overview intro: fact and fiction what is science? what science is not –stereotypes –myths...
TRANSCRIPT
Overview
• Intro: fact and fiction• What is science?• What Science is not
– Stereotypes– Myths
• Assumptions of science– Empricism– Determinism– parsimony
• Methods of Science– Objectivity– control– Operational definition
THE PROCESS OF SCIENCE
• Ideas• Problem Formation• Goal or Hypothesis• Design• Conduct/Collect• Analysis• Interpretation• Report• FEEDBACK
SELF-REPORT approaches…….SURVEYS
• What do we want to know• Construction
– Questions– responses
• Who do we want to know about -Administration– sampling
Survey construction and reactivity
• Topic areas• Confidential and Anonymous• Instructions-honest and accurate• Inclusion of “lie-detectors?”
– Reverse ordered questions– Catch questions– etc
Other factors that may affect reactive responses
• Length• Clarity of questions/grammar
– Jargon, acronyms,– “how effective is the SSC in meeting your
campus needs?”– “Does the student newspaper obfuscate the
truth?”
Avoid Leading Questions
• “do you agree that grades should be abolished in public schools?”
• Most doctors feel that a high fiber diet is healthy. How strongly do you agree or disagree?”
Avoid emotionally loaded or “charged” words or biased wording• “should “junkies” be sent to jail or given
treatment?”
• “charged” words?.....Lie, hate, jealous, promiscuous, etc…
• Do you think pornography is vile and disgusting?
Avoid double-barreled questions; 2 questions in one
• Ham and eggs?• Sent to jail and given therapy?
• etc
Don’t assume knowledge
• Should society adopt the medical or the moral model to deal with drug addicts?
Psychological scales
• Sets of questions that collectively measure an attribute, attitude, disposition, etc…
Population vs. sampleSampling: representation vs bias
• Identification of target population?• Approach and demeanor of researcher
– Simple random sampling– Stratified random sampling– Proportionate sampling
Sample Size?
Final comments on Self-report studies
• ISSUES OF– Reactivity vs Control– Knowledge– Accuracy – Motivation
– Interpretation…relationship to behavior???
Correlational studies may reveal interesting relationships between different variables
• E.G.
• Traffic Accidents and time of day?• Age and Aggression• Internet Dating and Personality factors• Drug use and social attitudes• Eye contact and “liking”
The Limits of correlation: 3rd Variables
Let’s consider these ideas:
Does Nicotine cause addiction?
Does Nutra-sweet cause cancer?
Does Social romantic attraction cause/lead to social affiliation?
3rd variables
Does Nicotine cause addiction?
-personality? Stress?
Does Nutra-sweet cause cancer? -Other food substances, genetics, exposure to environmental hazards etc..
Does Social romantic attraction cause/lead to social affiliation?
-liking? Loneliness? Politeness? Intellectual interest? Other motivations…
Directionality is another problem in this respect
Does Nicotine cause addiction?-could addictive personalities lead to tobacco use
Does Nutra-sweet cause cancer? -Could predisposition for cancer lead to higher use of certain substances?
Does Social romantic attraction cause/lead to social affiliation?
Could affiliation lead to attraction?
-
• Correlations are Descriptive
• they do not control for 3rd Variables • or Directionality
• Therefore Scientists refrain from making causal conclusions based on Correlation evidence.
• The Aspartame Controversy• While questions about saccharin may persist, the safety of another artificial
sweetener, aspartame, is clear cut, say FDA officials. FDA calls aspartame, sold under trade names such as NutraSweet and Equal, one of the most thoroughly tested and studied food additives the agency has ever approved. The agency says the more than 100 toxicological and clinical studies it has reviewed confirm that aspartame is safe for the general population.
• This message would not necessarily be apparent to consumers surfing the Internet, especially those who use Web-based search engines to find information about sugar substitutes or artificial sweeteners. Websites with screaming headlines and well-written text attempt to link aspartame consumption to systemic lupus, multiple sclerosis, vision problems, headaches, fatigue, and even Alzheimer's disease. One report distributed nationally over e-mail systems claims that aspartame-sweetened soft drinks delivered to military personnel during the Persian Gulf War may have prompted Gulf War syndrome.
• No way, says FDA, along with many other health organizations such as the American Medical Association. David Hattan, Ph.D., acting director of FDA's division of health effects evaluation, says there is no "credible evidence," to support, for example, a link between aspartame and multiple sclerosis or systemic lupus. Some Internet reports claim that patients suffering from both conditions went into remission after discontinuing aspartame use. "Both of these disorders are subject to spontaneous remissions and exacerbation," says Hattan. "So it is entirely possible that when patients stopped using aspartame they might also coincidentally have had remission of their symptoms."
• It is true, says Hattan, that aspartame ingestion results in the production of methanol, formaldehyde and formate--substances that could be considered toxic at high doses. But the levels formed are modest, and substances such as methanol are found in higher amounts in common food products such as citrus juices and tomatoes.
So why use correlation rather than experimentation?
• Allows study of non-manipulatable variables– Gender, drug use, etc– May identify or discover new relationships– Suggests possible causal relationships
Overview
• The experimental approach• Variables and terminiology• Between vs Within Designs
• Issues in Experimental design• Inferential Stats• Reliability and Validity• Experimental Control