The NIH toolkit has a variety of cognitive measure that are time efficient because the measures are based on item-response theory. However, I am not sure if they can be included in survey research. A proxy variable like education or questions about cognitive difficulties might be a better approach. You should consider that being asked to do cognitive tasks in the midst of responding to a survey may cause a lot of respondents to discontinue participation.
You may wish to look at the Multidimensional Aptitude Battery (MAB) published by Sigma: http://www.sigmaassessmentsystems.com/assessments/multidimensional-aptitude-battery-ii/
It contains subtests that assess crystallized knowledge (Verbal Scale subtests), those that assess fluid ability (Performance Scale subtests and Arithmetic).
Is the survey among adults or aging population ? Fluid intelligence is likely to decline in the declining years. I'm inclined to Multiple Intelligences than to Cattell's Fluid/Crystallized division of cognitive acquisition.