In my experience it's usually enough to just explain that that's not what linguistics is about. Give examples of the difference---i.e., examples of prescriptive "rules" that no one actually follows, and examples of descriptive "rules" that linguists are actually interested in (examples to show people's abstract linguistic knowledge--I find things like pseudowords useful for this, and things like island or subjacency violations, the sorts of things that no one is explicitly taught about in a prescriptive grammar class). It can also be useful to give examples of the descriptive rules that even non-"standard" grammars follow---for example, in a United States context, lots of young students are taught prescriptive lessons about how certain constructions in e.g. AAVE are "wrong", but in a linguistics class it's useful to give examples of how AAVE constructions actually follow a grammar too.

Dear Wong,

You have have rightly broached up a matter of concern facing teachers in linguistic classes. Students enter Linguistic classes with certain preconceived notions about the betterment of certain rules over the others.In my own classes, I provide information helping them to realize that social normative mores do not make a given language any better than a language used by primitive speakers. What makes a language worthy of  analysis are the inner mechanisms defining it as a system of communication. Exteralinguistic subjective judgments lack the objectivity required for linguistic analysis.

Best regards,

R. Biria

Teacher-trainees agree that pupils cannot tolerate cases of ambiguity. On the one hand, we need to honor this as teachers. On the other hand, prescriptive and proscriptive grammars do not seem to be the solution, as they create false assumptions about the body of data (language) being analysed.

http://shanklin.sdsu.edu/Linguistics/CommunicationL.pdf

USUALLY STUDENTS COMING FOR LINGUISTIC CLASSES AWEILL HAVE THE LANGUAGE BARRIER AND THERE WILL BE ADDITIONAL BURDEN FOR THE TEACHER TO TRY TO MAKE HIM OR HER UNDERSTAND THE MEANING AND PATIENCE IS A BIG VIRTUE IN SUCH CASES.

I explain descriptive rules as arising from observed patterns of behavior, whereas prescriptive rules are more a part of social politeness phenomena. This means that prescriptive rules are real, too, just not as linguistic patterns -- instead, they are part of a set of ideological beliefs about how one ought to behave, like which fork to use for which course at a fancy dinner or whether a man should stand up when a woman enters the room. My students usually are in my class because they have been successful learners, including learning prescriptive rules, and I do not want to create confusion or ideological battles in my classroom by denying the reality of prescriptive rules. They are real, but they arise from different sources (for English, many were created by major religious figures in the 18th century), they are learned explicitly in the educational system rather than implicitly via immersion in the linguistic environment, they are usually called for in formal situations (definitely not the vernacular!), and they are markers of (high) social status rather than an integral part of the communicative system.

This would be a good starting point: http://xkcd.com/1735/

If you are to confront your students' misconceptions, better do it through humour.

http://xkcd.com/1735

On the first day of my introduction to linguistics class, I start my lecture by giving my students a task to complete. Here are some examples (the items listed are not exhaustive):

A. Below are some grammar “rules” you may have been taught. After looking them over, indicate the degree to which you follow the rules. Then, discuss your responses with a partner(s).

1. Don’t end a sentence with a preposition. (e.g., A preposition is not a good word to end a sentence with.)

a) Always follow     b) Almost always follow     c) Sometimes follow     d) Never follow 

2. Don’t split an infinitive. (e.g., I wanted to carefully explain to him why the decision was made.)

a) Always follow     b) Almost always follow     c) Sometimes follow     d) Never follow

B. Mark the following sentences as "okay," "bothers me somewhat," or "bothers me a lot," according to how you would react to hearing them SPOKEN. Would your answers be different if the sentences were written rather than spoken? Then, discuss your responses with a partner(s).

1. Who did they speak to?

a) okay     b) bothers me somewhat     c) bothers me a lot 

2. Kathy and me arrived first.

a) okay     b) bothers me somewhat     c) bothers me a lot

To these lists, you could add a number of other prescriptive rules and "violations" of these rules. 

Cheers!

Some reflections by a corpus language planner:

a) If descriptive linguistics of a language has to describe what belongs to this language, a previous distinction from what does not belong to it has to be made: is not this a kind of "prescription" about what it is of this language and what it is not?

b) Some prescriptive rules not belonging to the child competence in the beginning of his/her schooling are finally adopted by him/her: do they, then, belong to his/her current competence?

c) Are linguists (and all language-related professionals) prescription-free when doing their job?

d) As Milroy/Milroy (Authority in language: investigating language prescription and standardisation, 1991, 11) say: “The study of linguistic authoritarianism is an important part of linguistics, and as linguists we feel an obligation to attempt to close the gap between specialist and non-specialist views on the nature and use of language. One reason for that is that attitudes to language have practical consequences […]. But the best reason for studying prescription is simply that it is interesting in itself.”

Best regards,

Joan Costa-Carreras

Costas: I literally just shared that comic with my students yesterday!