Hypothesis Information and Courses from MediaLab, Inc.
These are the MediaLab courses that cover Hypothesis and links to relevant pages within the course.
Learn more about laboratory continuing education for medical technologists to earn CE credit for AMT, ASCP, NCA, and state license renewal and recertification. Or get information about laboratory safety and compliance courses that deliver cost-effective OSHA safety training and continuing education to your laboratory's employees.
| Using p values in antibody identification When p values are calculated for antibody identifications, we think of the null hypothesis as meaning, "the relative proportions of one variable (panel cell being antigen-positive) are independent of the second variable (patient's plasma reacting with the cell). In other words, the results could be due to another cause (different antibody, combination of antibodies, or spurious reactions), not the antibody that we have identified as being probable.Therefore, a p value of 0.05 can be interpreted as meaning that the same results produced by another antibody or cause would be expected to occur by chance alone only one in 20 times (5% of the time), given the number of cells tested. By scientific tradition, this is an acceptable level of uncertainty.A p value of 0.05 does not mean that we have identified the correct antibody. | View Page |
| Using p values in medical research Let's briefly review using p values in medical research. A simple example would be a randomized clinical trial to assess whether a new drug decreases levels of low-density lipoprotein (LDL) more than an established drug. Data are collected from subjects treated with new Drug A and established Drug B. Let's suppose that the mean LDL of Drug A is lower than that of Drug B. We want to know whether the difference is due to an effect of Drug A or if the difference is due to chance. There is no way we can ever be certain whether the observed difference reflects a true difference (Drug A is more effective in lowering LDL) or is just a coincidence of random sampling. All we can do is calculate probabilities (the p value) based on a null hypothesis. A null hypothesis states that there is no difference between the drugs. The p value is the probability of observing a difference as large or larger than was observed in the study, if the null hypothesis of no difference were true. | View Page |
| The p value in this case This CaseWith the panel done 2 weeks post-transfusion, 5 panel cells that were Jk(a+) reacted and 5 that were Jk(a-) did not. This yields a p value of 0.004, which is less than the standard of 0.05, and therefore is more than acceptable statistically. In other words, an antibody other than anti-Jka would be expected to produce these panel results only 4 times in 1000 (which is pretty unlikely).Th true p value is much lower because many more cells were tested than in the panel alone.Concluding that the antibody is anti-Jka is further strengthened because the patient's red cells type as Jk(a-).Learning points: The most important things to know about statistical tools such as p values are that they Relate to the probability of getting the observed results if the null hypothesis were true (the panel results were due to another antibody) Do not substitute for technical and clinical judgment. | View Page |