Comparisons

Comparisons are frequently made in the results sections of papers, and it is especially important to compare "like with like."

One common error made by non-native authors is overlooking this simple rule and leaving the reader to make an assumption about what is being compared. At best, the language will appear unnatural but the meaning clear; at worst, the wrong meaning can be imparted. As an example, the sentence "Expression levels of p53 in smokers were compared with non-smokers" should actually be "Expression levels of p53 in smokers were compared with those in non-smokers."

Another frequent error with comparisons is the use of relative terms (higher, greater, more, etc.) without a reference. In the sentence "transgenic mice showed higher levels of cortisol" it is unclear what these levels were higher than. Thus, a "than clause," such as "than control mice," is required. The reader might make this assumption automatically, but in some cases alternative inferences will be possible, and the goal of accurate scientific writing has to be the removal of all assumption. Because comparisons of results are critical to their interpretation and, ultimately, their significance, it is critical that you convey to the reader exactly what is being compared.

Finally, the word "between" should be used for comparisons of two findings, but "among" should be used for comparisons of three or more.

Examples:

  • "The levels of ubiquitinated proteins were higher in patients than in control subjects." In this sentence, the "than clause" provides a reference for the term "higher."
  • "The levels of ubiquitinated proteins in patients were higher than those in control subjects." Unlike the first example, where patients and controls are both on the same side of the comparing term, that is, they are both mentioned after "higher," here, patients and controls appear either side of the comparing term. Therefore, it is necessary to add "than those" to compare like with like.
  • "There was no significant difference in the levels of ubiquitinated proteins between patients and controls." The word "between" is appropriate here for a comparison of two groups.
  • "There were no significant differences in the levels of ubiquitinated proteins among AD patients, PD patients and controls." The word "among" is appropriate for comparisons of more than two groups (note the change to the plural differences because more than one type of difference is possible with more than two groups).