Wednesday, April 18th
Good work this week class. We got caught up today and are right on track with the syllabus. Here is what we covered this week;
Experiments
Factors for causation (Casual order, covary, no intervening)
Control
Attribute variables
Random assignment
Pretests
Double blind procedure
Confounding variables (suppressor, reinforcer, lurking)
Types of experimental designs (One group posttest only, one group pretest-posttest, pretest-posttest control group, Solomon 4 Group design)
Chi square
t-test
In class we discussed the difference between independent samples and paired samples t-test. For posting, I would like you to come up with an independent samples t-test and then a similar paired samples t-test. (In class, we used the example of a training film at a car dealership. An independent samples t-test would give half the participants the training video and compare their sales productivity after the video to a control group that did not receive the video. If the group who watched the video had higher productivity, then the video is successful.A paired samples example would be to measure sales individuals' productivity, show them all the training film, and then measure their productivity again. If their was an increase, the video is successful.
Experiments
Factors for causation (Casual order, covary, no intervening)
Control
Attribute variables
Random assignment
Pretests
Double blind procedure
Confounding variables (suppressor, reinforcer, lurking)
Types of experimental designs (One group posttest only, one group pretest-posttest, pretest-posttest control group, Solomon 4 Group design)
Chi square
t-test
In class we discussed the difference between independent samples and paired samples t-test. For posting, I would like you to come up with an independent samples t-test and then a similar paired samples t-test. (In class, we used the example of a training film at a car dealership. An independent samples t-test would give half the participants the training video and compare their sales productivity after the video to a control group that did not receive the video. If the group who watched the video had higher productivity, then the video is successful.A paired samples example would be to measure sales individuals' productivity, show them all the training film, and then measure their productivity again. If their was an increase, the video is successful.
posted by Dr. Jeffrey Bolt at 5:03 PM
7 Comments:
One thing that you could be testing is the taste of a product. The independent t-test would be to give half the room your product and the other half of the room someone elses product. You would then analyze the taste and see which group was happier. The paired sample would be to have a group of people taste a supply of products without identifying which one was yours. You would then see which product was most liked.
Machelle Montgomery
These types of tests could be useful for new trials of new medicines. For example, maybe a new drug was developed to reduce the symptoms of Parkinson's disease. A researcher could give half of a group the medicine while not giving the other half of the group the medicine (possibly a placebo pill instead). You could then measure any change in symptoms in both groups. This is an independent samples t-test.
A paired samples example would be to have the entire group of Parkinson's patients take the new medicine and measure any decrease in symptoms.
Lauren Birmingham
this tests could also be used for sports testing. the independent t-test would be to give a group of golfers a new invented driver and another group a standard driver. then an analysis would show which group drove the ball further and more straight. with the pared sample you would have the whole group swing a few clubs without them knowing which one was the new driver, and then see which club the group prefered.
Dan Caparso
One thing that could be tested is if people can taste the difference between Pepsi and Coke. The independent t- test would be to give half the room Coke. You would then analyze the taste and ask people if that was Coke or Pepsi. You would then give the whole class another soft drink that is not either Coke or Pepsi and ask them if it was Coke or Pepsi.
A paired sample example would be if you gave everyone Pepsi and see if they could tell the difference.
Richard Toth
One could test the effect of energy drinks on people's athletic performance. With the independent t-test, someone would distribute the energy drink to only half of the group of athletes being tested, then have the whole group perform some kind of athletic activity and see which group felt the best or had more energy during and after the activity was completed. For the paired sample test, one would have the group of people perform an athletic activity and see how they felt or where their energy level was at during and after the activity. Then give all of them the energy drink being tested, then having them complete the physical activity once again. After this, the atheletes would again be asked about their energy level during and after the physical activity. These two test would see if the energy drink really impacted or had a positive affect on their athletic performances.
Ashley Mayle
One thing that could be tested is the difference between Graeters' chocolate ice cream (from Cincinnati) and Handels' chocolate ice cream (Northeast Ohio). The independent t-test would be to give half of the room Graeters and the other half Handels.Then people would examine the ice cream's taste and determine which one was which.Then the researcher would give the participants chocolate ice cream from Katie's Corner and ask if it were Graeters or Handels.
A paired sample example would be if all of the participants were given Graeters' chocolate ice cream and to see if they could tell the difference.
Erin Satzger
An example of an independent t-test would be American sports cars are faster than European sports cars. To test this, simply line the cars up and see which performs the best.
An example of a paired t-test could be: Attitudes toward the war in Iraq are more positive with a republic president in office. To test this, gather a sample group and see their opinion on the war of Iraq with a republic president in office, or right now. Then take that sample group and see what the results are in there is a democratic president in office.
Marc Shupp
Post a Comment
<< Home