Psych2
Terms
undefined, object
copy deck
- Pseudo-Experiment
- a research design in which you test a claim about a variable by exposing people to the variable of interest. You then note that these people feel, think, or behave as expected.
- Biggest Problem of Pseudo-Experiment
- Some specific people happen to be very good at playing the piano, bowling, taking the SATs.
- Internal Validity
- The extent to which your findings provide information about causality.
- To fix problem with Internal Validity
- since people are different...use the experimental method (with large N, a control group, random assignment)
- A. Threat To Internal Validity: Selection Bias
- Sampling people from an unrepresentative sample
- A. Threat To Internal Validity: Nonresponse Bias
- a type of selstion bias which results from the fact that people who choose to answer survey are systematically different from people who choose not to answer surveys.
- B. Threat To Internal Validity: History
- -changes that are occurring across the board in a very large group of people(a nation or culture) -Ex: my treatment outcome study on chronic stress
- B. Threat To Internal Validity: Maturation
- -Specific developmental changes that are occurring in a particular person or age cohort. -Ex: TV commercials for pain relievers
- C. Threat To Internal Validity: Regression to the Mean
- -tendency for people who receive high or low scores on a particular measure to score closer to the mean on later testing. *see NOTES
- D. Threat To Internal Validity: Repeated Testing
- the tendency for participants to perform better on a test or personality measure the second time they take it. *see notes for solutions
- 3 Reasons for Repeated Testing
- 1. Learning (IQ tests) 2. Practice (athletic performance test) 3. Figuring out what kinds of responses are "socially desirable" (personality tests)
- E. Threat To Internal Validity: Attrition(Experimental Mortality)
- -Dropping out. -Big problem for very long ("longitudinal") studies that continues for several months or years.
- 3 reasons for Attrition
- 1. People get bored. 2. People move. 3. People die.
- Homogeneous Attrition
- when you have an equal number of people dropping out in each group
- Heterogeneous Attrition
- when attrition rates differ between groups
- Ways to avoid the problem of attrition
- 1. Communicate importance of the study with authority and enthusiasm. 2. warn people that the study may not be all that exciting. 3. other significant compensation such as money.
- F. Threats to Internal Validity: Reaction Bias
- Sometimes reffered to as the Hathorne Effect- gets its name from the Hawthorne plant in Chicago. -worker productivity studies -new coke Biggest Problem: when people realize they're being studied, they may behave in ways that they normally wouldn't. *see NOTES
- F. Threats to Internal Validity: Experimenter Bias
- When the experimenters expectations about the study biases his/her own experimental observations. *SEE NOTES
- Confounds
- -is any design problem in which some additional variable varies systematically along with the IV. -Usually occurs when the experimenter didn't think about what other variables might be playing a role in participants behavior.
- Ways to Avoid Confounds
- 1. Conduct true experiments- less likely to contain confounds because you're holding everything constant (but still not a cure-all!) 2. Be CAREFUL: give a lot of thought to "what else might have an effect on the DV." 3. Measure other variables: that might be confounded with the IV-you may be able to control for them statically later on! (Ex. temp&humidity for panic study)
- One-way Designs
- -One-way designs are the simplest possible experimental designs in that there is only one IV. there are two types of one-way designs: 1. two groups design 2. multiple groups design
- Two-groups Design
- Two groups design: there is only one IV, and this IV has only two levels. -There are two ways of presenting the two levels 1. Experimental Group vs. Control Group: experimental group receives the manipulation, and control group does not. 2.Two levels of the manipulation -there is no strict "control group". Rather, both groups receive the manipulation in different doses or qualities.
- Multiple Groups Design
- Solves the major limitation of the two-groups design (the non-linear relationship limitation). -have one IV, but that IV has multiple levels.
- Factorial Design
- -one-way designs are limited in that they only allow us to look at ONE IV at a time. -To look at more than one IV, use factorial designs... Factorial Designs- contain 2 or more IVs that are completely crossed, meaning that every level of every IV appears in combination with every level of every other IV. So basically you create every possible combination of all of the levels of all your IVs.
- Threats to Internal Validity
- A. Selection Bias B. History & Maturation C. Regression to the Mean D. Repeated Testing E. Attrition -homogeneous -heterogeneous F. Reaction Bias G. Experimenter Bias Confounds