Search
  • Antoni C. Maroto

Biased Perceptions

Last year I enrolled in Salt Lake Community College (SLCC) because I considered studying for an associate's degree in Radiologic Technology. Although I eventually switched to a master's program in Medical Translation at a different institution, I decided to complete the first semester at SLCC. That included taking a course about Statistics.


Since I have a background in humanities, I thought I could benefit from studying Statistics. Math subjects were not my forte in high school, so I took the challenge. Not only did I ace the course, but I also gained some invaluable knowledge and skills.


One of the things that I learned is that predictions based on statistics may not necessarily be accurate. That is because they could be flawed. A common flaw in statistics is sampling bias, which can happen unintentionally.


The word bias can be defined as a preference for selecting some individuals over others. It could also mean that certain results are more likely to occur in the sample than in the population.


In Statistics, a sample cannot contain bias. Otherwise, the sample will not be representative of its population.


That is why we should be careful when we make certain statements based on data collection and analysis. Here is an example:


A translation agency approaches you to work for them, but they want you to take a test before you start collaborating.


They claim they have had a few bad experiences with some translators who misrepresented themselves.


Therefore, they began testing translators to see if they were suitable matches.


To their surprise, a high percentage of candidates failed their tests, which solidified their belief that translators must be tested before working with them.


Where is the problem here?


There could be sampling bias because some of us choose not to take translation tests, especially if we will not be compensated for them.


We prefer to get verified by other means, preferably those who will not make us spend too much time filling out countless forms because we are generally busy people!


Better yet, we would rather pursue professional relationships based on trust.


READ: Proving Oneself — My Take on Translation Tests


Therefore, I do not believe the samples in cases like the example above are representative of the entire population of translators.


One of the most powerful lessons I learned when I took Statistics is to use critical thinking. Statistics are, unfortunately, easy to manipulate.


As you can see, predictions resulting from data collection and analysis are not necessarily accurate. Statistics are only reliable if they are free of bias.