Journal of Abnormal Psychology 1977, Vol. 86, No. 5, 561-564
Subjective Characteristics of Sleep Efficiency Frederick J. Evans The Institute of Pennsylvania Hospital, Philadelphia, and University of Pennsylvania
A 33-item questionnaire concerning typical self-report sleep habits and behavior was factor analyzed. After rotation of the principal-components solution, 5 matching factors were isolated in 2 samples of 92 and 180 volunteer college students. These factors were tentatively interpreted as (a) Voluntary Control of Sleep, (b) Sleep-Onset Difficulty, (c) Difficulty Maintaining Sleep, (d) Recall of Dreaming, and (e) Cognitive Control of Sleep Mentation, Parametric data on estimated factor scores for 640 college students are reported. While it is not necessary that the dimensions of sleep efficiency should correlate with physiological and electroencephalographic data, some data suggest that they may do so successfully. The implications of these subjective dimensions of sleep quality for the understanding of the functions of sleep and the treatment of sleep disorders are briefly discussed. There has been a considerable amount of research on the physiological parameters of nighttime sleep, but relatively little attention has been given to those characteristics that an individual would typically use to describe his own sleep patterns and habits. A person may report that he slept deeply, or that he habitually has difficulty falling asleep, yet he usually remains blissfully unaware of his rapid eye movement (REM) percentage time or the relation between his falling asleep and the time to his first electroencephalographic (EEG) sleep spindle and other objective measures that have been adopted to describe sleep behavior (Rechtschaffen & Kales, 1968). Even when an individual asserts he has difficulty in falling asleep, it cannot be assumed that he necessarily takes longer to fall This research was supported in part by Contract #DADA17-71-C-1120 from the U.S. Army Medical Research and Development Command and in part by a grant from the Institute for Experimental Psychiatry, Boston, Massachusetts. I wish to thank my colleagues Harvey D, Cohen, Mary R. Cook, Charles Graham, Betsy E. Lawrence, Emily Carota Orne, Martin T. Orne, Helen M. Pettinati, Pamela A. Markowsky, Tony Van Campcn, William M. Waid, and Stuart K. Wilson for their substantive comments during the preparation of this report. I also wish to thank Jeremy DeLong, Eileen Grabiec, James Hamos, Alexander Myers, and Neal Shore for their technical assistance. Requests for reprints should be sent to Frederick J. Evans, Unit for Experimental Psychiatry, The Institute of Pennsylvania Hospital, 111 North 49th Street, Philadelphia, Pennsylvania 19139.
561
asleep by EEG criteria than others who claim they fall asleep easily (Kales, Kales, & Bixler, 1974). The few laboratory studies that have examined the relation between subjective and objective criteria of sleep have mostly focused on insomnia and sleep deprivation (Dement, 1972; Webb, 1973). Although there have been some extensive surveys of nonclinical self-reported sleep length and related parameters (e.g., Johns, Gay, Goodyear, & Masterton, 1971; Lewis, 1969; Tune, 1969), little is known about how people describe their own sleep in terms of such variables as how easily they fall asleep, how well they sleep, when they sleep best, how easily they waken, to what extent they awaken during the night, and so on. Attempts to explore the functions and quality of sleep should be sensitive to the experiences reported by the individual as well as to the objective indices of his sleep behavior. Self-report variables that are found to affect sleep satisfaction will eventually require validation and comparison with objective sleep criteria. Although the task of exploring subjective factors involved in sleep efficiency cannot be limited to the collection of self-report data, documenting the meaningful experiential patterns is nonetheless an essential first step. Factor analysis was chosen as an appropriate method to explore interrelations among self-reported characteristics of sleep. Thus, the aim of the present article is to investigate subjective sleep characteristics using factor analytic techniques to present a preliminary descriptive account of variables affecting sleep efficiency.
562
SHORT REPORTS Method 1
The 33-item sleep questionnaire included a range of questions carefully sampling the domain of self-reported sleep habits. Two typical questions were "Do you take catnaps during the day?" and "Do you have difficulty falling asleep?" The subject checked the most appropriate of five response choices: always (scored 5), usually, sometimes, rarely, or never (scored 1). Responses of 92 and 180 volunteer male students (Samples 1 and 2, respectively) were intercorrelated and factor analyzed using the principal-components method (inserting unities in the diagonals). After six factors were rotated by the varimax method, additional blind orthogonal hand rotations to achieve simple structure (maximizing zero factor loadings) were completed. The five factors retained accounted for 44% and 49 % of the total matrix variance in Samples 1 and 2, respectively. Estimated factor scores were obtained on an additional sample of 640 volunteer students (Sample 3).Results Estimated factor scores derived from the best five defining items (range of 25 to S) were not significantly correlated (except for the first two factors reported). Factor matching was conducted primarily by inspection of the significant factor loadings without reference to the item content. Correlations between the two sets of factor loadings were mostly high. The correlation between factor loadings on the 33 variables for a given factor in Sample 1 and its matching factor in Sample 2 was always higher than its correlation with any other combinations of factors in Samples 1 and 2. Voluntary Control oj Sleep The Voluntary Control of Sleep factor accounts for 24% and 20% of the reliable variance in Samples 1 and 2, respectively. The factor loadings for 33 variables correlate .82 across the two samples. The mean and standard deviation for the estimated factor scores in Sample 3 (TV =640) was 13.9±2.84. The person who scores highly on this cluster reports that he falls asleep easily at night (the respective factor loadings for this item in Samples 1 and 2 were .S3 and .SO); falls asleep easily in a variety of unusual surroundings, such as during a movie or concert (.51 and .47) or
in a plane or car (.66 and .63); catnaps regularly (.62 and .49); and has no difficulty falling asleep ( — .43 and —.44). In general, such an individual can go to sleep "at will" (.65 and .62). Thus, the high-scoring individual reports being ready, willing, and able to fall asleep at any time and in any place, and apparently has control over the sleep process. Such a skill may involve the ability to efficiently utilize the 24hour day to obtain sleep at the most appropriate time. Sleep-Onset
Difficulty
Accounting for 20% and 16% of the reliable variance, the loadings correlate .54 across Samples 1 and 2 for the Sleep-Onset Difficulty factor. The Sample 3 mean estimated factor score was 11.7±3.08. The subject with a high score on this factor reports difficulty in falling asleep (factor loadings in Samples 1 and 2 were .65 and .64, respectively), particularly before examinations or similar important events (.69 and .47); takes sleeping medications (.37 and .40); and is unable to fall asleep easily (—.55 and —.61) or at will (—.30 and —.49). This dimension may reflect mild or incipient sleep-onset insomnia. The mediating roles of stress and anxiety will be important to investigate. Difficulty
Maintaining Sleep
The Difficulty Maintaining Sleep factor accounted for 12% and 20% of the reliable variance with a correlation between loadings of .44 in Samples 1 and 2. The mean estimated factor score in Sample 3 was 14.0±3.26. This factor is most strongly, but inconsistently, defined by a question that was repeated in the following positive and negative formats: "Are you a deep sleeper?" (—.24 and —.81) and "Are you a light sleeper?" (.05 and .76). 1 The original version of this questionnaire was developed by the late Donald N. O'Connell. Copies of the questionnaire (and a more detailed summary oi the factor analytic results, procedures for obtaining estimated factor scores, as well as parametric data) are available from the author. 2 Harvey D. Cohen, Mary R. Cook, Julio Dittborn, John F. Kihlstrom, Edgar P. Nace, Donald N. O'Connell, and Ronald E. Shor collected these data at the Unit for Experimental Psychiatry, The Institute of Pennsylvania Hospital, Philadelphia, Pennsylvania.
SHORT REPORTS The consistent items include awakening to an expected sound (.48 and .59), awakening during the night (.52 and ,42), difficulty in sleeping in strange surroundings (.27 and .42), and having to get up during the night (.29 and .37). A subject with a high score on this factor has difficulty in staying asleep at night, although he does not necessarily report difficulty in falling asleep. Recall of Dreaming Accounting for 16% of the variance in both Samples 1 and 2, the factor loadings correlated .81. The mean estimated factor score in Sample 3 was 15.8±3.19. The high-scoring subject reports he dreams every night (.82 and .82), never has dreamless sleep (—.81 and —.79), dreams in color (.35 and .52), and dreams about the previous day's experiences (.58 and .42). The relevant questions all involve dreaming. The close association between dream recall and awakening from REM sleep during the night is widely known from the popular press, and the effects of response bias should be considered. The factors responsible for wide individual differences in dream recall remain obscure. Cohen (1974) found little empirical support for explanations in terms of dream salience, interference effects, and repression. The simplest hypothesis that some dreams are recalled only when we awaken directly (either spontaneously or by an alarm) and, perhaps, abruptly from REM sleep has apparently never been tested. This factor most likely relates to individual differences in the frequency of dreaming or the ability to recall dreams. Cognitive Control of Sleep Mentation The Cognitive Control of Sleep Mentation factor accounted for 16% and 12% of the reliable variance in Samples 1 and 2, respectively; factor loadings for these samples correlated .59. The estimated mean factor score in Sample 3 was 11.0 ±3,33. The high-scoring subject claims he can change dream content at will (.53 and .43), can decide ahead of time what he should dream about (.65 and .41), and can perhaps incorporate external sounds into his dreams (.56 and .27). This less clearly denned factor seems to tap the purported ability to influence and change one's dreams and cognitive activity during sleep. The relatively sparse evidence available pertaining to the successful cognitive manipulation of
563
sleep mentation and information processing during sleep has indicated that such control is at least feasible (Evans, 1972). 3 Discussion This study explored some of the parameters of self-reported sleep patterns, with the aim of learning more about those variables that might contribute to the effectiveness and quality of sleep. The five factors isolated would seem to have heuristic value in describing subjective sleep patterns. One factor appears to involve the ease with which individuals could fall asleep, maintain, and voluntarily control sleep in a variety of circumstances. Another involves individual differences in dream recall. Two factors may relate to different types of incipient insomnia: one involving sleep-onset problems and the other involving an inability to maintain sleep once it has been achieved. Another factor apparently identified individuals who claimed that they could cognitively manipulate sleep and dream functions. The conceptual and empirical distinction between sleep-onset insomnia and sleep-maintenance insomnia has important implications for understanding reports of sleep disturbances. These two kinds of insomnia are associated in clinical lore, respectively, with depression and neuroses. Whether these dimensions are normal manifestations of severe sleep disturbances, and if so, whether they are prognostic of incipient psychopathology requires further investigation. Additional research will be necessary to validate the utility of the subjective parameters of sleep. Future investigations will need to include studies of clinical populations in which sleep disturbances are reported and treated, studies of the predictive interrelations using subjective reports, and appropriate studies that utilize EEG definitions of sleep and its stages. Several illustrative studies can be summarized. For 20 students with insomnia, Graham (Note 1) reported significant increases in the Voluntary Control
3
A factor accounting for 11% of the reliable variance in Sample 1 had no counterpart in Sample 2. Only three items loaded significantly: (a) the ability to sleep late on Sundays (.58), (b) likes to sleep (.61), and (c) tends to sleep more than 8 hours per night (.74). This factor might involve the distinction between "long" and "short" sleepers, that is, individual differences in the amount of sleep required by a person.
564
SHORT REPORTS
of Sleep factor and significant decreases in the Sleep-Onset Difficulty factor 6 months after the completion of a self-hypnosis, relaxation treatment program. Compared to 169 non-nappers, 261 students who nap scored significantly higher on the Voluntary Control of Sleep dimension (p < .0001) and reported that they were able to fall asleep easily and readily in a number of unusual situations, including on long plane, train, and car trips; during lectures, speeches, movies, plays, and television shows; after a good meal; and at times of stress (Evans, Cook, Cohen, Orne, & Orne, in press). Subjects with higher scores on the Voluntary Control of Sleep factor fell asleep more quickly at night (N — 19, p< .05; Evans, 1972) and during a laboratory nap (N = 33, p < .001; Evans et al., in press) than those with lower scores, using standard EEG criteria of sleep onset. Subjects who have high Voluntary Control of Sleep scores are also more hypnotizable than those with lower scores (Evans, in press). Subjects able to respond behaviorally to meaningful suggestions administered during sleep (Evans, 1972) scored higher on the Cognitive Control of Sleep Mentation dimension. While it would be encouraging if these subjective patterns correlate with the relevant objective criterion measures of sleep, the usefulness of the present data does not depend on such correlations. For example, the patient reporting insomnia may require treatment for his experiential insomnia regardless of the number of hours he may sleep according to EEG criteria. Nevertheless, to the extent that there is convergence between subjective and objective patterns of sleep, this line of research is important for understanding qualitative aspects of sleep, for selecting appropriate research samples, and for the treatment of sleep pathology. Within the limits of the domain of the subjective sleep characteristics sampled, these are some of the dimensions related to the ways in which an individual describes his subjective experience of sleep. Further research will indicate whether they also form the basis of patients' clinical symptomatologies concerning sleep.
Reference Note 1. Graham, K. R. Changes in the subjective quality of sleep during the treatment of insomnia. Unpublished manuscript, Muhlenberg College, 1976. References Cohen, D. B. Toward a theory of dream recall. Psychological Bulletin, 1974, 81, 138-154. Dement, W. C. Some must watch while some must sleep (2nd ed.). San Francisco: W. H. Freeman, 1972. Evans, F. J. Hypnosis and sleep: Techniques for exploring cognitive activity during sleep. In E. Fromm & R. E. Shor (Eds.), Hypnosis: Research developments and perspectives. Chicago, 111.: Aldine, 1972. Evans, F. J. Hypnosis and sleep: The control of altered states of awareness. In W. E. Edmonston (Ed.), Conceptual and investigative approaches to hypnosis and hypnotic phenomena. New York: Annals of the New York Academy of Sciences, in press. Evans, F. J., Cook, M. R., Cohen, H. D., Orne, E. C., & Orne, M. T. Appetitive and replacement naps: EEG and behavior. Science, in press. Johns, M. W., Gay, T. J. A., Goodyear, M. D. E., & Masterton, J. P. Sleep habits of healthy young adults: Use of a sleep questionnaire. British Journal of Preventive and Social Medicine, 1971, 25, 236-241. Kales, A., Kales, J., & Bixler, E. 0. Insomnia: An approach to management and treatment. Psychiatric Annals, 1974, 4, 28-44. Lewis, S. A. Subjective estimates of sleep: An EEG evaluation. British Journal of Psychology, 1969, 60, 203-208. Rechtschaffen, A., & Kales, A. (Eds.). A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Washington, B.C.: Public Health Service, U.S. Government Printing Office, 1968. Tune, G. S. Sleep and wakefulness in S09 normal human adults. British Journal of Medical Psychology, 1969, 42, 75-80. Webb, W. B. (Ed.). Sleep: An active process. Research and commentary. Glenview, 111.: Scott, Foresman, 1973. Received June 7, 1976 Revision received March 3, 1977 •
Subjective Characteristics of Sleep Efficiency Frederick J. Evans The Institute of Pennsylvania Hospital, Philadelphia, and University of Pennsylvania
A 33-item questionnaire concerning typical self-report sleep habits and behavior was factor analyzed. After rotation of the principal-components solution, 5 matching factors were isolated in 2 samples of 92 and 180 volunteer college students. These factors were tentatively interpreted as (a) Voluntary Control of Sleep, (b) Sleep-Onset Difficulty, (c) Difficulty Maintaining Sleep, (d) Recall of Dreaming, and (e) Cognitive Control of Sleep Mentation, Parametric data on estimated factor scores for 640 college students are reported. While it is not necessary that the dimensions of sleep efficiency should correlate with physiological and electroencephalographic data, some data suggest that they may do so successfully. The implications of these subjective dimensions of sleep quality for the understanding of the functions of sleep and the treatment of sleep disorders are briefly discussed. There has been a considerable amount of research on the physiological parameters of nighttime sleep, but relatively little attention has been given to those characteristics that an individual would typically use to describe his own sleep patterns and habits. A person may report that he slept deeply, or that he habitually has difficulty falling asleep, yet he usually remains blissfully unaware of his rapid eye movement (REM) percentage time or the relation between his falling asleep and the time to his first electroencephalographic (EEG) sleep spindle and other objective measures that have been adopted to describe sleep behavior (Rechtschaffen & Kales, 1968). Even when an individual asserts he has difficulty in falling asleep, it cannot be assumed that he necessarily takes longer to fall This research was supported in part by Contract #DADA17-71-C-1120 from the U.S. Army Medical Research and Development Command and in part by a grant from the Institute for Experimental Psychiatry, Boston, Massachusetts. I wish to thank my colleagues Harvey D, Cohen, Mary R. Cook, Charles Graham, Betsy E. Lawrence, Emily Carota Orne, Martin T. Orne, Helen M. Pettinati, Pamela A. Markowsky, Tony Van Campcn, William M. Waid, and Stuart K. Wilson for their substantive comments during the preparation of this report. I also wish to thank Jeremy DeLong, Eileen Grabiec, James Hamos, Alexander Myers, and Neal Shore for their technical assistance. Requests for reprints should be sent to Frederick J. Evans, Unit for Experimental Psychiatry, The Institute of Pennsylvania Hospital, 111 North 49th Street, Philadelphia, Pennsylvania 19139.
561
asleep by EEG criteria than others who claim they fall asleep easily (Kales, Kales, & Bixler, 1974). The few laboratory studies that have examined the relation between subjective and objective criteria of sleep have mostly focused on insomnia and sleep deprivation (Dement, 1972; Webb, 1973). Although there have been some extensive surveys of nonclinical self-reported sleep length and related parameters (e.g., Johns, Gay, Goodyear, & Masterton, 1971; Lewis, 1969; Tune, 1969), little is known about how people describe their own sleep in terms of such variables as how easily they fall asleep, how well they sleep, when they sleep best, how easily they waken, to what extent they awaken during the night, and so on. Attempts to explore the functions and quality of sleep should be sensitive to the experiences reported by the individual as well as to the objective indices of his sleep behavior. Self-report variables that are found to affect sleep satisfaction will eventually require validation and comparison with objective sleep criteria. Although the task of exploring subjective factors involved in sleep efficiency cannot be limited to the collection of self-report data, documenting the meaningful experiential patterns is nonetheless an essential first step. Factor analysis was chosen as an appropriate method to explore interrelations among self-reported characteristics of sleep. Thus, the aim of the present article is to investigate subjective sleep characteristics using factor analytic techniques to present a preliminary descriptive account of variables affecting sleep efficiency.
562
SHORT REPORTS Method 1
The 33-item sleep questionnaire included a range of questions carefully sampling the domain of self-reported sleep habits. Two typical questions were "Do you take catnaps during the day?" and "Do you have difficulty falling asleep?" The subject checked the most appropriate of five response choices: always (scored 5), usually, sometimes, rarely, or never (scored 1). Responses of 92 and 180 volunteer male students (Samples 1 and 2, respectively) were intercorrelated and factor analyzed using the principal-components method (inserting unities in the diagonals). After six factors were rotated by the varimax method, additional blind orthogonal hand rotations to achieve simple structure (maximizing zero factor loadings) were completed. The five factors retained accounted for 44% and 49 % of the total matrix variance in Samples 1 and 2, respectively. Estimated factor scores were obtained on an additional sample of 640 volunteer students (Sample 3).Results Estimated factor scores derived from the best five defining items (range of 25 to S) were not significantly correlated (except for the first two factors reported). Factor matching was conducted primarily by inspection of the significant factor loadings without reference to the item content. Correlations between the two sets of factor loadings were mostly high. The correlation between factor loadings on the 33 variables for a given factor in Sample 1 and its matching factor in Sample 2 was always higher than its correlation with any other combinations of factors in Samples 1 and 2. Voluntary Control oj Sleep The Voluntary Control of Sleep factor accounts for 24% and 20% of the reliable variance in Samples 1 and 2, respectively. The factor loadings for 33 variables correlate .82 across the two samples. The mean and standard deviation for the estimated factor scores in Sample 3 (TV =640) was 13.9±2.84. The person who scores highly on this cluster reports that he falls asleep easily at night (the respective factor loadings for this item in Samples 1 and 2 were .S3 and .SO); falls asleep easily in a variety of unusual surroundings, such as during a movie or concert (.51 and .47) or
in a plane or car (.66 and .63); catnaps regularly (.62 and .49); and has no difficulty falling asleep ( — .43 and —.44). In general, such an individual can go to sleep "at will" (.65 and .62). Thus, the high-scoring individual reports being ready, willing, and able to fall asleep at any time and in any place, and apparently has control over the sleep process. Such a skill may involve the ability to efficiently utilize the 24hour day to obtain sleep at the most appropriate time. Sleep-Onset
Difficulty
Accounting for 20% and 16% of the reliable variance, the loadings correlate .54 across Samples 1 and 2 for the Sleep-Onset Difficulty factor. The Sample 3 mean estimated factor score was 11.7±3.08. The subject with a high score on this factor reports difficulty in falling asleep (factor loadings in Samples 1 and 2 were .65 and .64, respectively), particularly before examinations or similar important events (.69 and .47); takes sleeping medications (.37 and .40); and is unable to fall asleep easily (—.55 and —.61) or at will (—.30 and —.49). This dimension may reflect mild or incipient sleep-onset insomnia. The mediating roles of stress and anxiety will be important to investigate. Difficulty
Maintaining Sleep
The Difficulty Maintaining Sleep factor accounted for 12% and 20% of the reliable variance with a correlation between loadings of .44 in Samples 1 and 2. The mean estimated factor score in Sample 3 was 14.0±3.26. This factor is most strongly, but inconsistently, defined by a question that was repeated in the following positive and negative formats: "Are you a deep sleeper?" (—.24 and —.81) and "Are you a light sleeper?" (.05 and .76). 1 The original version of this questionnaire was developed by the late Donald N. O'Connell. Copies of the questionnaire (and a more detailed summary oi the factor analytic results, procedures for obtaining estimated factor scores, as well as parametric data) are available from the author. 2 Harvey D. Cohen, Mary R. Cook, Julio Dittborn, John F. Kihlstrom, Edgar P. Nace, Donald N. O'Connell, and Ronald E. Shor collected these data at the Unit for Experimental Psychiatry, The Institute of Pennsylvania Hospital, Philadelphia, Pennsylvania.
SHORT REPORTS The consistent items include awakening to an expected sound (.48 and .59), awakening during the night (.52 and ,42), difficulty in sleeping in strange surroundings (.27 and .42), and having to get up during the night (.29 and .37). A subject with a high score on this factor has difficulty in staying asleep at night, although he does not necessarily report difficulty in falling asleep. Recall of Dreaming Accounting for 16% of the variance in both Samples 1 and 2, the factor loadings correlated .81. The mean estimated factor score in Sample 3 was 15.8±3.19. The high-scoring subject reports he dreams every night (.82 and .82), never has dreamless sleep (—.81 and —.79), dreams in color (.35 and .52), and dreams about the previous day's experiences (.58 and .42). The relevant questions all involve dreaming. The close association between dream recall and awakening from REM sleep during the night is widely known from the popular press, and the effects of response bias should be considered. The factors responsible for wide individual differences in dream recall remain obscure. Cohen (1974) found little empirical support for explanations in terms of dream salience, interference effects, and repression. The simplest hypothesis that some dreams are recalled only when we awaken directly (either spontaneously or by an alarm) and, perhaps, abruptly from REM sleep has apparently never been tested. This factor most likely relates to individual differences in the frequency of dreaming or the ability to recall dreams. Cognitive Control of Sleep Mentation The Cognitive Control of Sleep Mentation factor accounted for 16% and 12% of the reliable variance in Samples 1 and 2, respectively; factor loadings for these samples correlated .59. The estimated mean factor score in Sample 3 was 11.0 ±3,33. The high-scoring subject claims he can change dream content at will (.53 and .43), can decide ahead of time what he should dream about (.65 and .41), and can perhaps incorporate external sounds into his dreams (.56 and .27). This less clearly denned factor seems to tap the purported ability to influence and change one's dreams and cognitive activity during sleep. The relatively sparse evidence available pertaining to the successful cognitive manipulation of
563
sleep mentation and information processing during sleep has indicated that such control is at least feasible (Evans, 1972). 3 Discussion This study explored some of the parameters of self-reported sleep patterns, with the aim of learning more about those variables that might contribute to the effectiveness and quality of sleep. The five factors isolated would seem to have heuristic value in describing subjective sleep patterns. One factor appears to involve the ease with which individuals could fall asleep, maintain, and voluntarily control sleep in a variety of circumstances. Another involves individual differences in dream recall. Two factors may relate to different types of incipient insomnia: one involving sleep-onset problems and the other involving an inability to maintain sleep once it has been achieved. Another factor apparently identified individuals who claimed that they could cognitively manipulate sleep and dream functions. The conceptual and empirical distinction between sleep-onset insomnia and sleep-maintenance insomnia has important implications for understanding reports of sleep disturbances. These two kinds of insomnia are associated in clinical lore, respectively, with depression and neuroses. Whether these dimensions are normal manifestations of severe sleep disturbances, and if so, whether they are prognostic of incipient psychopathology requires further investigation. Additional research will be necessary to validate the utility of the subjective parameters of sleep. Future investigations will need to include studies of clinical populations in which sleep disturbances are reported and treated, studies of the predictive interrelations using subjective reports, and appropriate studies that utilize EEG definitions of sleep and its stages. Several illustrative studies can be summarized. For 20 students with insomnia, Graham (Note 1) reported significant increases in the Voluntary Control
3
A factor accounting for 11% of the reliable variance in Sample 1 had no counterpart in Sample 2. Only three items loaded significantly: (a) the ability to sleep late on Sundays (.58), (b) likes to sleep (.61), and (c) tends to sleep more than 8 hours per night (.74). This factor might involve the distinction between "long" and "short" sleepers, that is, individual differences in the amount of sleep required by a person.
564
SHORT REPORTS
of Sleep factor and significant decreases in the Sleep-Onset Difficulty factor 6 months after the completion of a self-hypnosis, relaxation treatment program. Compared to 169 non-nappers, 261 students who nap scored significantly higher on the Voluntary Control of Sleep dimension (p < .0001) and reported that they were able to fall asleep easily and readily in a number of unusual situations, including on long plane, train, and car trips; during lectures, speeches, movies, plays, and television shows; after a good meal; and at times of stress (Evans, Cook, Cohen, Orne, & Orne, in press). Subjects with higher scores on the Voluntary Control of Sleep factor fell asleep more quickly at night (N — 19, p< .05; Evans, 1972) and during a laboratory nap (N = 33, p < .001; Evans et al., in press) than those with lower scores, using standard EEG criteria of sleep onset. Subjects who have high Voluntary Control of Sleep scores are also more hypnotizable than those with lower scores (Evans, in press). Subjects able to respond behaviorally to meaningful suggestions administered during sleep (Evans, 1972) scored higher on the Cognitive Control of Sleep Mentation dimension. While it would be encouraging if these subjective patterns correlate with the relevant objective criterion measures of sleep, the usefulness of the present data does not depend on such correlations. For example, the patient reporting insomnia may require treatment for his experiential insomnia regardless of the number of hours he may sleep according to EEG criteria. Nevertheless, to the extent that there is convergence between subjective and objective patterns of sleep, this line of research is important for understanding qualitative aspects of sleep, for selecting appropriate research samples, and for the treatment of sleep pathology. Within the limits of the domain of the subjective sleep characteristics sampled, these are some of the dimensions related to the ways in which an individual describes his subjective experience of sleep. Further research will indicate whether they also form the basis of patients' clinical symptomatologies concerning sleep.
Reference Note 1. Graham, K. R. Changes in the subjective quality of sleep during the treatment of insomnia. Unpublished manuscript, Muhlenberg College, 1976. References Cohen, D. B. Toward a theory of dream recall. Psychological Bulletin, 1974, 81, 138-154. Dement, W. C. Some must watch while some must sleep (2nd ed.). San Francisco: W. H. Freeman, 1972. Evans, F. J. Hypnosis and sleep: Techniques for exploring cognitive activity during sleep. In E. Fromm & R. E. Shor (Eds.), Hypnosis: Research developments and perspectives. Chicago, 111.: Aldine, 1972. Evans, F. J. Hypnosis and sleep: The control of altered states of awareness. In W. E. Edmonston (Ed.), Conceptual and investigative approaches to hypnosis and hypnotic phenomena. New York: Annals of the New York Academy of Sciences, in press. Evans, F. J., Cook, M. R., Cohen, H. D., Orne, E. C., & Orne, M. T. Appetitive and replacement naps: EEG and behavior. Science, in press. Johns, M. W., Gay, T. J. A., Goodyear, M. D. E., & Masterton, J. P. Sleep habits of healthy young adults: Use of a sleep questionnaire. British Journal of Preventive and Social Medicine, 1971, 25, 236-241. Kales, A., Kales, J., & Bixler, E. 0. Insomnia: An approach to management and treatment. Psychiatric Annals, 1974, 4, 28-44. Lewis, S. A. Subjective estimates of sleep: An EEG evaluation. British Journal of Psychology, 1969, 60, 203-208. Rechtschaffen, A., & Kales, A. (Eds.). A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Washington, B.C.: Public Health Service, U.S. Government Printing Office, 1968. Tune, G. S. Sleep and wakefulness in S09 normal human adults. British Journal of Medical Psychology, 1969, 42, 75-80. Webb, W. B. (Ed.). Sleep: An active process. Research and commentary. Glenview, 111.: Scott, Foresman, 1973. Received June 7, 1976 Revision received March 3, 1977 •
