Data are considered "valid" if they accurately represent social reality. Validity is often associated with qualitative research, which focuses on the meanings of, connections between, and explanations for social phenomena rather than simply enumerating features of social life. It is possible to obtain very informative and valid data through qualitative studies with relatively small samples (Haralambos and Holborn 1991). Ideally, a study should give the researcher insight into the social reality or realities of the respondents as they themselves see it—not in terms of frameworks imposed by the researcher. In this respect the analysis of the letters arguably produces more strongly valid results than that of the interviews, since parents are expressing themselves in their own words, with their own emphases, unprompted by anyone outside the family. In addition, the letters convey the parents' private, most deeply felt concerns to someone who is also the parent of a deaf child, with whom they are likely to be able to communicate most honestly and unreservedly. However, the letters tend to be short, providing a limited amount of information—they reveal key facts about the family's situation with regard to their deaf child and make clear the parents' main preoccupations, but they are silent about many details that a researcher might like to know. By contrast, interviews that systematically cover a broad range of topics produce more comprehensive information, thereby providing a fuller picture of parents' circumstances and their response to them. That fuller picture in itself contributes to validity, since it lessens the possibility that the researcher will draw false conclusions because relevant or possibly relevant factors were not considered; this argument is likely to be especially significant when the researcher is undertaking original research in a different cultural context.