Immigrant families whose first language was not English and who may have been living in low-income situations were not included. This second exclusion was because of inherent difficulties in communicating effectively with these families in a language that was not their birth language and because of a concern about introducing confounding cultural variables (Klassen & Burnaby, 1993). Given these restrictions, it would be prudent to recognize that those interviewed may or may not be representative of the larger population living in low…income circumstances.

In my qualitative research design, rather than using formalized sampling procedures common in quantitative research, each of the participant families was chosen through purposeful and selective sampling (Merriam, 2001). Since it was necessary to depend on others in this identification process, the question remains whether these cases represent the population studied.

The decision to conduct my research in only one school was made for practical purposes. This may make it difficult to establish widespread replicability or transferability of the findings to other locations. However, I chose qualitative methods to provide a deeper and broader understanding of the phenomena of family literacy, not to make generalizations. The case study method attempts to overcome this limitation by providing thorough descriptions of the cases and research context, in addition to identifying the assumptions and presuppositions that were central to the research. The results and conclusions drawn from within these boundaries must be weighed with this description I have offered in order to determine how reasonable it is to transfer the findings to other settings and draw out meaningful extrapolations.

My sample consisted of eight families. This large number for a qualitative research design was selected as a safeguard in case there was attrition. Since the eight families participated fully in all aspects of the study, the data collection and analysis were time consuming. Due to the iterative data collection design, there is the potential that the necessary delays between data analysis of previous events and the next data collection event may have resulted in information loss by spreading the process over a longer period.