Research Strategies – Mod 2 Notes
Research in psychology is necessary in order to create logical, scientific, informed conclusions about psychological phenomena. Intuition and “common sense” are unreliable because they tend to vary from person to person. Often, the real reasons behind phenomena are counter- intuitive.
Bias is any unjustifiable (non-scientific) influence that will increase the likelihood of a particular outcome in analysis and research. Examples of bias include: Increased or decreased attention to certain results because of prior expectations. Behaving as expected rather than normally during observation. Not investigating phenomena due to internal value judgments.
These are good for defining questions about observed phenomena, including figuring out what variables pertain to a particular phenomenon. Descriptive studies include: Case studies Interviews Surveys Direct observation: natural or lab Longitudinal and Cross-Sectional Studies
Correlation studies look at the statistical relationship between observed phenomena and make predictions about outcomes on the basis of statistical probability. These often arise out of descriptive studies. Correlation is not the same thing as cause-and-effect.
Experimental studies are the most important method for trying to determine cause-and-effect relationships. Critical to experimental studies is establishing research protocols, including operational definitions and a hypothesis, and the control of variables.
Variables are any factors that might influence an outcome in an experiment. These include: Independent variable: the variable that is actively manipulated to cause a change in the dependent variable. Dependent variable: the research variable influenced by the independent variable. This is usually the phenomenon under question. Confounding variables: other variables that could influence the dependent variable that are not the independent variable.
Experimental groups include: Control group: the group that is not exposed to the independent variable. (Often, a placebo is used to help achieve this.) Experimental group: the group that is exposed to the independent variable. Controlling for confounding variables: random assignment and the double-blind method help to control for confounding variables.
Statistical considerations include: Size of the research population: (=n) Variation between results in the groups. Variation of results within groups. Reliability of the result: (? +/-.95)
For Human Research, there must be… Informed Consent Protection from harm and discomfort Confidentiality Debriefing...unless the test subject waives these rights.