removed.