Surveys provide the chance to express participants’ ideas and we can get precise answers to our questions. I think the secret of an excellent survey is the structure of the questions and having very clear the goals of the study. One of the best surveys I have reviewed is the following:
Rosales-Dordelly, C.L. and Short, Edmund C. (1985) Curriculum Professors’ Specialized Knowledge. Lanham, MD: University Press of America. This book reports a survey among professors between Canada and the US.
# 1 Validity, Reliability, Trustworthiness
A. Summary by Nelson Dordelly-Rosales: McMillan (2007) summarizes and provides suggestions to monitor the threats that put at risk internal validity in randomized field trial studies or control trials or, on quasi-experiments in which there is “equating” of pre-test differences: unit of randomization and local history (equivalence of the groups), intervention-treatment (fidelity, consistency with theory), differential attrition (mortality-tracking participants), testing (instrumentation variations and procedures), subject effects (selection-maturation interaction), diffusion of intervention (treatment, asking appropriate questions), experimenter effects (checking values, biases, needs) and novelty effects (changes to normal routines). It is the responsibility of researchers “to include design features that will lessen the probability that the threat is plausible” (p.5)
B. Experience: As graduate student, my recent research activity is in the area of interpretation/construction (my theses). But, I recently reviewed an excellent research by Chauncey Monte-Sano (2005) which resembles another study done at the Ministry of Education in Venezuela on comparative studies. The researchers monitored and supervised the threats to validity and trustworthiness through intervention fidelity (pre- and post-test essays, interviews, observations, teacher feedback, assignments, and readings; analysis of students’ progress within each classroom and between both classrooms, assessing any changed observed in the students’ work).
C. Suggestions: I think that in studies in which there are pre-test differences, it is necessary to include design features to monitor all plausible threats. Particularly, I would suggest increasing the number of homogeneous comparative groups.
# 2 Different types of sampling methods
Summary: Cui, Wei Wei (2003)aims to help the readers to understand what is sampling (a technique of selecting a representative part of a population for the purpose of drawing conclusions of the whole population), the different types of sampling (probabilistic, non-probabilistic, simple, systematic, stratified, cluster, purposeful), potential sources of error (sampling error, non-coverage error, non-response error, and measurement error) and how to reduce error in mail surveys and interviews (avoid unrepresentative number, enlarging the sample size; avoid bias of interviewer or survey researcher in favouring the selection of units that have specific characteristics; improving survey return rates, etc).
The value for me as an educator and as a consumer of educational research: I should use appropriate sampling methods and an adequate response rate for a representative sample. However, I should also evaluate different factors that may affect the quality of data from a research study, for example, procedures, questions asked, validity of questionnaire, among others.
# 3 Two concepts regarding data analysis
a. Helberg (1996) warns the researchers on the paradox that statistics can produce dissimilar or contradictory results. The author provides suggestions about how to cope with sources of bias, errors in methodology, and misinterpretation of results. To that end, he explains how to assuring representative sampling and valid statistical assumptions; recommends using methods available for taking measurement error into account in some statistical models and applying more precision and accuracy in interpretation of results.
b. Oliver-Hoyo snd Dee Dee (2006) reviewed data collection through three qualitative methods to study qualitative variables (meaning constructed by individuals): surveys, journal responses and field notes. The authors are persuasive in that relying on more than two methods is invaluable to avoid gross errors when drawing conclusions in surveys.
c. How the articles’ information could be of value to you as an educator and consumer: From these readings, I learned that multiple methods of data collection and analysis (quantitative and qualitative) help us to develop a more complete view of the problem and the solution. So, in the example provided regarding accountability, I think that the right approach is integrating different “assessment strategies” so that educators can take advantage of all the information.