OUR TEAM
LAB 2 WRITEUP
Descriptive Statistics
Experiment 1: Rats
0 mg:
Mean= 10.5 pg/ml
Standard Deviation= 2.23 pg/ml
Standard Error= .995 pg/ml
10 mg:
Mean= 11.1 pg/ml
Standard Deviation= 7.40 pg/ml
Standard Error= 3.31 pg/ml
Experiment 2: Humans
0 mg:
Mean= 3.83 pg/ml
Standard Deviation= 1.52 pg/ml
Standard Error= 0.482 pg/ml
5 mg:
Mean= 8.93 pg/ml
Standard Deviation= 1.59 pg/ml
Standard Error= 0.504 pg/ml
10 mg:
Mean= 6.16 pg/ml
Standard Deviation= 30.1 pg/ml
Standard Error= 9.52 pg/ml
15 mg:
Mean= 658 pg/ml
Standard Deviation= 213 pg/ml
Standard Error= 67.3 pg/ml
Results
Experiment 1: Rats
We used a ttest for the rat experiment since there were only 2 experimental groups. The resulting Pvalue for the ttest was 0.8674. This Pvalue leads us to have only 13% confidence that the LPS had an impact on Inflammotin levels. This means that there is a 87% chance that our results were due to chance. Due to this, we can determine that for the rats, there was not a significant difference between Inflammotin levels in rats who were given 0mg of LPS and 10 mg of LPS.
Experiment 2: Humans
We chose to use the ANOVA test for the human experiment due to the fact that there were more than 2 experimental groups. The Pvalue that was calculated after doing the ANOVA was 1.4x10^16. This means that we are 99.9% certain that the amount of LPS did affect Inflammotin levels. After determining that there was significance with the data, we performed individual ttests for each experimental group. In order to find out if each piece of data had significance, we used the Bonferroni correction method. This means that for each ttest we did for the human study, we wanted a pvalue of 0.0083. For the ttest done with 0mg and 5mg we obtained a pvalue of 8.59E07, which is well below 0.0083, thus it is significant. For the ttest done with 0mg and 10mg we obtained a pvalue of 9.94E06, which is well below 0.0083, thus it is significant. For the ttest done with 0mg and 15mg we obtained a pvalue of 1.39E08, which is well below 0.0083, thus it is significant. For the ttest done with 5mg and 10mg we obtained a pvalue of 3.01E05, which is well below 0.0083, thus it is significant. For the ttest done with 5mg and 15mg we obtained a pvalue of 1.57E08, which is well below 0.0083, thus it is significant. For the ttest done with 10mg and 15mg we obtained a pvalue of 6.48E08, which is well below 0.0083, thus it is significant.
Analysis
Experiment 1: Rats
Due to the fact that our Pvalue was much greater than .05, we fail to reject the null hypothesis. We can conclude there was not a significant difference between the Inflammotin levels for rats who received 0 mg and rats who received 10 mg of LPS.
Experiment 2: Humans
We used the Bonferroni Correction in order to establish that our Pvalue needed to be less than .008. It was indeed less than .008, so we rejected the null hypothesis and concluded that the difference was significant. Humans receiving 0 mg, 5 mg, 10 mg, and 15 mg of LPS had different levels of Inflammotin.
Summary/Discussion
In the rat study, the results were not significant, while in the human study, the results were found to be significant. This means that for rats, we cannot prove that increasing the dose of LPS made any difference on Inflammotin levels, but for humans, the amount of LPS did change the level of Inflammotin. The two tests were linked in the dosages given to both the rats and the humans. Even though these were the same dosages, these tests should have taken into account other factors that would change the outcome of the study. The weight, as well as the age, of the rats is much smaller than the human patients so the tests cannot be compared conclusively. It seems like Kristen should have tested the rat study first and with a smaller dosage, then used that information to move onto the human tests. Even though the test was successful, she could have potentially harmed everyone in her human study if the drug had turned out to be dangerous.
