The paper on e-collars generated some discussion on a FB group that struck me as baffling. I suppose its important to be able to critically analyse scientific papers to some extent, but this takes some practice and guidance. Here are some quick tips: 1. Ask yourself what the research question(s) is/are. Be very specific. How do the conclusions answer those questions? A study is NOT flawed if it fails to answer questions that were not research questions. 2. Make sure you know what the independent variable (variable of interest) is, and what the dependent variables (measured by researchers) are. Are there any potential DVs that are not measured or controlled for? 3. Is the study experimental (IV is manipulated) or non-experimental (observational)? 4. How valid are the results - can they be applied to situations and populations beyond the study (external validity)? How sure can the researchers be that the effects they recorded are in fact due to the dependent variables they measured (internal validity)? Some knowledge of simple statistics may help. How big is the effect size? Is the sample size big or small? How variable are the results? Remember that small sample sizes can still produce meaningful results. 5. Controls for bias - randomisation, blind conditions (experimenter/recorder does not know which treatments or conditions are given to whom, when), null (no treatment) accounts for placebo effect. NB: Some of these are not always relevant to the research question. E.g. If there is no treatment, how are you going to blind experimenters or test a placebo effect? 6. Controls for possible dependent variables - there may be some DVs researchers do not want to or may find difficult to measure. They may be controlled for by techniques such as matching samples, selecting a specific demographic, or randomly allocating conditions.
Posted on: Sun, 07 Sep 2014 23:31:17 +0000
Trending Topics
Recently Viewed Topics
© 2015