Procurando dialogar com o texto de Davis (1995), James Brown, da Universidade de Hawaii, Manoa, faz considerações sobre a pesquisa qualitativa que resumimos a seguir. O texto completo encontra-se em http://jalt.org/test/bro_22.htm
How Can We Know If Qualitative Research is Systematic and Principled?
Research can be systematic and principled in many different ways. In general terms, good quantitative research (at one end of the qual-quant continuum) will be judged in terms of its reliability, validity, replicability, and generalizability, while sound qualitative research (at the other end of the continuum) will be judged in term of its dependability, credibility, confirmability, and transferability.
Dependability involves accounting for all the changing conditions in whatever is being studied as well as any changes in the design of the study that were needed to get a better understanding of the context. Dependability can be enhanced by using overlapping methods, stepwise replications, and/or inquiry audits. Overlapping methods use carefully planned METHODOLOGICAL TRIANGULATION, or multiple data gathering procedures (e.g., observations, interviews, and questionnaires), in order to create overlapping (and therefore cross-validating) data. Stepwise replications involve TIME TRIANGULATION, that is, gathering data on multiple occasions (e.g., at the beginning, middle, and end of a school year), which helps in examining the consistency of the data and interpretations over time. Inquiry audits involve enlisting an outside expert “auditor” to verify the consistency of agreement among data, research methods, interpretations, conclusions, etc.
Credibility requires demonstrating, in one or more ways, that the research was designed to maximize the accuracy of identifying and describing whatever is being studied, especially as judged by the groups of people being studied. Credibility can be enhanced by using one or more of the following strategies: prolonged engagement, persistent observation, triangulation, peer debriefing, negative case analysis, and/or member checking.
Prolonged engagement involves investing sufficient time and persistent observation involves using adequate numbers of observations, meetings, interviews, etc. so that participants feel enough confidence and trust in the researcher to allow for adequate study of the cultural context and adequate checks for misinformation.
Various forms of triangulation may also enhance credibility: SOURCE TRIANGULATION involves gathering data from multiple sources (e.g., people in different roles, like students, teachers, and administrators) in order to minimize and understand any differences/biases held by people in various roles. INVESTIGATOR TRIANGULATION involves using multiple researchers to interpret the data in order to minimize and understand any differences/biases the researchers may have. LOCATION TRIANGULATION involves gathering data at multiple sites (e.g., three different schools) in order to minimize and understand any differences/biases that might be introduced by the participants in each of the institutions.
Credibility can also be enhanced by using peer debriefing (i.e., critical examination and evaluation by a qualified outside researcher of the study design, data collection, analyses, etc.); negative case analysis (i.e., intentionally searching for and analyzing examples of data or participants that contradict the overall interpretations in a study); and member checking (i.e., verifying the researcher’s interpretations and conclusions with the various groups of participants themselves).
Confirmability entails full revelation of the data upon which all interpretations are based, or at least the availability of the data for inspection. In other words, the reader of the research report should be able to examine the data to confirm the results/interpretations.
Transferability involves demonstrating the applicability of the results of the study in one context to other contexts. Transferability can be enhanced by providing what is often referred to as thick description (i.e., giving enough detail so the readers can decide for themselves if the results are transferable to their own contexts). Thick description also “involves an emic perspective, which demands description that includes the actors’ interpretations and other social and/or cultural information”.
Source: How Can We Know If Qualitative Research is Systematic and Principled?
by James D. Brown - http://jalt.org/test/bro_22.htm (Adaptado) – Acesso em 06/05/2012