SCALES

Air Force Research Laboratory and Department of Psychology, University of Plymouth

THIS INSTRUMENT MAY BE OBTAINED FROM THE PRINCIPAL INVESTIGATOR FOR USE BY PROFESSIONALS IN PSYCHOLOGY AND CHRONOBIOLOGY. NOTE THAT IT IS SUBJECT TO CERTAIN CONDITIONS WITH PROPRIETARY RIGHTS EXISTING IN SOME COUNTRIES

The term ‘circadian’ denotes: "The near 24 hour physiologic rhythm that has been observed under free-running conditions, at every system level in nearly all plants and mammals, under near constant environmental conditions" (Brown, 1982, p. 9). Within individual differences approaches to this phenomenon, there is a stated assumption that there exists definite types corresponding to diurnal preference (e.g., Kerkhof, 1985; Roberts, Irvine, & Kyllonen, 1998; Tankova, Adan, & Buela-Casal, 1994). This notion has a long history with anecdotal evidence deriving from a number of sources, including Johnson’s purported "love of lying till noon" (Boswell, 1785/1961), to populist notions of the ‘early bird’ (or ‘lark’) and ‘night owl’. Following Kleitman’s (1939/1963) pioneering research, these propensities were formally conceptualized as a trait, lying along a continuum, that has come to be known as the Morningness-Eveningness (M-E) dimension.

However, extant measures of diurnal preference (also known as circadian type or sometimes also chronotype) appear prone to a number of conceptual and psychometric limitations (e.g., see Roberts, 1997; Roberts et al., 1998; Roberts & Kyllonen, 1998). This interim report serves to provide the following:

(A) Sample items comprising the LOCI. The complete inventory measures three dimensions of chronotype (i.e., ‘Morningness’, ‘Eveningness’, and ‘Propensity for Sleep Debt’) such that sample items are given for each scale. The LOCI protocol has been validated (at the time of writing) in four studies totaling nearly 1800 participants. Data from these studies are still being analyzed, though it is clear that our intention to create a conceptually richer instrument, with impressive psychometric qualities, has been realized. Note that there are two forms of this instrument: LOCI (Alpha) and LOCI (Beta). Our decision to create two forms allows the researcher (or practitioner) the opportunity to examine the stability of circadian type over time, treatments, or experimental interventions.

(B) Preliminary data that demonstrates (in a concise manner) critical statistical properties (means, reliabilities etc) possessed by each scale. The reader familiar with the literature on human chronotypes should find this information valuable for assessing the veracity of our claims surrounding the quality of LOCI.

(C) Validation data in which self-reported LOCI scores are compared (and contrasted) with peer-report LOCI, a sleep-wake diary, and ‘biodata’. This study, conducted with an Australian university population who had friends and family members rate them (and each other) on the LOCI scales, offers the first ‘cross-cultural’ comparisons of chronotype dimensions since three previous studies were conducted with USAF enlistees.

The complete test instrument may be obtained from the Principal Investigator (PI) named in this document providing that it will be used for research purposes. Because of a variety of commercial applications (e.g., selecting appropriate individuals for shiftwork) and strong interest in marketing this instrument, it is restricted to this use and this use only. A technical report detailing conceptual precursors, development, validation, and correlates of this instrument (along with various peer-reviewed articles) will be available in the near future. Any inquiries concerning the use of this scale (which is copyrighted in certain countries) should be directed to the following email address: richardr@psych.usyd.ed.au

LOCI is (in the fashion of the vast majority of self-report psychometric tests) easy to administer. Instructions to Form Alpha are included in the present package, which the experimenter can read aloud with the participant (if they see fit). The only item presenting notable problems to participants in any previous study is Item 4 of LOCI (Beta), wherein a small percentage of military (USAF) enlistees reported not understanding the meaning of ‘flexi-time’. The experimenter is encouraged to answer this query if it arises in the test session.

The six-point rating scale used in the LOCI represents one founded on a solid research tradition. Note that the no items contain qualifiers (e.g., ‘usually’) in the stem, which otherwise is thought to confuse respondents. The readability statistics of Forms Alpha and Beta are particularly impressive: Flesch Reading Ease = 90.3%, Flesch-Kincaid Grade Level = 2.8. This suggests that individuals having three or more years of formal schooling easily understand the scale.

Users of LOCI are encouraged to score the test from 1 (‘Never’) to 6 (‘Always’) for each and every item, entering it directly into an SPPS data file. Roberts and Irvine (1998) have written programs for correcting reverse keying, summing composites, running factor analysis and so forth. These are available from the PI upon request. Note that information by item is valuable because it allows the researcher to check reliability, factor structure and so forth across the various samples that are tested. It is hoped that users of LOCI (in the research context) will make available all findings to either the PI or his associate. In so doing, we hope to build an impressive array of correlates, outcomes, treatments, interventions etc related to this instrument.

Results from three (of the first) validation studies of LOCI may be found at the bottom of the section detailing sample items from each scale. In each of these initial studies, participants were military enlistees undergoing their sixth week of basic training at Lackland AFB, TX.

The PI has just completed a large scale validation study of the LOCI wherein approximately 200 students were administered Form Alpha. These 200 students then administered this scale to two other people (denoted Person A and Person B) who in addition to completing self-reports made peer-reports on the experimenter and the other person. The student sample also completed a sleep-wake diary and various ‘biodata’ questions related to their sleep patterns and meal times. These results, presented in Section 5, suggest that the LOCI possess exceptional validity since the correlation between peer-report and self-reports are among the highest found in the individual differences literature.

Three additional studies have been conducted with the LOCI (total sample size =1600), each of which are currently being scored and collated. The results of these studies will be incorporated into this report as they come to hand. Four new studies are planned for late 1999. These latter studies will focus on student and shift-work populations in Australia, the USA, England, and Uruguay. The study in Uruguay means the LOCI has been translated into Spanish and copies of this translation are available from the PI upon request. It is our practice (in all studies) to examine a variety of psychological and biological phenomena that may correlate with the various scales (e.g., personality, intelligence, mood-state, sleep quality etc).

On the pages below, you will find a number of statements that you should read and decide how well each one of them describes the way you think or feel. Here is an example:

You should reply to all statements. Give your first impression of whether each statement describes the way you think and feel about it. Don’t spend too long on deciding what your answer should be. Answer all statements even if you're not entirely sure of your answer. There are NO right or wrong answers.

LOCI (Morning) is composed of 13 items, two of which are reverse-keyed. Participants respond on the six-point rating scale given in Section 3. (Note all of these specific properties apply to both forms of this scale). Example of an item from each scale is given below. Item numbers correspond to an item’s order of presentation in the full form.

Summary statistics and reliabilities for both LOCI (Morning-Alpha) and LOCI (Morning-Beta), obtained in three studies, are given below. The correlation between parallel forms is also given. Study 3 involved a computer-administered version of LOCI.

LOCI (Evening) is composed of 13 items, four of which are reverse-keyed (indicated below by an asterisk). Participants respond on the six-point rating scale given in Section 3. (Note all of these specific properties apply to both forms of this scale). Examples of two items from each scale are given below. Item numbers correspond to an item’s order of presentation in the full form.

Summary statistics and reliabilities for both LOCI (Evening-Alpha) and LOCI (Evening-Beta), obtained in three studies, are given below. The correlation between parallel forms is also given. Study 3 involved a computer-administered version of LOCI.

LOCI (Sleep) is composed of 12 items. Unlike the LOCI (Morning) and LOCI (Evening) scales, all items are common to both the LOCI (Alpha) and LOCI (Beta) forms of LOCI (Sleep). No item is reverse-keyed. Participants respond on the six-point rating scale given in Section 3. Examples of an item from each scale is given below. Item numbers correspond to an item’s order of presentation in the full form.

Summary statistics and reliabilities for LOCI (Sleep) obtained in three studies, are given below. In these early studies, time constraints dictated that questions composing this scale be ‘spread across’ both the LOCI (Alpha) and LOCI (Beta) forms. (Recall that in the final forms of this scale, all items are the same [i.e., essentially they represent a replication]). Therefore, these data are relevant only to a single administration. Study 3 involved a computer-administered version of LOCI.

This validation study was conducted with LOCI (Alpha) only. In the Tables below, Person E refers to participants composing the student population, while Person A and Person B are people that they tested on both self-report and peer-report versions of LOCI. Although there are some missing data, the average sample size for all three groups approaches N = 200. This table contains descriptive statistics and reliabilities for the self-report data obtained as part of this investigation.

Measure	Morning	Evening	Sleep
Person E (Self-Report)
Mean	24.40	22.40	43.00
S.D.	11.10	10.10	7.70
Cronbach a	0.91	0.89	0.74
Person A (Self-Report)
Mean	28.50	16.30	38.90
S.D.	11.60	11.70	8.70
Cronbach a	0.89	0.90	0.77
Person B (Self-Report)
Mean	30.10	16.20	37.80
S.D.	12.40	11.10	9.10
Cronbach a	0.90	0.88	0.80

The reliabilities of the peer-report measures were only marginally lower than that obtained for the self-report measures on all three scales (i.e., Morningness, Eveningness, Sleepiness). Moreover, the means and standard deviations were remarkably similar. These data are available from the PI upon request. In the table below, correlations between self-report and peer-report LOCI scores are reported.

Peer-Report	Morning (Self-Report)	Evening (Self-Report)	Sleep (Self-Report)
Person E about Person A	0.66	0.68	0.50
Person E about Person B	0.69	0.62	0.56
Person A about Person E	0.57	0.60	0.37
Person A about Person B	0.73	0.63	0.58
Person B about Person E	0.59	0.56	0.40
Person B about Person A	0.61	0.74	0.51

It is worth noting that this data set also has high discriminate validity (i.e., correlations between unrelated measures is generally near zero). The investigation also included information on a sleep-wake diary and biodata completed by person E. This information was correlated with Person E’s self-reported Morning, Evening, and Sleep score and is provided in the table that follows.

Measure	Morning (Self-Report)	Evening (Self-Report)	Sleep (Self-Report)
Person E’s Bedtime (Diary)	-0.45	0.60	0.29
Person E Wake Time (Diary)	-0.52	0.42	0.42
Person E’s Bedtime (Self-Report)	-0.42	0.57	0.14
Person E Wake Time (Self-Report)	-0.50	0.32	0.32
Person E’s Breakfast (Self-Report)	-0.42	0.34	0.18
Person E’s Lunch (Self-Report)	-0.22	0.18	0.15
Person E’s Dinner (Self-Report)	-0.24	0.31	0.07

Boswell, J. (1785/1961). Journal of the Tour of the Hebrides with Samuel Johnson. New York: McGraw-Hill.

Brown, F. M. (1982). Rhythmicity as an emerging variable for psychology. In F. M. Brown & R. C. Graeber (Eds.), Rhythmic Aspects of Behavior. Hillsdale, NJ: Erlbaum.

Kerkhof, G. A. (1985). Inter-individual differences in the human circadian system. A review. Biological Psychology, 20, 83-112.

Kleitman, N. (1939/1963). Sleep and Wakefulness. Chicago, IL: University of Chicago Press.

Roberts, R. D. (1997). Individual differences in diurnal preference. Paper presented at the Air Force Office of Scientific Research (AFSOR) Chronobiology and Neural Adaptation Conference. Colorado Springs, CO, USA, September, 1997.

Roberts, R. D., & Irvine, S. (1999). Development and construct validation of the Lark-Owl (Chronotype) Indicator (LOCI): Technical Report. Manuscript in Preparation.

Roberts, R. D., Irvine, S., & Kyllonen, P. C. (1999). Personality and circadian type: A review and empirical investigation. Manuscript in Preparation.

Roberts, R. D., & Kyllonen, P. C. (1999). Morningness-Eveningness and intelligence: Early to bed, early to rise will likely make you anything but wise! Personality and Individual Differences. In press.

Tankova, I., Adan, A. A., & Buela-Casal, G. (1994). Circadian typology and individual differences: A review. Personality and Individual Differences, 16, 5, 671-684.

Alpha-Form	Beta-Form
1. I find it easy to get up in the morning.	4. When working on ‘flexi-time’, I'd show up early in the morning.

	Study 1 (N = 465)	Study 2 (N = 300)	Study 3 (N = 367)
Alpha Form
Mean	28.30	30.60	32.70
S.D.	11.00	10.50	11.50
Cronbach a	0.90	0.88	0.88
Beta Form
Mean	28.60	31.10	33.90
S.D.	11.10	11.00	12.00
Cronbach a	0.90	0.90	0.90
Combined
R (forms)	0.94	0.93	0.93

Alpha-Form	Beta-Form
5. I enjoy working unusual hours.	6.* I need to relax, not doing much, in the evening.
7.* I feel tired next day when I stay up late, past midnight.	8. I can manage on a few hours sleep.

	Study 1 (N = 465)	Study 2 (N = 300)	Study 3 (N = 367)
Alpha Form
Mean	22.90	22.40	21.70
S.D.	9.40	9.30	9.90
Cronbach a	0.84	0.83	0.82
Beta Form
Mean	23.00	22.50	22.60
S.D.	8.80	8.10	9.20
Cronbach a	0.80	0.76	0.78
Combined
R (forms)	0.91	0.90	0.91

Alpha-Form	Beta-Form
11. On the weekends, I sleep past noon.	25. I need 8 hours sleep.