- Slides: 16
Learning Outcomes 1. Identify a quasi-experimental design 2. Describe the differences between experimental and quasi-experimental designs 3. Generate ways to control for confounds in an quasiexperimental design
Subject vs. Independent Variables ● Subject variables allow you to make claims of association. ● Independent variables allow you to make tests of causality.
Quasi-Experimental Design Quasi-experimentsare used when: ● the researcher is interested in subject variables that cannot be randomly assigned because of practical or ethical reasons (e. g. , innate characteristics). ● Participants may be self-selected into groups.
Quasi-experiments have: ● Limited internal validity ○ the experimental group may differ on variables other than the independent variable (extraneous variables) ■ e. g. , teacher competency, morning vs. afternoon class time, birth order
Review How are true experiments different from quasi -experiments?
Review: True experiments ● Determine cause-and-effect relationships ● Participants are randomly assigned to conditions ● Control for all extraneous variables
Matching Activity For each scenario, decide whether it is an experimental or quasi-experimental design. to indicate your choice when prompted.
1 A team of researchers visits a small Kansas town hit by a tornado. They want to find out whether the disaster had increased stress levels among residents. They check all the pharmacy records in the town to compare prescriptions for sleeping pills 6 months before and 6 months after the tornado. http: //psc. dss. ucdavis. edu/sommerb/sommerdemo/experiment/quiz/quasi. htm
2 To see if lighting level increases clerical errors, Dr. Johnson gives a proof-reading test to 60 company employees, 30 men and 30 women. Half of each group wears corrective lenses (glasses or contacts). Each employee completes a third of the proof-reading under each of the following conditions: low, medium, and high illumination. The presentation order for each subject is set using a table of random numbers.
3 To assess the effect of sleep on attention, a researcher randomly assigns 10 individuals to sleep for 5 hours at night and 10 individual to sleep for 10 hours at night. The next morning all 20 participants are given a reaction time task.
4 A researcher hypothesizes that oldest siblings will be higher academic performers compared to later-born children because their parents could provide them with more individualized attention. To assess this, she recruited a sample of 357 participants with 200 oldest siblings and 157 younger siblings to complete an online assessment of academic achievement and parental involvement.
Think-Write-Pair-Share Suppose you hypothesize that children whose parents read to them will score higher on a reading test than children whose parents do not read to them. Since many parents feel strongly about reading to their children and are unlikely to give it up, random assignment will not be appropriate for this study. 1. What are the confounds in this study? 2. What might you do to manage the confounds of this quasi-experimental design? https: //psych. hanover. edu/classes/Research. Methods/Readings/Quasi_experimental_designs. pdf
Potential Confounds - The kinds of books being read Amount of time spent reading Language of the books Language of the test Parental enthusiasm for reading Parents’ schedules Age of the child Accessibility of new reading material Gender (of parent or child) Parental education
Think-Write-Pair-Share Suppose you hypothesize that children whose parents read to them will score higher on a reading test than children whose parents do not read to them. Since many parents feel strongly about reading to their children and are unlikely to give it up, random assignment will not be appropriate for this study. 1. What are the confounds in this study? 2. What might you do to manage the confounds of this quasiexperimental design? https: //psych. hanover. edu/classes/Research. Methods/Readings/Quasi_experimental_designs. pdf
Controlling for Confounding Variables ● Matching ○ e. g. , bedtime stories (IV), parental enthusiasm towards reading, or total contact time with parents (potential confounds) ● Statistical differences ○ e. g. , measure the confounding variables to account for potential variation. ● Create an intervention ○ e. g. , assign the same books for all parents to read to their children over a several week period ● Utilize a self-selecting control group