UNDERSTANDING RESEARCH METHODS - Tusculum College 了解研究方法- Tusculum学院.doc

UNDERSTANDING RESEARCH METHODSby Mildred Patten WHY STUDY RESEARCH METHODS? Leaders rely on research to make decisions Must be able to sort/evaluate info Often conduct research in job Lifestyle decisions based on research Need to read/report research for classes Empirical Research Simple observations can be misleading Plan systematic observation (so not misled) Why observe - need for study (purpose/significance) Who observe - population or a sample of one, not biased against any subgroup (subjects) How observe - tests, interviews, surveys, direct observation (measurement in numbers or words) When observe - existing groups or experimental ones (descriptive or cause and effect research design) Process is to describe existing situation (literature), produce new data (empirical data collection), draw conclusions Experimental Versus Non-Experimental Studies Experiments – treat then observe changes in behavior – to establish cause and effect Two groups assigned at random (equal chance to be in either group) Treatment group (experimental group) behavior observed versus control group behavior Quasi-experimental (causal-comparative) has no randomization Randomization essential in TRUE experiments!!! Non-Experimental Research No treatments given Observe and describe Often called descriptive research The type of measurement used does not indicate whether or not research is experimental Cause and effect determined by true experiments, only suggested by quasi Experimental Versus Causal Comparative Studies Experiments establish cause and effect Often unable to experiment due to legal, ethical, physical, financial reasons Alternatives ? Quasi-Experimental See an effect that has occurred Look at past to determine cause (ex post facto research) Use controls such as matching Dangers in Quasi-Experimental Studies (aka Causal-Comparative) Common cause for both the cause and the effect being investigated stress causes smoking and cancer Difficult to establish that experimental an