外语教学研究方法—随机化控制实验.pptVIP

Table of Contents How are variables are controlled and manipulated? Random assignment Within and between subjects design Threats to internal validity Two examples of randomized experiment Other experimental designs Evaluation of randomized experiments Introduction Randomized experiments are highly specialized tools which are excellent for some jobs and poor for others. They are ideally suited for the task of causal analysis. It is safe to say that researchers cannot make bold causal statements if they have not conducted randomized experiments. The chief strength of randomized experiments is their internal validity, which is accomplished through the researcher’s assumption of control over the independent variables in the design. This kind of control, which may be easily achieved in laboratory settings that may be quite different from everyday settings in a variety of ways that limit the researcher’s ability to generalize his or her results to other settings, may compromise external validity. How are variables are controlled and manipulated? (1) Variables, as the name suggests, must vary and have at least two values. For example, if you want study the effect of age of starting to learn a foreign language, you need to study at least two age groups which are substantially different from each other. An independent variable is one which the experimenter should manipulate in order to study its relationship with another variable called independent variable, or the result caused by the independent variable. Experimenters study variables that either they or someone else can manipulate, such as the age of starting to learn a foreign language, the materials to be taught, the methods used etc. They may decide when to start to teach a foreign language, what to be taught etc. Subject variables, like religion, aptitude, learning style, education, income etc, are ones that people bring with them to the study, and that are virtually impossible to manipulate. How are variables a