Woltjen Lab

Stem Cells and Genome Engineering

Written by: on May 30, 2012 @ 00:30

Here’s an interesting article from “the Node” on the dangers of shooping in scientific research.

When it comes to public presentation of data, we are often tempted to show only the most beautiful images, or make “flawed” images more attractive. We recently had a related discussion in the lab around the practice of cropping Western blot images to remove secondary bands – which may eliminate relevant, or even critical, data.

Really, this problem probably begins before the data becomes digital. How many times have you scanned the plate trying to find the perfect colony, disregarding all misshapen brethren? Cropping by selective image capture with the microscope field, or even tweaking the exposure time on the gel doc has a similar effect in biasing data.

By selective capture or beautifying data after acquisition, we show only what what we want others to see. The danger lies in fooling ourselves by obscuring what the data truly wants to show us.

Categories: Data, Opinion