Correlations: Description or Inference?
暂无分享,去创建一个
This paper was solicited and the review and publication in The Journal managed by the Editor . Research is often conducted in agricultural education with the purpose of explaining and/or predicting phenomena. This research is frequently labeled associational, relational, correlational, a m facto or causal-comparative depending primarily upon which author one studied when learning about research. The purpose of explaining is most often addressed by what is known as the correlation question: “Do two variables vary together?” The second, a prediction question, examines: “How well can 'Y' be predicted from knowledge of 'X'?" As might be expected, the statistic used to answer the correlation question is a correlation and regression analysis is used to address the second question. These types of research are called descriptive by most research writers. No cause-and-effect (causality) relationships can be drawn since the independent variables (causes) are not under the control of the researcher, but are naturally occurring or selfselected by the subjects. The researcher does not “control” which subject gets which level of the independent variable or variables. Since the researcher cannot exercise control, the research is descriptive. Further, the end sought (purpose) of the research is description in contrast to explanation or control of outcomes.
[1] Jon Wagner,et al. Ignorance in Educational Research Or, How Can You Not Know That? , 1993 .