Homework #1 Statistical Concepts

BIO/BTEC180 Biostatistics

Name:___Richard Beck
MiraCosta College

Example 1: Dolphin Therapy?
1. Why did the researchers include a comparison group in this study? Why didn’t they just
see how many patients showed substantial improvement when given the dolphin therapy?
T
The comparison group acts as a control. They are given as a reference value to compare
for your null hypothesis. If there was no group to compare against it would be impossible to tell
if they got better because of the dolphins
or some other external influence.
2. Why do you think patients with mild to moderate depression only were allowed to
participate?
It could be dangerous for patients with extreme depression to be that far ourside of thier
surroundings or be weaned off medication.
3. Do you think the age range is important?
The age range is important. For it to be more significant it should be focused on narrower
age range to tell if the
results only impact certain groups. For instance perhaps the
unresponsive people in the sudy were only old or
only young. It would be impossible to well
without a smaller age range.
4. Why the “random assignment” when they got to the island?
They were randomized to prevent bias from influencing the results of the study.
Randomizing helps make sure the results are accurate and only because of the experimental
infleunce.
5. Do the data appear to support the claim that dolphin therapy is effective? A useful first
step is to calculate the proportion who improved in each group. Calculate these
proportions. Did the dolphin therapy group have a higher proportion who showed
substantial improvement than the control group?
Dolphin group= 10 / 15 had improvement (67% Improvement) Control= 3/12 showed
improvement (25%
Improvement)
Yes the dolphin group had higher proportions of improvement 67% vs 25%. The data
appears to support the
idea that the dolphin group has greater improvement.
6. How would you describe the results of the simulation? That is, how would you describe
what the results looked like. Think about the shape, where it’s centered, how spread out
the data is. Draw a picture if it helps.
The simulation showed that the probablility of having a random improvement similar to
the results was not improbable and likely happened as a result of random chance rather
than the result of a significant improvement. The P value of the simulation did not show
significance and did not disprove the null hypothesis. The graph showed a bell curve
shape with the data centered between the points of 4 and 9. It tapered quickly off on both
extremes.

Based on the second re-sampling results.0009) that couples are more “inclined “ to lean to the right when they are kissing. it is possible and likely that the seed group’s higher proportion was due only to random assignment rather than a significant effect. Based on the in-class re-sampling simulation results. The seed yawning group did not exhibit a significant ammount of yawning compared to the control group (p=0. 3. as compared to the control group? Seed group = 10/34 yawned = 29% Control group = 4/16 = 25 Yes. The difference between the two groups is small and the sample size is small.51). Did the “yawn seed” group have a higher proportion of yawners. the seed group had a slightly higher proportion of yawns compared to the control.383 10000 Frequency 2000 1623 1614 1500 1000 560 574 500 0 0 0 1 6 2 117 121 4 6 8 Dolphin Improvers 10 16 1 12 (80 of 124 couples leaning to the right) provides strong evidence that couples in general really do tend to the lean to the right more often than the left? Yes the second re-sampling results show significance to the experimental hypothesis (p=0. . I would not conclude that there is enough evidence to say that couples tend to lean to the right more than the left.505 1. and random assignment alone is responsible for the observed difference in proportions of yawners between the two groups? (This is what The Skeptic would say. 2. Do the data appear to support the claim that yawns are contagious? The data from the small smaple size shows that it has a slightly higher proportion of yawns than the control group.) Yes. The experiement does not have enough pwer. would you conclude that the researchers’ data (8 of 12 couples leaning to the right) provides strong evidence that couples really do tend to the lean to the right more often than the left? Explain the reasoning behind your answer. Example 3: Leaning While Kissing? 1. 4.Example 2: Contagious Yawns? 1. would you conclude that the researchers’ data Histogram of Dolphin Improvers Normal 3000 Mean StDev N 2715 2652 2500 6. 2. Is it possible that yawning is not contagious. Calculate the proportion of yawners for each group. What were the results of the in-class re-sampling simulation? The results in resampling were that the effect of the seed yawning was not significant and did not discprove the null hypothesis.

What’s a clinical trial? What are the phases? 2. This gives the second experiment much more power and less probablity that the results were due to random probability. there was 21 percentage difference favoring the treatment. Example 4: EPO for Neuropathy 1. Why do we get such different results if the proportion (about 2 out of 3) of people leaning to the right is the same? The results of the the two experiments are so different. What two characteristics of a medicinal therapy do companies need to demonstrate to get FDA approval? 3. Explain why the difference in proportions favoring the treatment is basically the same as in the EPO trial for neuropathy and yet we come to very different conclusions. in the first trial. in the CABG trial. what can we conclude about EPO for the treatment of peripheral neuropathy? Example 5: EPO for tissue protection following CABG 1.3. even though they have the same proprotion. For example. there was a 23 percentage difference favoring the treatment and. is because the second sample size was much bigger than the first. . With respect to the question above.