George Weldon: The Consistency of Aggression, Sexuality, and Pop Culture in a Long Dream Series

Because of the massive amount of data presented to us in this study, there was no way we could have coded George's dreams using every single Hall/Van de Castle category. The only way to make a study feasible was to select a few elements and simply count how many dreams contained them. We chose aggression because it is easy to code and quite common. We picked sexuality because it is very easy to code, and moreover, we might expect fluctuations in the sexual content of people's dreams as they mature, meet new lovers, break up, etc.

The third category we examined was "pop culture." We invented this category specifically for this study because we noticed that George's dreams contained many references to pop music, television, movies, comic books, etc. In fact, in many of his dreams, George was interacting with famous actors or fictional characters, or he was actually in a TV show, movie, or comic book. Because these pop culture references were so frequent in his dreams, we thought it would make a good topic of study. Unfortunately, there are no norms to which we could compare our findings on George. But, as long as we carefully operationalized our definition of the new pop culture category, it was a perfectly valid measure with which to measure consistency.

Aside from the raw numbers, there are two larger reasons we are publicizing the results of this study:

• First, it shows that you can plow through incredible amounts of dream data (such as a dream series that spans several years) by using the "at least one" method of coding. That is, all you have to do is count how many dream reports contain "at least one" of some coding element, and you can do an excellent study with much less tedious busywork than it would take to do a complete content analysis.
• Second, we have demonstrated that you can create new categories if need be. If you do this, you must clearly define the coding rules and then make sure that you have good intercoder reliability. The drawback, of course, is that there will be no norms to compare your new measure to.

Now let's get back to George Weldon. The graph below shows the results of breaking up all the dreams into twelve sets of 200 dreams each. (The dates along the horizontal axis are the mean dates of each subset.) There are a few interesting things to point out:

• For the most part, the percentages of dreams containing our three categories did not vary too much (although the amount of aggression fluctuated somewhat during George's college years). In other words, there was quite good consistency over the 15 years.
• With the exception of a "blip" in 1984, the order of the three content categories remained the same; there were almost always more dream reports with aggression than pop culture, and always more with pop culture than sex.
• The findings on aggression are a little below the male norms (47%); the findings on sexuality are very close to the male norms (12%).

George Weldon: Dreams with Aggression, Pop Culture, and Sexuality by Subsets of 200, 1981-1996

Go back