Gackenbach, McDonnall, and Estrada (2018) reported that dreams collected from college students asking for media in the dream could be classified as active or passive media. That is a classic distinction in the communication studies literature between media that allows interaction and one where the consumer is simply a passive viewer. Active media includes video games, social media, and texting while passive media includes television, videos and movies. They found that self selected dreams that included active media were more pleasant than those that included passive media. In fact, nightmares were more associated with passive media in dreams for these college students. The present study is a follow-up on the Gackenbach et al report where a larger sample of dreams was available. David Olds (personal communication, Oct 18, 2015) provided the author with over a million dreams from the online dream social media website DreamCloud, with the permission of the website owner. After duplication’s were removed 992,022 dreams remained which were collected from late in 2006 through late in 2013. The first three years of the sample were fairly small, below one percent of the total dreams. After 2008 about 100 to 300 thousand dreams were collected each year. These were identified as including active media from gaming (i.e., video game, Xbox, PlayStation) and social media (i.e., Facebook, Pinterest, Twitter) or passive media (i.e., television, you tube, movies). The largest number of dreams from these media was from twitter. The 157,356 dreams from 2013 were entered into a preliminary sentiment analysis using NVivo plus. The most recent generation of textual analysis software allows analysis for emotional valence, sentiment, based on advances in artificial intelligence. The categories of dreams examined were active gaming, active social media, and passive media dreams. As is the case in most dream content analysis most the media dreams determined to be codable in terms of sentiment, across category these were coded as very or moderately negative but the vast majority of dreams were coded as neutral. About a quarter to a third of the dreams coded for sentiment were positive in each category with the fewest such classifications with gaming dreams. However, considerable refinement of the sentiment coding is needed as is coding of the other years of dreams collected. Additionally, other elements of media in dreams needs to be coded. While the sheer number of dreams is a strength of this study, they tend to be quite brief. Full analysis will be computed and reported upon for the conference.
Gackenbach, J.I., McDonnall, B. & Estrada, E. (2018, May). Individual differences in dreams and video game play. Paper presented at the Canadian Game Studies Meeting, Regina, Sask.