Answer
1. Sample size is the number of subjects used in an experiment. In general, the public will have more trust in results from an experiment that was done using 300 subjects than in one which had a sample size was just 30. A large number of subjects reduces the effect of individual variation on the experimental observations.
2, Control group. A control group is like the experimental (treated) group in all possible ways except the
treatment being studied. At the end of the experiment, data from the control group and the experimental group are compared. Only if the control group is included can the experimenter be sure that the results from the experimental group are due to the treatment and not to some chance influence. The use of appropriate controls strengthens the trust in experimental results.
Double-blind method. This is an experimental design used in experiments on human beings. In this design , two groups --as alike as possible in age sex , SES, occupation etc-- are chosen. The members of one group receive the drug being tested: the other group --the placebo control group--receives an inert chemical substance. The experiment is called double-blind, because neither the persons in the control group, nor the person or persons who supervise the experiment, hand out the substances, or collect and analyse the dats know which group receives the drug, and which receives the placebo. This is done to obviate psychosomatic effects in the subjects, and subconscious biases on the part of the research personnel. Both these sets of effects have been shown to confound research observations, effects, and conclusions. The use of a double-blind design vastly improves trust in research results from drug tests.
Work Step by Step
Statistical testing is done to derive a number that expresses the degree of the researcher's confidence that the results of his experiment are not due to chance. There are many simple, and some very complex statistical tests used in the determination of the experimental error of a set of data. Some of these are the chi-square test, the t-test, the analysis of variance (ANOVA), and Minitab.
All experiments are bedevilled by systematic and random errors. Random errors that represent variation from the true value can be reduced by statistical analysis. Confidence that an experimental result is real or due to the experimental procedures, and not to random error( variation), is usually expressed in terms of confidence. The confidence limits for acceptance of results as real are set by scientific disciplines and editors of scientific publications. For most biomedical sciences, the researcher must show by statistical analysis, that that he justifiably has at least 99.5% confidence that the results he/she reports is not due to random variation. This is necessary for his data to be accepted for publication.
Peer is scrutiny of a researchers work by colleagues in the discipline.. This is done before and after publication. It ensures honesty, objectivity of researcher, and high quality and trustworthiness of published research.
