I define a study as a small version of a larger painting I contemplate doing, and I have mixed opinions about the value of studies.

Studies are good for:
1) Working out color and value issues on a small scale and in less time, using smaller amounts of paint before laying out larger and expensive amounts for a larger painting.
2) Working out size issues - what size and shape canvas, where to put things  on the canvas, what should be completely removed from the painting, etc. (The study for the image on this blog used to have a wall behind the figure. Removing the wall made a much better image).
3) Having studies on hand for the many requests artists get to "donate a piece of art" for auctions, charity sales, and so forth. (That topic may be the subject of another blog down the line. Would love to hear from other artists on this subject).

Studies also have some negative aspects:
1) If you like the way the study turned out, it is a temptation to sort of "copy" the study, rather than respond afresh to what is being painted. That could be deadly.
2) Some of the "thrill of the chase" is missing, when the image has already been painted and you are doing it again, only larger.
3) You can't really transpose the image of the study onto a larger canvas exactly; some things read better in a small format, some read better in larger spaces. Colors are an example.

Don't know how many artists habitually do studies before a larger work. I like to do one when I am trying something new, but usually it just seems to be more fun to dive right in and trust to my instincts.