VetZone vet zone dog cat veterinary medicine veterinarian technician tech health CE continuing education preventive diplomate ACVPM DACVPM health Christopher Anthony Lee Kayla Wells MPH DVM biostats epidemiology blue line dot meaning positive predictive value negative sensitivity specificity confirm confirmatory test p-value p value worth felv fiv leukemia feline diagnostic IDEXX test SNAP false idol Irene of Athens Constantine VI Byzantine Empress
VetZone, vet, zone, dog, cat, veterinary, medicine, veterinarian, technician, tech, health, CE, continuing, education, preventive, diplomate, ACVPM, DACVPM, health, Christopher, Anthony, Lee, Kayla, Wells, MPH, DVM, biostats, epidemiology, blue, line, dot, meaning, positive, predictive, value, negative, sensitivity specificity, confirm, confirmatory, test, p-value, p, value, worth, felv, fiv, leukemia, feline, diagnostic, IDEXX, test, SNAP, false, idol, Irene, Athens, Constantine, VI, Byzantine, Empress,

The most powerful woman in the world blinded her own son to uphold her beliefs

Gorgeous, brilliant, and powerful, the Empress Irene perched upon her throne and contemplated her Eastern Roman Empire. She sought to reunite it with the West through marriage with none other than Charlemagne. Luckily for the great Carolingian king, pleuritis killed him shortly before the wedding. For despite her beauty and talent, her obsession with being an iconodule, or supporter of venerating religious images, consumed her. To resolve a dispute over power and to stamp out all iconoclasm, Irene ordered that her own son have his eyes gouged out in the room that she had given birth to him. The severity of his injuries seized not only his life but the dreams of iconoclasts throughout what was now, unquestionably Irene’s empire. Without Irene, the ability to have people, such as Jesus, Mary, or the Apostles, immortalized within the art might not exist due to the rise of iconoclasm in Christianity and other faiths at that time. So, despite her act of filicide, she was canonized, and St. Irene’s feast day is August 9th.

Podcast: Free Audio File

If you prefer to listen to podcasts, feel free to play the audio version of this blog by clicking on the player above.

Podcast: P-value, the False Idol
Length: 5 min 27 seconds
Written and read by the author

VetZone, vet, zone, dog, cat, veterinary, medicine, veterinarian, technician, tech, health, CE, continuing, education, preventive, diplomate, ACVPM, DACVPM, health, Christopher, Anthony, Lee, Kayla, Wells, MPH, DVM, biostats, epidemiology, blue, line, dot, meaning, positive, predictive, value, negative, sensitivity specificity, confirm, confirmatory, test, p-value, p, value, worth, felv, fiv, leukemia, feline, diagnostic, IDEXX, test, SNAP, false, idol, Irene, Athens, Constantine, VI, Byzantine, Empress,

What are p-values and what can and can’t we infer from them?

While no veterinarian is likely contemplating murder, filicide or other, over their view of a p-value, the beliefs of p-value-odules endure. When reading a study, we might be tempted to search for a p-value or series of p-values. If they constitute the correct diminutive size, then there is a temptation to throw blind faith behind the conclusions of the study. Like Irene, following extreme black-and-white choices may take us down roads which we regret. So, what are p-values and what can and can’t we infer from them?

VetZone, vet, zone, dog, cat, veterinary, medicine, veterinarian, technician, tech, health, CE, continuing, education, preventive, diplomate, ACVPM, DACVPM, health, Christopher, Anthony, Lee, Kayla, Wells, MPH, DVM, biostats, epidemiology, blue, line, dot, meaning, positive, predictive, value, negative, sensitivity specificity, confirm, confirmatory, test, p-value, p, value, worth, felv, fiv, leukemia, feline, diagnostic, IDEXX, test, SNAP, false, idol, Irene, Athens, Constantine, VI, Byzantine, Empress,

P-Values are Probability Values

P-values, or Probability Values, are the probability that we could receive a value as extreme or more than the study value assuming the null hypothesis is true. Basically, “What is the possibility that this result occurred from random chance?” The scientific convention sets a standard of five percent.

So, a P-value of 0.05 always means that the result could only have occurred randomly with a 5 % chance, right? Not so fast. The context is important. Imagine a bucket of marbles: 95 white ones and 5 red ones. Red marbles represent random chance. The probability of picking a red marble “on one try” is 5%. If we pick 5 times (keeping the number of balls the same), our chance of picking at least one red marble is now 22.6%. If we pick 20 times, 64.15% and 30 times, 78.5%. These additional chances to grab red marbles are referred to as the problem of multiplicity. If we run enough data, we are almost sure to find a red marble. See the problem?

VetZone vet zone dog cat veterinary medicine veterinarian technician tech health CE continuing education preventive diplomate ACVPM DACVPM health Christopher Anthony Lee Kayla Wells MPH DVM biostats epidemiology blue line dot meaning positive predictive value negative sensitivity specificity confirm confirmatory test p-value p value worth felv fiv leukemia feline diagnostic IDEXX test SNAP false idol Irene of Athens Constantine VI Byzantine Empress

“Facts are stubborn things, but statistics are pliable.”

Multiplicity tosses p-values and statistics upside down on its head. To make matters worse, we might then take those red marbles and generate hypotheses to support our overall general goal. After all we did grab from the bucket containing our original yet defunct hypothesis. This is called HARKing or Hypothesizing After Results are Known. It presents as questionable research practice and a significant problem for science. Like Mark Twain said, “Facts are stubborn things, but statistics are pliable.” Mark Twain also commonly quoted Benjamin Disraeli with, “There are three kinds of lies: lies, damned lies, and statistics."

VetZone vet zone dog cat veterinary medicine veterinarian technician tech health CE continuing education preventive diplomate ACVPM DACVPM health Christopher Anthony Lee Kayla Wells MPH DVM biostats epidemiology blue line dot meaning positive predictive value negative sensitivity specificity confirm confirmatory test p-value p value worth felv fiv leukemia feline diagnostic IDEXX test SNAP false idol Irene of Athens Constantine VI Byzantine Empress

“A finger pointing at the moon is not the moon.”

              So how do we protect ourselves? The key lies with understanding the context of the p-value. Does this refer to prospective or retrospective data? What are the null and alternative hypotheses of the study, and how does this p-value fit in? Could multiplicity or HARKing have raised their ugly heads?

Like icons, P-values may provide valuable insight. The direction and depth of that insight must, in part, be determined by you, the reader. There is a Buddhist saying, that, “A finger pointing at the moon is not the moon.” P-values point us in a direction, but they should never be the goal.

References and Further Reading

  1. Brownworth, L. (2010). Lost to the West: the forgotten Byzantine Empire that rescued Western civilization. New York: Crown .

[amazon_link asins='1118553985,1935660020,0199946647,130526892X,1607951789' template='ProductCarousel' store='vetzone-20' marketplace='US' link_id='aa3e86be-ce1c-11e7-ae16-0944a45c35a3']

One thought on “P-Value, the False Idol (revisited)”

Leave a Reply

Your email address will not be published. Required fields are marked *