I have gone through your work very thoroughly, and have made a very large number of changes, so that it now reads much more persuasively. I have also eliminated all the repetition and the more dubious of the assertions made. You do seem rather preoccupied with the idea that mailed questionnaires are a good idea whenever the sample is small, though in reality that can be, and are used for any size of sample - one of their great benefits is that they can be used wherever the popularion is disparate, and living a long way from the resarcher.
In addition to my corrections, I think the following points need to be incorporated into your essay if it is to achieve a very high mark:
1. The response rate is almost always low - May (1993) says that it seldom rises above 40% and often falls to below 10%.
2. This underlines the fact that those do reply may not be representative of the population, and so limits the validity of the results. Replies may be systematically biased towards one segment of the population. For example, a postal questionnaire focusing on health issues will almost invariably receive a greater response from high income than from low income households. Since there is a positive correlation between income and health, it follows that the overall result will tend to suggest that the population is healthier than it actually is.
3. In principle, postal questionnaires are cheap to administer, though the researcher will have to send a reply-paid envelope. There is also the cost of the covering documentation - explaining the purpose of the exercise and requesting cooperation. The cost mounts whenever reminders have to be sent to those who have failed to reply. In such cases there is usually the cost of having to print and send the questionnaire again, and this cost escalates if this has to be repeated several times.
4. If the identities of those returning the questionnaire are anonymous, as is good practice, then the researcher will be unable to establish who has responded and who has not, and therefore cannot re-submit the questionnaire to the latter group.
5. The questions must be very carefully compiled and must lack any kind of ambiguity. Otherwise it is impossible to know whether they have been correctly understood, since the researcher cannot then clarify the meaning. Moreover, since there is no face-to-face contact, it is impossible to kow whether the questions have been answered honestly. Clear instructions for completing the questionnaire must also be included, while in practice relatively few questions can be asked, since evidence suggests that the response rate to longer questionnaires is almost always low.
6. Nevertheless, postal questionnaires are popular as the researcher is able to select a sample from what might be a geographically disparate population; they are also cheaper and less time-consuming to administer than those conducted through interview; moreover, many researchers don't enjoy meeting respondents face-to-face! At the same time, postal questionnaires make it virtually impossible to use either a stratified or a quota sample, and yet it is precisely these samples that normally give the results their greatest validity.
I hope all this is helpful.