CFS 395 Notes for 2/9 and 2/11
CFS 395 Notes for 2/9 and 2/11 CFS 395 001
Popular in Intro Research Methods/Stats
Popular in Child and Family Studies
This 4 page Class Notes was uploaded by Lorena Roberts on Thursday February 11, 2016. The Class Notes belongs to CFS 395 001 at University of Tennessee - Knoxville taught by Dr. Greer Fox in Spring 2016. Since its upload, it has received 26 views. For similar materials see Intro Research Methods/Stats in Child and Family Studies at University of Tennessee - Knoxville.
Reviews for CFS 395 Notes for 2/9 and 2/11
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 02/11/16
February 9, 2016 Chapter 4: Fundamental Research Issues Validity: is it true? - Types of validity: o Construct: pertains primarily to measurement issues Refers to the adequacy of operational definitions o Internal: nature of the research project itself Refers to ability to draw conclusions about causal relationships among variables o External: Refers to the ability to generalize findings beyond the study sample and setting Variables: - Something that varies: has at least two values Operational Definitions: - The set of procedures used to measure the variables of interest - Construct validity: how adequate is the operational definition- does it reflect the theoretical meaning of the variable? Relationships among Variables: - Positive linear - Negative linear - Curvilinear - No relationship Correlation Coefficient: - Strength of a relationship between two variables - Ranges from -1.0 through 1.0 - Review your introduction to “r” Reduction of uncertainty: - Page 75-76 of textbook - Very important logic - Why is this important to us? - Uncertainty, lack of predictability How do we reduce uncertainty? - Find relationships among variables - These will increase the predictability of the world Nonexperimental Research Designs: - Make observations, take measurements or variables of interest- ask, look, record - Which variable “describes” the relationship? o Measure both independent and dependent variables - Look for a relationship between variables of interest - Do they co-vary? - How? Describe their relationship. Problems! - Do we really know which variable is the driver? The independent variable? Maybe it’s the other way around. - What if some other variable is the “True” driver of the relationship? Maybe the relationship is just happenstance. - If so, we have a SPURIOUS relationship caused by a confounding variable. - The confounding variable accounts for the relationship. Experimental Designs: - Avoid the problems - Manipulate (rather than simply measure) the independent variable - Measure the outcome variables - Control for potential confounding variables through o Randomization o Hold them constant so they don’t vary between values of the independent variables o Independent groups must be “the same” except for their value February 11, 2016 Example of a Spurious Causal Relationship: Experimental Designs: - Avoid the problems of nonexperimental designs - Manipulate (rather than simply measure) the independent variable (2+ values) - Measure the outcome variables - CONTROL for potential confounding variables through: o Randomization o Hold them constant so they don’t vary between values of the independent variable o Independent groups must be “the same” except for their value Manipulate the Independent Variable: - What does manipulate mean here? - In experimental designs, the researcher exerts control over the independent variable Randomization: - Want experimental groups to be the same on everything conceivably related to the DV o *Except for the “experimental condition” - Randomly assign P’s to IV groups o This is the most efficient way to control for random variation - “Without bias” means the experimenter didn’t affect the assignment to experimental groups - Allows assumption that groups are the same except for the exposure to “the film” With random assignment, if the outcome measure of the DV (attitudes re LGTB) differs between the two IV groups then we can be confident that viewing the film caused the difference in attitudes re LGTB because we are assured they were alike in all other respects This is internal validity: the ability to draw conclusions about causality with confidence Experimental designs are superior to other research designs re internal validity Sometimes random assignment isn’t possible; the experimenter can use information about participants to “control” for variables that may affect the outcome variable. So why doesn’t all research use experimental design? - External validity - Artificiality of experiments - Practicality of experiments - Ethical considerations: cannot randomly assign participants to dangerous/non-dangerous independent variables (thanks to Belmont Principle) - Some variables cannot be manipulated - Much research is designed to DESCRIBE behavior rather than infer causality - Some research is diagnostic in design - Multiple methods are the better alternative when possible Units of Analysis: - Individuals - Groups - Organizations - Neighborhoods, cities, states - Social Interactions - Social Artifacts Time Studies: - Cross-Sectional Studies: o Observations of a sample or a cross-section of a population or phenomena that are made only once or at one time point (most social surveys, the US census) - Longitudinal Studies: o Two or more observations of the same phenomenon over time o Trend Studies: A type of longitudinal study that examines change within a population over time Do not pertain to the SAME individuals over time o Cohort Studies: Examine specific subpopulations, or cohorts, as they change over time Cohorts are sets of people that share an experience at the same time point- like high school graduates or telephone poles o Panel Studies: Examine the same set of people each time over time (interview same sample of voters every month during an election campaign) Pre-Quiz for the test on TUESDAY. Operation definitions: a. Describe how a researcher will measure a given construct b. Often require researchers to state in behavioral terms wat they would need to see in order to infer that they’ve observed a given phenomenon c. Are vital to obtaining reliable and valid data d. All of the above Which of the following Is not an example of an operational definition of self-esteem? a. The number of positive self-statements a person makes each day b. The number of times per week that a person makes a statement of self- worth c. Feeling good about yourself d. B and C Nonexperimental research methods do which of the following? a. Test causal relationships between variables b. Only describe characteristics of existing phenomenon c. Can be descriptive, historical, or correlational d. Examine factors that are not related
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'