Method Section -
Undergraduate Psychology students attending classes at San Francisco State University (SFSU) during the 2010 winter and summer semesters were invited to take the online Dreams Questionnaire (DQ) provided through SurveyGizmo.com, for extra-credit or not depending on the discretion of class instructors. Of the 101 participants completing the DQ, age range was 41 years, spanning from 18- through 59-years (M = 25.35; SD = 6.89). The most frequent ages were 21-, 22-, and 23-years, reported by 19, 17, and 10 participants, respectively. Eighty-eight percent of participants were between ages 18- and 30-years.
Thirteen participants were male and 88 were female (no effort was made to obtain equal sex numbers). Average age was 26.85-years (SD = 7.37) for males and 25.13-years (SD = 6.83) for females. Male age ranged 26- while female age ranged 41-years.
Dream Threat Scale. Antii Revonsuo and Katja Valli (2000), the researchers responsible for originating and promoting the TST for the function of dreaming were gracious in their assistance and instruction on use of the DTS upon request from this study’s Principal Investigator (PI). The DTS is a work-in-progress and not every study examining dreams in a TST context will need to use all of its standard elements, and may in fact need to expand on its scope, as was the case in this study.
The initial attempt to identify the number and category of threats and quantity of words describing threats in the dream reports involved three coders. Sentences describing threat were generally agreed upon by the PI and two assistants but the total number of threats per dream was difficult for individual coders to decide upon and for the team to agree on. It was decided that an attempt to quantify the number of threats per dream was too subjective and would be abandoned while more focus would be applied to refining (and ultimately expanding) the DTS. Words extraneous to actual dream content, including preceding material (e.g., “I had a dream in which…”) and waking commentary (e.g., “This dream happened after…”) were omitted from total word counts and threat word counts. Threat target(s) and severity level were only coded for when one or more threats were identified in a dream report.
Themes for dream reports were divided into five main categories (Aggressions, Misfortunes, Failures, Other Threats, and Non-Threats) and 60 sub-categories (as listed in the Results Section). Each dream could be coded as having more than one main category theme, but only one sub-category per main theme. For instance, a dream report could have both a sub-category within both Failures and Misfortunes, but not more than one type of Failure or Misfortune would be coded for per dream report. Therefore, on the rare occasion that a dream contained multiple sub-categories within a main theme category, the most salient, as ultimately judged by the PI, was selected. This approach was maintained to instill relative brevity into what was otherwise a very time-intensive endeavor, and can be construed as resulting in a somewhat conservative frequency estimation of threatening content in the sample dreams. Because the DTS was modified over the course of the study to best accommodate the full variety of dream themes, the final coding was only performed by the PI. Validity in the final results can be judged by the very close parallel of frequency data when compared to previous TST research.
One main theme category was labeled as Non-Threat, and included sub-categories such as Sex/Marriage/Dating and Flying/Floating. All other main theme categories coded for threatening content. A dream report was only threat-free if its content included Non-Threat themes exclusively. Also, Non-Threat themes could potentially refer to threat scenarios, such as if Flying was coded for but the flying was actually an attempt to escape a hostile pursuer, which would concurrently call for coding in the main-category: Aggression, sub-category: Pursue (Pursuits were distinguished as perpetrated by Human, Animal, Monster, Unknown, or As Outlaw).
Participants were requested to use initials for dream characters to protect anonymity; general characters and places were not coded for in this study. Dreams were coded as being “Recurring” or not depending on each participant’s response to a question addressing this in regards to Earliest Remembered Dreams (ERD), or by considering participant commentary in the Most Important and Recent dream reports (with only the ERD results being truly reliable because of the explicit attempt to draw these data from participants. Future studies would remedy this).
Typical Dreams Questionnaire. The same 55 dream theme options used by Nielsen, et al., (2003) were used in the TDQ page of this study’s Dreams Questionnaire. The TDQ was modified in 2003 by omitting themes which were never well endorsed since the TDQ’s inception in 1958 by Griffith and Miyagi, and by adding theme options which have repeatedly been found in dream studies. The only difference in the 55 themes for this study was in their order randomization in the DQ, a feature which was not a part of previous studies and may have resulted in order effects. Questions which were related to the TDQ can be viewed in Appendix A.
Dreams Questionnaire. Potential participants were provided with a link-address to gain access to the Dreams Questionnaire (DQ) designed by this Author. The full questionnaire as it appeared online at SurveyGizmo.com is presented in Appendix A. The DQ was designed to take ~15-min to complete, however, there was no time limit for responding to the DQ, and participants could opt to save their progress and return to the survey at a later time by way of individualized access codes.
After a preliminary Implied Consent page was electronically signed, the DQ randomly presented the subsequent four pages, one focusing on material drawn from the Typical Dreams Questionnaire (TDQ) (Nielsen, et al., 2003), and one each concerning full reports on Earliest Recalled Dream, Most Important Dream, and Most Recent Dream. Page randomization and randomization of all 55 TDQ theme options was used to minimize order effects in participants’ responses. On each page of the DQ, participants had an option to go back one page or to continue onto the next page, until completion of the survey.
Analyses. The TDQ (Nielsen, et al., 2003) portion of the DQ and all non-open-ended questions could be statistically analyzed within the SurveyGizmo.com platform, and these data were uploaded into a SPSS package for more detailed statistical computation. Open-ended questions (i.e., full dream reports) were content analyzed with the DTS (modified), which accounted for threatening and non-threatening dream themes and occurrences, threat severities, and threat targets. Total words in each dream report and total threat describing words were counted. Frequencies for these data were then tallied. Words in dream reports which were extraneous to the actual dream description (e.g., “”In this one dream I had…”; “I think that this dream happened because I was eating cereal before bed…”) were not included in the total word counts. Words describing threats were considered to be, in most cases, entire sentences in which threat was perceived by the researchers, and in some instances commas differentiated the beginning or end of a threat word-string. Threats were identified by any dream content which, if encountered or occurring during waking would be regarded as threatening by most persons. Comparisons between the three dream types and of the present findings to previous studies are dealt with in the Discussion section. See Appendix B for examples of ERD’s, MRD’s, and MID’s and how these were coded.