It seemed like a fairly straightforward research project.
After spending years teaching about the ways in which movies functioned as agents (and evidence) of socialization I settled down, post teaching career, to work on a related research project. Instead of movies I wanted to study books. Specifically books written for and popular with the middle classes in England and the United States. I was particularly interested in examining the ways in which gender, class and ethnicity were treated in books and stories published between 1900 and 1950. Obviously the job of reading all the popular/successful books written during those five decades was beyond my time and means and so I narrowed the scope of the project to financially successful mystery/detective books that were, if not taken seriously by the literary critics of the time, considered acceptable reading for the educated middle class.
I had two reasons for this choice: As a long time fan of the detective/mystery novel I owned/had already read several hundred books that fell into that category and I wanted to focus on the type of entertainment that people create and consume with comparative unselfconsciousness.  The fact that many of the books I would need to read were in the public domain (and thus available for free or at a small cost via the internet) or could be picked up cheaply at used book sales was also a factor in my choice.
As turns out so often to be true things turned out to be more complicated than that.
First, I quickly found that just because I had read a book several decades ago it didn't mean that I didn't have to read it again. Just as the books original audience had read the book without conscious awareness of the issues/attitudes about gender, class and ethnicity I had originally read it not looking for such things. Reading the books consciously now was not the same as reading the books unconsciously then. And, of course, the me of two decades ago would not even have been aware of many of the things I am conscious of now.
Second, I soon realized that I needed to read fiction written at the same time in other countries in order to disentangle change over time from difference in cultures.
Third, I also needed to read non-mystery/detective books written in the same time period in the same countries in order to disentangle genre-related attitudes towards gender, class and ethnicity from those of the wider society. Some of those books would have to be within the category of general fiction and some from other genres.
Fourth, I need to read books in all three categories (books from other countries, general fiction, other subgenres) from decades before and after the time I am focusing on in order to determine whether attitudes changed at different periods of time in different categories.
This is not an exhaustive list of how my research project moved from reading several hundred books and making notes on several hundred books already read to reading (and rereading) well over a thousand books. It does explain why I have no worries about being bored or at loose ends for the foreseeable future.
 Although books were censored and faced publication bans and these were not as restrictive or onerous as those facing American and British filmmakers. Indeed one could make an argument that although the movie going audience was generally unconscious of the presumptions and limitations of anything that could be shown on the screen many of the people actually involved in making those films were very aware of at least some of them.
 Clover makes a similar point about the makers (and audience) of modern horror films in Men, Women, and Chain Saws: Gender in Modern Horror Film.
 For example, it is difficult to determine the general attitude of American society towards women on the basis of reading only certain sub-genres of American science fiction.