Midterm Chapter 5 Flashcards
what is the difference between internal and external validity?
external validity is the degree to which research findings generalize beyond the specific context of the study being conducted
internal validity is the degree to which an experiment is methodologically sound and confound-free
how are internal and external validity similar and different in experimentational and correlational studies?
internal validity is more prioritized within experimental studies which require tight procedure which limit generalizability.
why does the presence of a confound (but not an extraneous variable) make it impossible to infer causality in an experiment?
a confound has direct impacts on both the IV and the DV and therefore, it is impossible understand causality as the confound is intertwined with both. However, extraneous variables MAY have an impact on the DV but doesn’t necessarily impact the relationship between the IV and DV.
describe and explain the difference in the kids of conclusions that can be drawn from studies using manipulated variables versus subject variables
w/ manipulated variables, conclusions about the causes of behavior can be made with some degree of confidence bc the experimenter has more control of the IV. The independent variable precedes the DV and can be considered the most reasonable explanation for the results.
w/ subject variables, experimenters can vary a factor, but cannot hold all else constant. Meaning there is not guarantee groups will be equal, this makes drawing conclusions less confident.
describe a study in which love is operationalized as a dependent variable, subject independent variable, and manipulated independent variable
DV: Love will be measured by the amount an individual’s pupils grow when looking at the eliciting stimulus.
SIV: In love group will be measured by individuals in long term relationships, the not in love group will be measured by individuals not in relationships
MIV: Love will be manipulated by having the experimental group watch a romantic movie, and the control group will watch an action movie with no romance
For a study with only a pretest and a posttest, list and describe six potential threats to internal validity and explain how inclusion of a control group enables the researcher to evaluate those threats.
- History: when an event outside of the study occurs between pre and post testing that produces large changes unrelated to the treatment program itself
- Maturation: physiological processes occurring within the participant that could account for any changes in their behavior
- Regression to the mean: the tendency for participants who receive extreme score when tested, to have less extreme scores on subsequent retesting.
- testing: when participants are tested repeatedly, improved scores may be attributed to practice or knowledge of testing procedure
- Instrumentation: changes in measurement procedures that may result in differences between the comparison groups that are confused with treatment effects
- selection bias: any bias in selecting and assigning participants to groups that results in systematic differences between the participants in each group
Using a control group allows a comparison point so that researchers can consider these effects if results for both groups go in the same directions. Whereas if there was no control group, results would be attributed to the studied effect
What are the advantages and disadvantages in the use of a posttest-only control group design versus a pretest-posttest control group design?
advantage: you avoid the testing effect
disadvantage: it is hard to know if there is an effect with no pretest
Compare and contrast the confounds of history and maturation. Give an example of each.
History is when an outside variable has an impact on the dependent variable that is independent from the treatment effects. Similarly, maturation also produces effects that are independent from the treatment groups, but maturation is naturally occurring physiological processes rather than events.
Example History: A study is being conducted measuring the impact of a testing curriculum on the grades of elementary schoolers. In the middle of the study, 10 teachers quit their jobs. The students perform poorly on their tests even with the curriculum.
Example maturation: researchers are studying the impacts of a campus wide mindfullness initiative on the adjustment of first year students. It is found that first year students who completed the mindfullness course were more well adjusted than when they first got to school.
Compare and contrast the confounds of testing and instrumentation. Give an example of each.
Testing effect occurs when improvements in results are a function of participants practicing the test or having knowledge of the testing procedure, whereas instrumentation is when differences in results can be attributed to changes in the way something was tested/measured.
Example Testing: researchers are testing the impacts of eating breakfast in the morning on IQ. Participants are tested everyday for 6 days. The first 3 days, participants do not eat breakfast, and they do in the final 3 days. IQ scores increased by the end of the experiment.
Example Instrumentation: Researchers are interested in how athletic activity impacts self acceptance. Participants are instructed to workout 3 days a week and fill out a survey after each workout measuring self-acceptance. In weeks 2-3 participants filled out the Johnson self-acceptance scale, and in the 4th week, participants fill out the Anderson self acceptance scale. researchers saw a dip in self acceptance come the 4th week.
Compare and contrast the confounds of selection effects and mortality/attrition. Give an example of each.
Selection effects and mortality/attrition all have to do with the participant pool. Selection effects are when treatment groups are not equal, and mortality can make groups unequal but it where selection effect was done by initially selecting unequal groups whereas mortality creates unequal groups through participants dropping out of the study.
Example Selection Effects: two sections of psychology course are being offered and a researcher wants to compare a traditional lecture course with a combination of lecture and discussion. School policy allows students to choose to sign up for either course, and researchers have to fully disclose all information regarding the study. This makes random assignment impossible
Example Mortality/Attrition: Researchers are studying the impacts of a big brother program on the grades of inner city children. Participants were randomly selected from two neighborhoods. However, an evil corporation bought a bunch of property in the area and drove up rent. Half of the participants in that neighborhood moved away.
Briefly explain the confound of regression to the mean. Give an example.
When an individual scores extremely they tend to go back to their “normal” or mean
ex.) Maggie has taken the ACT 5 times,, her results were as follows: 22,24, 31, 23, 22
- Explain what is meant by a subject selection confound and give an example of how this problem can interact with other threats to internal validity.
A subject selection confound is when differing treatment or control groups are selected in differing ways. For example, group one is selected through taking volunteers from x, and group 2 was randomly selected from x.
If populations are different then it is possible threats to internal validity have differing effects. One group may be more likely to drop out than another (mortality), or be undergoing differing rates/types of maturation
