Intro to Research: How Psychologists Ask and Answer Questions Correlation and Description
Jan 13, 2016
Intro to Research: How Psychologists Ask and Answer Questions
Correlation and Description
What you need to know by the end of the day….
• Differentiate types of research with regard to purpose, strengths, and weaknesses.– Why use each type of research method
• Describe how research design drives the reasonable conclusions that can be drawn.– How can each research method best be used
based on their results
Correlation
POSITIVE CORRELATION• Two variables increase
together or decrease together
• Example: The number of gallons of
gas pumped is positively correlated to the amount spent on gas.
NEGATIVE CORRELATION• Variables move in opposite
directions
• Example: Miles traveled is
negatively correlated to the amount of gas left in your tank.
The measure of the relationship between two items or variables
No relationship (0.00)Perfect negative
correlation (-1.00)
Scatterplots
Perfect positivecorrelation (+1.00)
Scatterplot is a graph comprised of points generated by values of two variables. The slope of points depicts the direction, and
the amount of scatter the strength of relationship.
Correlation Coefficient
Correlation coefficient
Indicates directionof relationship
(positive or negative)
Indicates strengthof relationship(0.00 to 1.00)
r = 0.37+
Statistical measure that shows the degree of relationship between two
variables
Number will always fall between -1.0 and +1.0
The higher the number the stronger the relationship
Correlation Coefficient Examples
• - 0.78moderately strong negative
• + 0.05very weak positive
• - 0.43moderately weak negative
• + 0.92very strong positive
What kind of relationships do the following correlations have?
Another example….Scatterplot showing relationship between height
and temperament in people. What kind of relationship is depicted?
Moderately positive… +.63
or
Correlation and Causation
Knowing that two items are correlated does NOT necessarily tell us why or exactly how they are related!!
CORRELATIONCAUSATION
3 types of Research
1. Descriptive (Observational)- simply describing a behavior
2. Correlation- finding the relationship between two variables
3. Experiment- controlled setting, manipulated variables
Naturalistic ObservationObservation of human or animal behavior in the environment in which it typically occurs
Jane Goodall’s study of chimpanzee culture
Naturalistic Observation• Valuable where other methods are
likely to be disruptive or misleading• Problems with Naturalistic
Observation– If people know they are being observed,
they tend to act differently than they normally would
• Ethically you usually must tell people you are observing them, yet can combat by observing for a long period of time
– Observations can be distorted if observers expect to see certain behaviors
• Why would this be a problem? How could it be solved?
Case StudyIntensive examination of the behavior and
mental processes associated with a specific person or situation
http://behavioralhealth.typepad.com
Case study• Are useful when something is new,
complex or fairly rare– Often used in clinical work and
neuropsychology
• Limitations…– May contain evidence that a certain research
thought to be important (why would this be a problem?)
– Are unlikely to be representative of people in general
• Yet…provide valuable material for further research and serve as testing ground for new treatment, training programs & other applications of research
SurveyA technique for ascertaining the self-
reported attitudes, opinions or behaviors of people usually by questioning a
representative, random sample of people.
http://ww
w.lynnefeatherstone.org
Surveys
• Validity of data depends upon:– How questions are worded– Representativeness of people surveyed
(why?)
• Other limitations…– Willingness of people to honestly
complete the survey– People may say what they believe they
should say about an issue
• Still a great way to gather LARGE amounts of information
Quasi-experiments
Studies that have the same “control” as experiments yet do not include the random
assignment of participants
EXAMPLEResearchers want to test hypothesis
that a pregnant woman’s use of drugs will cause abnormalities in her
developing baby.
Could you ethically randomly assign women who are 8 weeks pregnant to a group that will be snorting cocaine 3 times a day? Why or why not?
Quasi-experiments
Conclusions are not as firm as those drawn from true experiments, yet they allow
research to be conducted on topics and in settings that would otherwise be
impossible