The ultimate goal of science is explaining cause and effect relationships. Only true experiments allow us to determine cause and effect relationships. However, even when using true experiments we can only suggest cause and effect relationships in terms of probability. Scientific information is not absolute and it is tentative in nature (subject to change).
True Experiment: A research method, using random assignment of participants to different groups, that allows researchers to determine cause and effect relationships through manipulation of a variable / variables and control of a situation.
So, in addition to describing and predicting phenomena in the observable universe only true experiments allow explanations of cause and effect relationships.
For Causation: Key implications
- Cause occurs before effect
- Cause present - effect more likely to occur
- Cause absent - effect less likely to occur
Confound: an uncontrolled extraneous variable or flaw in an experiment.
Confound is term often used when discussing research. There are a couple of key ways to reduce confounds when conducting true experiments:
- Randomly assign participants to groups so individual differences within groups average out
- Treat each group exactly the same except for the differences in the level of the independent variable
UNDERSTANDING RESEARCH METHODOLOGY
Thursday, November 15, 2012
Subscribe to: Posts (Atom)