Prof Asha Kaul and Vaibhavi Kulkarni wrote a paper with the name of Coffee, Tea or...? Gender and politeness in Computer Mediated Communication. They studied computer mediated communication (CMC), specifically work related emails in the Indian context (494 mails written by men and women were analysed). By the way: google images on politeness show up as scenes from previous centuries ...
They notice that "CMC provides a link between oral and written communication, producing a categorical change which affects not only the notion of what the conversation is but the speaking-writing dichotomy as well. CMC is not restricted to formal, organizational emails but is also widely used for informal communication. Gramatical markers are used and exclamation point and question marks added as well as emoticons." " CMC can facilitate emotional content and inequalities of status and expertise are reduced in task oriented CMCs."
Though not all researches share a similar view, one research cited in the paper states that rule of politeness governing face-to-face conversations seems to be less binding when there is no physical presence. And in mails there seems to be a sense of anonymity which could encourage people to be impolite and express their hostility or resentment explicitly. One cited research observes that employees communicating with their supervisors through email might lose their supervisor's sense of status and could end up in miscommunication.
Looking at gender, the study cites a lot of researches which found differences in language, with for instance, women using more 'graphical accents' than men. On the other hand more recent research suggests that men and women use interaction patterns which are organizationally fit rather than gender specific. In this study, the examples of adherence of politeness maxim was higher in women than in men. Mails written by women showed hesitancy to openly contradict or disagree. Mails were more appeasing and tentative rather than firm and clear. But looking at 'clusters' of answers, they use of language was more contingent on the needs of the situation or the organization.
They conclude amongst others with the statement that disparity between the intent of the sender and the message transmitted would need further research. I think that this is somehow the point. (and now completely non-supported by any research): The way people write (on similar tasks) is probably influenced by their gender, cultural background but also ease and experience in writing. For instance as an example of the cultural background: in the paper men for instance used flattery more than women in their emails, with contradicts other literature. A possible explaination can be found in the cultural background of India.
So realizing that messages may be misunderstood, you become more conscious about the way you write. That's actually why I try to write my mails always just slightly more positive than I would in talking. And it is of course an important point for online communication in international communities of practice which demands from the facilitators and participants alike not only the skill to write clearly, in a way which will be easily understood by people reading it with different mental models, but also the ability to read between the lines.
The other interesting thing here is the reduction of inequalities of status, because that would mean that people who feel inhibited to express themselves in a certain group due to this factor might find it easier to do so online (in computer mediated communication), which opens up the opportunity to hear voices which may not be heard in a similar face-to-face group.