Note 1: In the following, the term "déjà vu" will be used to stand for all forms of déjà experience, except where a more differntiated treatment is required.

Note 2: There is some overlap with both the history and explanations pages.

Note 3: There is currently an on-line déjà experience questionnaire which has been collecting data since December, 2004. Some early results from that survey are available there. They will be updated and made available on this website at a later date.

What has been learned from surveys and questionnaires

Down through the years, several facts about déjà experience have been discovered with the help of questionnaires and surveys. The earliest one that I know of was conducted by H. F. Osborne at Princeton and elsewhere, and mentioned in his 1884 article that he published in the North American Review. Unfortunately he doesn't tell how many persons were interviewed (by questionnaire), or anything else about their ages, sex, education, etc. He does say, though, that about one-half had had some form of déjà vu experience.

In 1898, Bernhard-Leroy submitted his (medical) doctoral thesis on the illusion of false recognition (as it was also called). During his research, he distributed a questionnaire (1000 copies) and published it in a French magazine and in an American one. He received only 67 replies of which 49 were usable. He added to these 36 from the published literature which brought the total to 85. He did manage to ask about and record the aspects which Osborne had neglected but he found only one correlated factor: age. Most of the respondents said that they had had more frequent and more intense déjà vu experiences when they were younger, especially during their adolescence. This fact had already been noted, however, by Sander in 1873 and Kraepelin in 1886 - 87. On the other hand, there seemed to be no correlation with sex, race, class, state of health or energy level (fresh or fatigued).

Heymans, in 1904 and 1906, made a critical review of Bernhard-Leroy's results and extended the data by questioning the students in his classes, young people between 20 and 25 years of age, as he noted. In 88 cases, he found 14 (16%) who had often experienced déjà vu, and 41 (47%) who had experienced it occasionally. There seemed to be not much correlation with experiences of depersonalization, a relationship he had hoped to prove. There was more of a correlation with fatigue (see chapter 5). Probably his most interesting finding was that persons prone to déjà vu tended to have a more emotional, more labile nature than their peers.

With regard to incidence, Chapman and Mensh interviewed 220 patients, ambulatory and in bed, in the hospital and clinics of the Washington University (St. Louis). Their results, published in 1951, showed that 35% of the women and 31% of the men recalled having had déjà vu experiences. This agrees amazingly well with Lalande's 1893 estimate of 30% (see Ellis, 1911, p. 240). There seemed to be a slight positive correlation with intelligence (seen from education and occupation) and with travel (those who never traveled tended to have fewer déjà vu experiences). This may still be intelligence-related.) The clearest correlation was the one with adolescence, the highest incidence occurring between ages 15 and 25. The relationship with intelligence may be a spurious one: persons given to thinking and reflection might be more struck by the unusual nature of the experience so that they remember better their encounters with it. It would, though, explain the higher incidence among students as found in Heymans' study.

In 1967, a similar survey was carried out with 161 neurological and 301 psychiatric patients. Richardson and Winokur discovered that 40% and 44% (respectively) had experiences of déjà vu. They found a slightly higher incidence among women psychiatric patients (48%), but otherwise no significant sex differences. They also found that younger patients tended to experience déjà vu more often than older ones. The correlation with profession, education, and travel was substantiated in their data. They also had some African Americans patients in their sample population and found that the incidence of déjà vu among African Americans is significantly lower (only 19%). In a subsequent paper, they investigated the correlation of déjà vu with various neuro- and psychopathologies. Those results will be discussed on the Explanations page.

A few years later, in 1974, a survey by mail was made of 700 persons, picked at random from a telephone directory of Charlottesville, Virginia, by two students of the university there. Palmer and Dennis also sent out 300 to fellow students. With some effort, they managed to collect useful questionnaires from 89% of the students and 51% of the townspeople. Of those responding, 68% from the town and 88% of the student sample reported having had déjà vu experiences. Of the townspeople under 30, 84% gave positive responses. I suspect, though, that these figures can be brought into agreement with the results of the previous studies if we can assume that persons in the town who had had no experiences didn't bother to return the survey form. 68% of 51% comes to 35%. Similarly, 51% of 84% comes to 43%, reflecting the increased incidence among the young. The intelligence factor may be reflected in the much higher incidence among the students (88% of 89%, or 78% in the worst case). McKellar and Simpson, in a 1954 study of hypnagogic imagery, among 110 subjects chosen from the students at Aberdeen University in Scotland, found that 71% had experienced déjà vu.

The table below provides a synopsis of the information obtained in surveys, arranged chronologically, concerning the incidence of déjà vu in the general, healthy population. The grand average on the last line does not include the pre-1900 survey results.

More recent surveys (Greeley, 1987; Gallup and Newport, 1991; Probst and Jansen, 1991; Levin, 1993) have also shown that déjà vu is a common occurrence in the general population with frequencies of incidence ranging from 54.6% to 89%, depending on how the question about déjà vu is phrased and which population is queried. It can thus no longer be considered paranormal (Greeley, 1975: Fox, 1992; Ross and Joshi, 1992) nor pathological (Hufford, 1992). As was found in the older surveys, these also found that it seems to be positively correlated with educational attainment (Fox, 1992).

McCready and Greeley (1976) surveyed 1526 persons within seven age ranges about their experiences of déjà vu. The average incidence for all 1526 came to 61%. The age dependence on incidence, as shown in the following table, is clearly seen from their results.

While they were at it, they looked into the incidence as related to religious convictions and found that déjà vu is claimed less often by Protestants and Catholics (58.5%) than among Jews and others (79.3%). Among 5 so-called mainline Protestant denominations, the Episcopalians were the most frequent experiencers of déjà vu.

A further, very interesting survey has been performed more recently by McClenon (1994) in which he interviewed students of various races and cultures in order to test what effects cultural source might have on so-called anomalous experiences. Based on questionnaire data from Chinese, Japanese and American universities as well as from members of the American Association for the Advancement of Science, he obtained the following frequencies for déjà vu incidence:

From 1984 to 1990, four students in the psychology department of the University of Hamburg under Prof. Paul Probst, conducted investigations about déjà vu in non-clinical populations. 42 students from four faculties at the university were recruited as subjects for the first study, the second was made with 130 police trainees, the third with 86 students, and the fourth with 97 medical students. Two concentrated just on how déjà vu was experienced while the other two were concerned with finding out what relationship there might be between depersonalization/derealization experiences and déjà vu. These have not been published in their entirety, but some of the results they obtained were summarized and made available in a (German) paper by Probst and Jansen (1991). A shorter version appeared in English in "The German Journal of Psychology" (Probst & Jansen, 1994).

Jansen (1991) tested 109 pupils (49 girls and 60 boys) with an average age of 16.1 from five classes in the 9th and 10th grades of a Niedersachsen primary school. They were subjected a battery of tests, including a 13 question déjà vu questionnaire that he had devised which included five questions (taken from Reed, 1979) having to do with Feeling of Knowing, Depersonalization, Tip of the Tongue, Time Gap, and Checking phenomena. A Cognitive Failure Questionnaire, an Everyday Attention Questionnaire, a Self-Attention Questionnaire, two Speed of Pattern Recognition tests, a Visual Memory test, and two Speed of Perception tests were also utilized.

A déjà vu incidence of 87.2% was obtained, but no significant influence of gender was found. A typical déjà vu experience lasted between 1 to 60 seconds for 70.5% of his subjects. 91.6% said it began suddenly, but only 51.6% said it ended suddenly. 36.9% said they found their déjà vu experiences to be pleasant while 48.8% said they were neither pleasant nor unpleasant. Only 4.8% found them to be very pleasant while 14.3% said they were unpleasant.

Like the previous table, the one below provides a synopsis of the information obtained from various surveys (including those of Prof. Probst's students), arranged chronologically, concerning the incidence of déjà vu in college-aged, healthy populations. One notes that the various incidences fall roughly into two groups: those around 70% and those around 85%. This may be due to how the déjà vu question was posed; this needs to be looked into.

For comparison, a study was made in Japan in the same year of 104 persons with temporal lobe epilepsy and it was found that 7, or 6.7%, had déjà vu and/or reminiscence (Sengoku, Toichi & Murai, 1994). This is a much lower rate than that found among healthy Japanese university students in the preceding study. The 1963 survey made by Cole and Zangwill produced similarly low values. These results along with those from surveys of other clinical populations are summarized in the table below.

In 1994 Sno et al. extended Neppe's Screening and Quality questionnaires and included questions from those of Heymann and also Chapman and Mensh to create a standardized scale for investigating déjà vu (IDEA: Inventory for Déjà vu Experiences Assessment). This instrument was then used to compare the occurrence of déjà vu with that of depersonalization in individuals. In their investigation, 48 outpatients suffering from major depression, 42 adolescents with schizophrenic spectrum disorders, 64 outpatients with complex partial seizures (TLE) and 59 healthy subjects were tested. The incidence frequencies of déjà vu among these categories were 72.9%, 76.2%, 71.9% and 71.2%, respectively. They conclude that their results support a link between déjà vu and dissociative phenomena, confirm that it is more frequent among younger people and that there does not seem to be any gender dependence. (English and Dutch versions of the IDEA can be obtained by contacting Dr. Herman N. Sno).

In a survey made using the internet (Snowden and Ito, 2001), questions from Sno's IDEA questionnaire and the Eysenck EPQ Personality Inventory scale were installed on a web-page and the data obtained from 352 "hits" was analyzed. 35 nations were represented among the respondents and a 97.4% affirmative rate provides further evidence that déjà vu occurs to individuals around the world. As to frequency, 49.6% reported ‘a few times per year’. Other frequencies were reported as: ‘a few times a month’ 27.9%, ‘less than one a year’ 12.5%, and ‘at least weekly’ 7.4%. 81.4% reported that they have had déjà vu incidents in situations that had occurred previously to them in a dream (cf. Funkhouser, 2000). 74.1% reported that while their déjà vu experience was happening, they felt they could predict what was going to happen in the next few minutes. 54.8% of 336 participants reported they had their first déjà vu experience between the age of 6 and 15. 9.8% said that their first experience occurred when they were between 16 and 25. About the same number (8.9%), though, said that their first experiences occurred when they were less than 5 years old. 26.2% said they were not sure. When asked to report at what age they had déjà vu incidents most frequently, 45.5% chose the age between 16 and 25, 12.3% for between 5 and 15, 6.9% for between 26 and 35, 0.8% for between 46 and 55, and 28.5% was not sure. 89.4% of the participants who experienced déjà vu reported their experiences lasted ‘less than a minute’ or a ‘couple of minutes’ and of that number (over half 52.1% claimed it lasted about a few seconds). Over 80% of the participants who reported déjà vu experiences denied having a negative mental condition at the time their déjà vu incidents, such as mental and physical fatigue, depression, nervousness, stress, anger, feeling of being frightened, drowsiness and/or physical illness. 61.9% of 331 reported, in fact, they were relaxed before the déjà vu experiences. 43.5% of 331 noted they were happy and cheerful before the incidents. Also, around 90% of the participants who reported déjà vu experiences noted that they have never had adverse physical conditions during their déjà vu experiences, such as headache, blackout, and epileptic seizures. Only one participant reported having a déjà vu experience during epileptic seizure.

Finally, regarding the EPQ-R scores, a one-way ANOVA analysis demonstrated a highly significant relation between déjà vu incidence and extroversion (p = 0.004), but this probably says more about the type of person who responds to such surveys on the Internet than about some relationship between extraversion and déjà vu. The ANOVA result for the déjà vu incidence and the neurosis dimension, however, is non-significant (p = 0.475) and the psychotic dimension showed non-significance as well (p = 0.138). These results support the hypothesis that N and P scores are not significantly abnormal within a non-clinical population in association with déjà vu incidence. Those wishing more information about this investigation can obtain it by contacting Kei Ito.