The first experiment was done on humans, and analyzed the effect the LPS dosage had on inflammaton protein. The control group with 0 mg of LPS had an average Inflammation of 3.834 pg/ml and had a standard deviation of 1.52301018 pg/ml. The group with a 5 mg LPS dose had an average inflammatonin of 8.932 pg/ml and a standard deviation of 1.593931547 pg/ml. The group with a 10 mg LPS dose had an average Inflammation of 61.622 pg/ml and a standard deviation of 30.11069386 pg/ml. The group with a 15 mg LPS dose had an average inflammation of 657.941 pg/ml and a standard deviation of 212.9429762 pg/ml. All studies had an endpoint of 10. The standard error that the control group had was 0.481618106 pg/ml and the standard errors the experimental variables: 5 is 0.504045412 pg/ml, 10 is 9.521837451 pg/ml, and 15 is 67.33848166 pg/ml.
Experiment 2
The second experiment was done on rats, and analyzed the effect LPS dosage had on inflammation protein. The control group with 0 mg of LPS had an average inflammation of 10.516 pg/ml and had a standard deviation of 2.22555162 pg/ml. The rats with a 10mg LPS dose had an average inflammation of 11.112 pg/ml and a standard deviation of 7.40288592. Both studies had an endpoint of 5. The standard error that the control group had was 0.99529694 and the standard error the experimental variable is 3.31067123.
Results
Experiment 1
This table shows the amount of inflammatory proteins (Inflammotin) in each of the human subjects after receiving a specific dosage (in grams) of the LPS. The averages of Inflammotin for each of the set amounts of LPS are shown in the graph below. The values calculated using the T-test are also shown below.
Experiment 2
This table shows the amount of inflammatory proteins (Inflammotin) in each of the rat subjects after receiving a specific dosage (in grams) of the LPS. The averages of Inflammotin for each of the set amounts of LPS are shown in the graph below. The values calculated using the T-Test are also shown below.
Analysis
Experiment 1Experiment 1 was performed on humans.
The group with 5 mg LPS dose had an average inflammation of 8.932 pg/ml while the group with 10 mg of LPS had and average inflammation of 61.622 pg/ml. We found that by analyzing the ANOVA test that there was a statistical significance between the groups. Therefore we performed a T-test, which gave us values for the six comparisons, each being less than 0.05. In experiment one we saw that the different dosage amounts of LPS has an effect on inflammation in humans. The table below shows that there was indeed a statistical difference between the inflammation in humans on LPS. The p-value was less than .05 so it is indeed accurate. Because the p-value is less, then it proves that if we were to retest, then the results would be accurate 95% of the time.
Experiment 2
Experiment 2 was Lab performed on Rats.
As seen in the highlighted table there was no statistical difference between the inflammation in rats on LPS and the ones in control. We used ANOVA testing to discover that the p-value was greater than .05. In order to be statistically significant the p-value must be less than .05. This p-value is necessary to prove a significant difference because it shows that if we retested that we could recreate the results 95% of the time (.05 being the 5% margin for error).
Summary/Discussion
The experiment with LPS dose effect on humans proved to be effective while the experiment with LPS dose effect on rats failed to show any results. We used statistical testing to give us information about our experiment. To test whether the experiment with humans had any statistical difference, we used the One-Way ANOVA test. We chose this test since it is best for investigating differences in more than two
groups (we had four groups). After executing the ANOVA test, we knew there was a difference between the groups because we got a p-value of 1.4E-16. In order to know that there was a difference, the p-value had to be less than alpha value 0.05 which correlates to a confidence level of 95% that indeed there is a difference between/among the means (the test can be repeated and there is a 95% chance we would get the same result). We have affirmed that there is a statistical difference between the groups but there's a problem. The ANOVA test tells us that there is a difference overall but we had more than two groups and we didn't know the statistical difference of each individual comparison of groups. To resolve this problem, we did the Post Hoc Test: Bonferroni Correction. We completed t-tests for each comparison (six tests). Since there is error associated with several comparisons, we have to adjust the alpha value (0.05) and divide that value with number of comparisons needed(six) which results in 0.0125 as the new value. After finishing the t-tests, we found that all six comparisons had an outcome of p-values less than 0.0125 which leads to the final conclusion that there was a statistical significance between the different LPS dose amount. Therefore, different dosage amounts of LPS do indeed have an effect on inflammotin in humans. For the rat experiment, we simply just used the t-test since there were only two groups (unpaired samples) to be compared. After completing the t-test, we found the p-value to be 0.8674. This p-value is more than 0.05 so that means the data received does not have any statistical difference between the two groups. Therefore, we can conclude that dosage amounts of LPS do not have an effect on Inflammotin in rats.
BME 100 Fall 2013
