They had several experienced computer users answer typical questions such as “What are my current charges” as quickly and as accurately as possible. The subjects answered the questions using either a table like “A” or one like “B” (there also were two other conditions that are not discussed here).

As they had predicted, the bold Field Values took reliably longer than the non-bold Field Values (7.3 seconds vs. 6 seconds), and the bold Field Values elicited reliably poorer accuracy than the non-bold Field Values (86% vs. 92%). Even more surprising, when they analyzed the user’s eye movements, they found that participants spent reliably more time looking at the bold Field Values than the non-bold, and had reliably more eye fixations when looking at the bold. The “bold” formatting seemed to require longer processing time -- either in the eye or in the brain.

They concluded that even subtle text formatting manipulations, such as the mis-use of bold and non-bold words, can have measurable (negative) effects on search time and accuracy.

References

Joseph, K.M., Knott, B.A. and Grier, R.A. (2002), The effects of bold text on visual search of form fields, Proceedings of the Human Factors and Ergonomics Society 46th Annual Meeting, 583-587.