Chapter 10 - One IV Experiment
Identify one way to prevent selection effects. (could be a technique or study design)
Random assignment to condition to ensure they are equivalent before the study begins.
Use a matched-groups design (matched on a key variable) and then randomly assigned to condition.
Use a within-groups design (all participants serve as their own control).
(A pretest/posttest design can only reveal whether there might be differences in groups before manipulation, but it cannot prevent/eliminate)
Which type of research design is particularly vulnerable to multiple threats to internal validity?
One group pretest/posttest design
It is possible that the IV does affect the DV, but the study was not designed/conducted carefully due to not enough ______-group difference on the DV or too much ______-group difference on the DV.
between; within
Use the following factorial notation to answer these questions: 2 x 2 x 3.
(1) How many IVs? (2) How many levels of IV 2? (3) How many main effects to examine? (4) How many interaction effects to examine?
This group reviews a study’s procedures to ensure that participants are treated ethically. What is the minimum number of members required, and what kinds of professional backgrounds must be represented?
Institutional Review Board (IRB); 5 members, scientist, nonscientist, community member.
This controls for order effects in within-groups designs
Counterbalancing
______ threats involve changes in participants that occur naturally over time (e.g., physical, social, or cognitive development).
______ threats involve a specific event or external factor that occurs between the pretest and posttest and affects most participants simultaneously.
maturation; history
A ____ effect is when scores on the DV are clustered at the high end, whereas a ____ effect is when scores on the DV are clustered at the low end.
ceiling; floor
A researcher plans a 2 × 2 study and wants 10 observations in each cell. How many total participants must be recruited for each of the following designs: (a) between-subjects, (b) within-subjects, and (c) mixed?
40, 10, 20
When a researcher asks participants to provide informed consent, they are adhering to which principle of the Belmont Report?
Respect for persons
A developmental psychologist places two toys in front of a 12-month-old infant (one red, one blue) and records which toy the infant reaches for first. What type of measure is the researcher using? Hint - it is a type of within-subjects design.
Concurrent measure design
Imagine a researcher is concerned about an attrition threat to internal validity in a one-group, pretest/posttest design. They are unsure whether the change in the posttest mean is due to the independent variable or because participants with extreme scores dropped out. Provide at least one recommendation for addressing this threat.
(1) Remove dropouts and only look at scores of people who completed both parts of study
(2) Check pretest scores of dropouts to see if they are near the average or are extreme
In a study looking at whether online reading games would make kids better readers, the reading quiz asked children to identify the first letter of their own name. Is this an example of a weak manipulation or insensitive measure?
Insensitive measure (DV is not operationalized with enough sensitivity -- if it's too easy, you won't detect differences between groups)
Is this an example of a crossover interaction or spreading interaction?
Spreading!
Spreading: IV has an influence on one level (but not all levels) of the other IV. There is an effect of # of others on consumption, but only when you’re with friends.
A crossover interaction = it depends. Ex: The effect of drug on # of symptoms depends on gender. The effect reverses.
______ research collects participants' names but separates from them from the data; ______ research does not collect participants' names.
Confidential; anonymous
Researchers want to test how mood affects creativity. Participants are randomly assigned to one of three mood conditions: Positive mood (watching a funny video), Neutral mood (watching a nature video), Negative mood (watching a sad video). Creativity is measured by the number of unique ideas or drawings participants generate in 10 minutes.
Question: Identify two questions you could ask about the construct validity of this experiment. How could you empirically test the construct validity of the independent variable (IV)?
1. How did you measure creativity? Is there evidence for face, content, convergent/discriminant, and criterion validity?
2. How did you manipulate mood? Did you incorporate a manipulation check to ensure that those who watched the funny video were in a more positive mood than neutral and negative?
A/an ____ threat involves the measuring instrument changing over time, whereas a/an ____ threat involves a change in the participant as a result of taking the DV more than once (e.g., practiced, tired).
Instrumentation; Testing
When we increase our sample size, use a within-groups design, ensure we have a strong and successful manipulation, and control situation noise, we are talking about increasing ________.
Power! Our ability to detect an effect if there is one there.
According to the ANOVA output, is there a significant main effect of age group on favorite attitudes?
No. p = .18, which is greater than .05.
During which part of a study are participants fully informed about all aspects of the research, and under what circumstances is this step required?
Debriefing; when deception has been used
A threat to internal validity occurs only if a potential design confound varies with the independent variable _______
systematically (e.g., handwritten getting easy test versus laptop getting difficult test)
A researcher is concerned that their study might be influenced by observer bias (the researcher unconsciously interprets results in a way that supports the intervention) and demand characteristics (participants alter their behavior because they think the researcher expects them to improve). What strategies could the researcher use to control for observer bias and demand characteristics?
1.Double-blind study – neither participants nor researchers who evaluate them know who is in treatment/control group
2. Masked design – observers do not know condition
Identify each factor below as contributing to either not enough variability between levels or too much variability within levels.
a. situation noise
b. ineffective manipulation of IV
c. measurement error
a. situation noise (any external distraction) - too much within-group variability (unsystematic variability)
b. ineffective manipulation of IV - not enough between-group variability (can't detect diffs between groups)
c. measurement error - too much within-group variability (DVs do involve a certain amount of ME, but if there are more sources of random error, could be too much variability within levels to detect diffs)
For which set of figures is there evidence of a significant 3-way interaction effect? Figure A or Figure B? Defend your answer.
Figure A - The two-ways are different as a function of the third IV. There is a two-way on the left, and no two-way on the right.
Figure B is NOT an example of a three-way interaction effect, because the two-ways are the same as a function of the third IV.
The APA outlines 5 general principles for guiding individual aspects of ethical behavior. Three of them are identical to the Belmont Report. What are the 2 additional principles?
Fidelity and responsibility
Integrity