Case Based Exam Question 60

[Karen C]

I am confused about the answer to this question which states:
60. A 65-year-old male was admitted to the hospital three weeks ago after experiencing rapid and progressive lower extremity weakness and paresthesia following an upper respiratory infection. The patient reported having to use his arms pull himself around his house on a rolling office chair before he agreed to call emergency personnel. Upon admission to the hospital, the patient was diagnosed with Guillain-Barre syndrome. His manual muscle test grades at admission were grossly 2/5 in the bilateral lower extremities (ankles, knees, hips), 3/5 in bilateral wrists and hands, and 5/5 in the elbows and shoulders. He had grossly diminished light touch and proprioception in the legs and hands.

The patient was in acute care for two weeks prior to discharge to inpatient rehabilitation. He is currently able to ambulate up to 30 feet with a platform walker and bilateral rigid ankle-foot orthoses.

As you were preparing to treat this patient, you noted a lack of well-designed studies related to physical therapy for Guillain-Barre syndrome. Your hospital frequently treats individuals with Guillain-Barre syndrome, so you decide to design a clinical trial to add to the body of evidence. Which of the following is a way to increase the statistical power of your trial?
a. Increase p-value
b. Increase the sample size
c. Experience larger effect sizes
d. All of the above

All of the above is supposedly the answer, however, I also know that if you increase the P value, you are saying that the result is more likely due to chance. A SMALLER P value means that there is a lower chance that the results were random.
Please help me understand.

[Aushem]

Increasing the P value (making it larger) makes it more likely to find a significant finding (aka increasing the power). A smaller P value lowers the power since it is much harder to find a significant finding.

[Karen C]

Thanks Aushem! But doesn’t it also make the finding meaningless since it is more likely to be by chance?

[Chrissy Durrough Lugge]

Hi Karen- yes, realistically, the researcher would not want to use this method for increasing power.

[Author]

Posts