This picture was taken in my laboratory, and the experiment here is exploring which isolates of a pathogenic microbe called Mycoplasma gallisepticum make the strongest biofilm (as seen by purple staining). As you can clearly see, there is no stastically significant difference between any of these strains.
Mycoplasma gallisepticum Biofilms (Image: M. May/K. Perez)
Wait, WHAT??, you may be thinking. "Strain B" is MUCH darker than "Strain A"! You cannot tell me that they are not different! The thing is, once the data (the intensity of purple color, in this case) are normalized, it turns out that they are quite similar. More on this in a minute, but here is why I want to bring it up. I saw this article by Phil Plait about climate data last week, and was rather shocked by the tenor of the conversation. While denial of climate change is nothing new, the insinuation that scientists at the National Oceanic and Atmospheric Administration (NOAA) are fabricating their data because they have made completely appropriate normalizations to measurements is egregious. Referring to the process as "altering the data" is slanderous, plain and simple. It impugnes the character of the investigators and breeds distrust between scientists and the public.
But didn't the climate scientists "alter the data?" They changed the numbers from their readings, after all. This is where I want to revisit the purple plates from my lab. Reporting the intensity of purple as read by the spectrophotometer as the amount of biofilm would in fact be misleading and dishonest. Why? Because each well did not have the same number of bacteria in it. If each bacterium produces a certain amount of biofilm, and there are more bacteria in a given well, then it is only rational that there would be more purple color. Instead of reporting the intensity of color as the machine read it, I reported it as purple intensity per cell. In other words, I "normalized" it to make it more truthful and accurate. Let me repeat that: I "adjusted the data" to make it more truthful and accurate. For the record this was rather disappointing, since I had hoped they would be different and this would explain the different disease states seen in the cases that these isolates came from. But as we say in science, it is what it is, not what we'd like it to be. The next time you see a flaming accusation about "adjusting data" being a bad thing, think back to the purple plate, and this graph of the findings, showing no statistically significant differences:
Normalized Biofilm Data (Image: M. May/K. Perez)
The take-home message? Scientific integrity is alive and well!
#Dataadjustment #Climatedata #BadAstronomer #Mycoplasmagallisepticum #Normalization #Scientificintegrity