The single most effective thing you can do is to change the white balance that the camera uses to process the JPEG image so that it favors no primary over another. Another way to say this is that the white balance coefficients should all be nearly one. This situation is called “UniWB”, which is textual and verbal shorthand for “Unity White Balance”. Most of this paper is about ways to do that.
Since camera sensors are usually more sensitive to green than to red or blue, the JPEG image of a camera adjusted for UniWB will appear unnaturally green. You will no longer be able to judge colors while chimping. This is a downside of the UniWB approach, and I don’t know of a practical workaround, although a magenta filter over the lens can ameliorate the situation with some side effects (more on that here).
Once you’ve set up your camera for UniWB, you can make a few tweaks to get the in-camera histogram even closer to the true raw histogram. Usually, you’ll want to reduce the contrast a bit.
How can I tell if the in-camera histogram is close to the real raw histogram?
You’re going to need a program that allows you to look at the values in the raw files that your camera makes and/or the EXIF data in those files. I currently favor a program called Rawdigger, which allows you to do both, and is available here. There are several other programs for this; one of the oldest is DCRaw. The reason I don’t generally recommend it is that is it a command-line program – it has no graphical user interface, although there are programs that provide a GUI for DCRaw. If you’re comfortable with the interface, go for it; DCRaw is very powerful and available free.
If you go with Rawdigger, you can use it both as part of a procedure to optimize your in-camera histogram, and as a way to see how close it is to the raw histogram.
Next: Shortcuts to UniWB