Next Diemand-Yauman wanted to see if the same effect would true in a more realistic setting. He took the study to a high school Ohio in the United States) and tested 220 students. He screened classes for those where the same teacher had been teaching at least two classes of the same subject and difficulty level and with the same learning material. The experimenters took all the worksheets and PowerPoint slides and changed the font. (The experimenters did not meet the teachers or the students or visit the class.)
Classes were randomly assigned to either a disfluent or a control category. The disfluent classes used material that was switched to one of these fonts:
Resttonschweler
Monotype Corsiva
Comic Sans Italic
In the control classes, no changes were made to the fonts. Teachers and students didn’t know the hypothesis that was being studied. They didn’t know whether they were in a fluent or disfluent group. The Material was taught the same way it normally was taught. No other changes were made in the classrooms or the instruction. Students in the disfluent condition scored significantly higher on their regular classroom tests. On a survey asking if they liked their course or course material, there were no differences in these preference ratings. There was no difference among the different disfluent fonts.
Excerpt from: 100 Things Every Designer Needs to Know About People (Voices That Matter) by Susan Weinschenk