a. State your hypothesis. Evaluate it.
b. What's the independent variable? What levels? Operational definitions?
c. What's the dependent variable?
d. Which group is treatment, which is control?
e. Evaluate your internal and external validity. For internal, consider both the IV and the DV.
2. Now for a harder one: When people are at a party following one conversation, they generally tune out other conversations in the room. But, if someone says their name, then they hear that. This is called the cocktail party phenomenon. I want to understand this phenomenon. How does it work? What governs your ability to pick up on something from across the room that you're not attending to?
Here are a couple of things I've identified as potentially important:
a. Voice characteristics (both the conversation I'm in and the one I overhear).
b. Message content (what's being said).
Design experiments to help me understand what role these things play in the cocktail party phenomenon. You will need at least two experiments, take me through the steps above for each.
3. Now, let's think about confounds. Are there any in the situations below? Why or why not? Fix any you find.
a. John conducts a study of diet and ability to learn. He
tests his good diet people in the morning and his poor diet people in the
b. Mary studies the effect of anxiety on recall. She has some people study in an anxious situation by having them work math problems for five minutes after hearing a list of words. Mary makes sure that the problems can't be solved in the allotted time. She has the no anxiety people sit quietly for five minutes. Mary finds a big impact of anxiety on recall.
c. Jason investigates word recognition. He has participants pronounce a word as soon as it appears on a screen. Some words from the degraded condition are park, fish, tree, dog. Some words from the not degraded condition are justice, poverty, romance. He finds no effect of degrading stimuli on naming time.
Back to Langston's Research Methods Page