Towards a deeper understanding of psychological aspects of lighting

(1)

Publisher’s version / Version de l'éditeur:

Journal of the Illuminating Engineering Society, 19, 2, pp. 155-160, 1990

READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE. https://nrc-publications.canada.ca/eng/copyright

Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.

Questions? Contact the NRC Publications Archive team at

PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.

NRC Publications Archive

Archives des publications du CNRC

This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.

Access and use of this website and the material on it are subject to the Terms and Conditions set forth at

Towards a deeper understanding of psychological aspects of lighting

Tiller, D. K.

https://publications-cnrc.canada.ca/fra/droits

L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.

NRC Publications Record / Notice d'Archives des publications de CNRC:

https://nrc-publications.canada.ca/eng/view/object/?id=57f20d02-31ad-4427-bae3-80f2fb8b1598 https://publications-cnrc.canada.ca/fra/voir/objet/?id=57f20d02-31ad-4427-bae3-80f2fb8b1598

(2)

S e r

T

H

1 National Research

Conseil national

1

*

1 Council Canada

de recherches Canada

I

no,

1342

Institute for

lnstitut de

Research in

recherche en

Construction

construction

Toward a Deemr Understanding

-

of P s y c h o l o g ~ l

Aspects of

Lighting

by D.K. Tiller

ANALYZED.

Reprinted from

Journal of the Illuminating Engineering Society

Vol. 19, No. 2,1990

p. 155-160

(IRC Paper No. 1842)

(3)

Toward a Deeper Understanding of Psychological

Aspects of Lighting

Dale

K.

TiRer

Interior and exterior spaces are illuminated so people can safely and accurately perform such visual tasks as walking and reading. In lighting for these kinds of activities, illuminating engineers and lighting designers are most concerned with the quantity of light in a space. Many believe lighting can also be used to cue particular impressions and emotional reactions in people. To successfully cue these subjective effects, designer and illuminating engineers must have a firm understanding of the quality of light in a space. Never- theless, more research has been directed at understanding the quantity of illumination required to perform different visual tasks than to investigating more qualitative effects of lighting. While a reasonable understanding of the factors that influence visibility has been achieved' and investigators agree on major unresolved issues,'research on the more qualitative and psychological effects of lighting has been sporadic and lacks a shared agenda to guide investigators.

However, vigorous activity by committees of the 11- luminating Engineering Society of North America (IESNA) and Commission International De L'Eclairage (CIE) will soon establish an agenda for future research on psychological aspects of lighting. The fragmented nature of previous research in this area is at least partly because powerful and flexible tools for characterizing luminous interiors, and statistical methods for relating physical measures to subjective effects have, until recently, been unavailable. This article reviews previous work on the psychological aspects of lighting. It also highlights some of the pitfalls of conventional measurement techniques typically used to investigate psychological aspects of lighting, and gives suggestions on how to avoid them. Finally, the article describes how new tools and statistical procedures can be integrated into a new experimental paradigm, leading to a more fun- damental understanding of lighting quality than previously available.

Previous research

The earliest work on psychological effects of lighting involved what could be called a "correlation" method: variations in some physically measured aspect of light were related to a single subjective

judgement. For instance, illuminance was often related to subjective preferences.

This simple approach gradually fell into disfavor, and was replaced by methods using more powerful and sophisticated statistical analysis techniques.

John Flynn and his associates were largely responsi- ble for popularizing these new "multidimensional" methods and for integrating them into an analysis protocol that is still being used today.

Correlational studies

Numerous investigators have examined relationships between subjective judgements and physical aspects of the luminous environment such as: (a) light level;-lo (b) spatial distribution of light, 2v"-18 (c) light source color 23 4. and; (d) lamp color

rendering.21-24

Taken together, these studies gave an indication of preferred light levels in office settings and of some psychological effects associated with changes in lamp color rendering. However, less agreement exists re- garding the subjective effects of lamp color temperature and the effects of different luminance ratios throughout the visual environment.

While these investigators met with varied success in their efforts to relate subjective impressions to variations in measured physical qualities of the luminous environment, the correlation approach gradually fell out of favor for at least two reasons. First, higher-order interaction effects between variables were not accessi- ble using this method. Studies were limited to the simultaneous manipulation of three or four independent variables at most. If more independent variables are studied, interpretation of all the possible relationships that might exist is difficult if not impossible. Rowlands, Loe, Macintosh, and ~ a n s f i e l d " faced just this difficulty in their ambitious attempt to relate subjective impressions to variations in measured physical

characteristics of real and realistic test offices. Their subjects judged a series of office spaces on 19 subjective dimensions. The subjective judgements were then correlated with 18 different physically measured characteristics of the different offices. Not surprising- ly, interpreting the resultant tables of correlation matrices was difficult. Hence, more complex effects are inaccessible with the correlational method.

Second. the subiective dimensions investigated in

_.,

-

a

hi^, Zndit&JbrReseamh in c~tu~ructiotl, N M R ~

council

~ ~all but the Rowlands, et al;5 study, (e.g., brightness, ~ ~ ~

of Canda, Otinzua, Ontatio preference) did not satisfy researchers and designers

Reprinted from the Journal of the IES Vol. 19, No. 2 with the permission of the Illuminating Engineering Society of North America.

(4)

who were interested in more complex aesthetic influences of different lighting designs. There was no guarantee that either perceived brightness o r preference were relevant to the everyday experience of different lighting schemes. Even if they were, it was at least implicitly assumed by the design community that any aesthetic influences of lighting would involve more than differences along these dimensions.

The development of the semantic differential seal- ing techniquCz6 multidimensional scaling analysi~,2~ and factor analysis, combined with readily available computer programs to perform these more complex analyses, provided investigators with the analytical tools to address these meatier issues.

Multidimensional studies

Semantic differential scaling was originally developed during the 1950s by Charles Osgood and his colleagues at the University of Illinois as a tool for measuring word and concept meaning. Semantic differential scales are bipolar, seven-category rating scales. The ends of each rating scale are defined by polar opposite adjectives (e.g., good-bad, large-small, spacious-cramped, hazy-clear, etc). People are asked to rate a variety of stimuli using each of a series of scales. Rather than assessing objects or environments along one o r two isolated scales (e.g., perceived brightness), lighting researchers could now examine the effects of different lighting schemes along many psychological dimensions.

Multidimensional scaling?' factor analysis,29 and special applications of multiple linear regressiong0 were the statistical methods used to analyze these data. They are families of statistical techniques used to describe patterns of similarity among groups of stimuli. Lighting researchers use them to determine which lighting schemes cue similar impressions. In- formation about the perceived similarities among the different lighting schemes is then used to formulate hypotheses about the physical characteristics of different luminous spaces that lead to the observed patterns.

Although several investigators have used the semantic differential, multidimensional scaling, factor analysis, and special applications of multiple linear regression to assess the psychological effects of different lighting schemes, John Flynn's work has been the most infl~ential.~l-~' Flynn's influence owes partly to his integration of these different techniques into a comprehensive analysis protocol. Perhaps more im- portantly, Flynn's work established a series of recom- mendations used by the design c o r n m ~ n i t y , ~ ~ ~ ~ which have even received the imprimatur of the IESNA.42

Concurrent with Flynn and his colleagues, several other investigators conducted multidimensional

studies of the psychological effects of

These investigators were less successful than Flynn and his colleagues in identifying subjective dimensions along which different lighting schemes varied.

The promise in early multidimensional scaling studies of relating variations in the luminous environment to complex psychological effects has remained largely unfulfilled. This was because the link between subjective responses and measured luminous characteristics of spaces was difficult to achieve. For example, the subjective impressions identified by Flynn were not related to detailed independent measures of the physical characteristics of the rooms he studied. In his studies the only definition of overhead lighting versus peripheral lighting was a subjective judgement.

Even when extensive physical measures were collected, investigators did not always relate variations in the physical characteristics of spaces to subjective ef- f e c t ~ . ~ " ~ ~ ~ While these investigators demonstrated that perceived differences among spaces could be ac- counted for by selected measures of illuminance, they did not relate the illuminance measures to any subjective effects that were cued by changes in illuminance.

What is needed is a bridge between the physical measures, or combinations of measures, and the different subjective impressions cued by the luminous stimulus. Only when the relationships among specific subjective effects and variations in physical qualities have been established can variation in objectively measured qualities be used to predict subjective effects. Unfortunately, such predictive power is not available to lighting designers who currently are guid- ed by Flynn's workg6 413 42 or their own intuiti0n.4~

The continued use of semantic differential scaling is also problematic, given recent questions about the validity of these procedures.47-50 Investigators would be well advised to identify and develop several measures that converge to supplement the data provided by semantic differential scaling. Converging operations51 testify to the robustness of inferred variables, in this case, subjective states. One strategy might involve the use of behavioral measures as com- plements to self-report data. For example, activity level might constitute a converging measure for the relaxed-tense semantic differential scale

Although studies investigating the behavioral influences of lighting are rare, we will briefly consider this work, before speculating on directions for future research.

Behavioral influences of lighting

Light serves as a powerful regulator of circadian rhythms in humans and other diurnal animals.52 Research on the behavioral effects of lighting sup- ports the notion that 1ighting.can be used to cue atten-

(5)

tion, orientation, and wayfinding in humans?5 5"55and increase activity levels in

human^^^-^'

and other animal~.~'-~I However, these results should be inter- preted with caution for at least two reasons. First, the range of light levels studied has been small. In many studies the effects of only two or three light levels are studied. Second, the effects of light level have been shown to interact with other independent Sim- ple linear functions are often insufficient for describ- ing relationships among different environmental variables; effects may be facilitated or inhibited depending on the presence of other moderating factors.62 Hence, further studies using a wider range of light levels and additional independent variables are required before firm conclusions about the effects of light level on orientation, wayfinding, and activity level can be drawn. Prospects for future research

Accurate characterization and prediction of the psychological effects of lighting requires concentrated effort in at least three areas. First, the subjective measurement tools used in assessing the psychological effects of different lighting designs must be re- evaluated and improved. Second, application of state- of-the-art instrumentation is required to achieve a comprehensive and accurate specification of the luminous environment. Finally, the experimental procedures and statistical analysis techniques that have been used by different investigators in the past need to be combined into a single experimental protocol, and supplemented with some newer techniques, to successfully relate psychological and behavioral effects to physical characteristics of spaces. Each of these requirements will be discussed in turn.

Subjective measurement tools

For the past 15 yrs., the semantic differential has been the measurement instrument of choice in research on the subjective influences of lighting, despite recent challenges to these procedures.47-m

Two crucial assumptions underlie the use of semantic differential scaling in lighting research. First, it is assumed that individual scales are used in a consistent manner across subjects and light settings. Second, it is assumed that changes in the subjective ratings are due to changes in the independent variables of interest, namely the different light settings.

Recent research suggests that neither of these assumptions are tenable. Although averaging semantic differential ratings across subjects can reduce the influence of random error, it can also hide important individual differences or erratic scale use, the ex- istence of which would cast doubt on the assumption that single scales are used in a consistent manner by different subjects. ~ e a ~ ' showed that mean semantic

differential ratings obscure great individual differences in scale use.

Rea's4' analysis also suggested that subjective responses collected in lighting experiments were not necessarily determined by changes in the experimental independent variables (i.e., light settings), as cor- relations between semantic differential ratings and ac- tual visual performance data were often weak. Hence, the semantic differential ratings made by Rea's subjects were not solely determined by changes in the light level or lighting geometry, which determined visual performance. Therefore, we cannot draw any conclusions about the subjective influences of lighting in Rea's4' experiment, because the lighting apparently did not consistently influence the semantic differential ratings. The factors that determined the subjective responses in this experiment remain largely unknown.

Tiller and Ream argued that individual differences in response scale use and unknown determinants of subjective responses are the result of inadequate definitions of the stimulus to be scaled in lighting experiments and of the response dimensions making up each semantic differential scale. The nature and effects of these two limitations are best explained by analogy to a more familiar measurement process, namely the measurement of objects using a meter stick. When several people measure different objects using a meter stick, their individual measurements of each object will usually agree because all the observers measure only the objects specified beforehand and consistently apply the meter stick to scale the same quality of different objects, namely spatial extent.

Obviously, measurement values would be useless if people were not told what specific objects to measure, but were instead given the vague general instruction to go out, measure something, and report back with their measurements. The instructions used in semantic differential scaling of lighting installations are often just as vague. Rarely is lighting defined explicitly as the stimulus to be evaluated. Without an explicit definition of lighting as the stimulus, there is no guarantee that different subjects are evaluating the same stimulus, or even if the stimulus being evaluated remains constant for individual subjects over time. Such data are as meaningless as a collection of measurements of unspecified and perhaps different objects.

Measurement values are also meaningless if observers do not use the meter stick consistently to scale spatial extent. If different observers do not consistently apply a measuring tool such as a meter stick, then the characteristics of the measured objects that were scaled are unknown. Investigators using the semantic differential assume that each individual

(6)

response dimension (eg., good-bad, bright-dim) will be used much like a meter stick, to consistently scale the same aspects of different stimuli.

However, Osgood and his colleagues found that the meaning of individual scales changed depending on the concept judged.*% This suggests that individual semantic differential scales are not always used as meter sticks to consistently scale the same features of different objects. Rather, the particular features measured by each semantic differential scale may change from object to object. Hence, semantic differential scaling does not always constitute an example of orthodox measurement. Orthodox measurement is defined as the assignment of numbers to objects following some consistent rule.64. 65

Careful selection of response dimensions that refer to salient and scalable aspects of the visual environment, pre-experimental training to highlight those aspects of the environment to which each response scale refers, and statistical analysis to identify incon- sistently applied scales will all be necessary to ensure that subjects consistently apply individual semantic differential scales to the same features of different stimuli.

Psychophysics is the measurement of physical characteristics of environments using behavioral responses from human observers. Semantic differential scaling is an ambitious form of psychophysics, however, it lacks the unambiguous response measures and rigorous stimulus control characteristic of more traditional psychophysical methods.6669 Nevertheless, direct application or adaptation of traditional psychophysical techniques to the study of psychological aspects of lighting would help improve definition of the stimulus and response dimensions.

The major difficulty that has faced investigators until now involves characterizing the luminous stimulus. Using conventional photometry, it could take months to specify the luminous characteristics of a space or scene.

Rea and his

colleague^^^^

71 have recently developed a computer-based luminance and image analysis system that allows for extremely rapid and precise measurements of luminance, contrast, and other aspects of the visual environment. These recent developments in computerized photometry make possible the rapid collection and specification of the luminous characteristics of a space. In this way, the influences of single and combined photometric measures on subjective impressions can be explored and modeled as never before.

Experimental protocols and analysis procedures No one investigator has yet combined all the steps necessary to link subjective impressions with physical

features of rooms. Each has concentrated on one aspect of the problem and neglected other equally important pieces of the puzzle. Early correlational studies, and other more recent British 43 44.45 focused on accurately specifying the photometric characteristics of the stimulus space, a painstaking endeavor subject to all the limitations imposed by conventional photometric techniques.

These investigators did not successfully relate observed physical variations to subjective effects. Flynn,31-40 on the other hand, concentrated on elucidating the nature of subjective effects and more o r less neglected to specify the luminous characteristics of the rooms he studied.

Future efforts must bridge these two areas. Multi- dimensional scaling and factor analysis will continue to be useful, as they provide information about the perceived similarities and differences among various lighting schemes. Extensive physical measures of the luminous characteristics of rooms are now possible using computerized photometry. The crucial step of relating the physical and subjective data can be achieved using either: (a) one of the more comprehensive property-fitting algorithms (eg., PROFIT^"^), or (b) canonical correlation analysi~.~" 74 These techniques are designed to relate one set of variables (i.e., subjective ratings) to a second set (i.e., physical measurements). Applying computerized photometry along with these analysis techniques will bridge the gap between subjective data and physical measures. Then, once relationships between measured physical characteristics and subjective reactions have been established, we can begin to vary the objectively measured qualities of the lighting in a space, and predict the subjective effects cued by these variations. Current research at IRC

The Institute for Research in Construction, Na- tional Research Council of Canada, has recently em- barked on a program of research to address these issues. Studies are being planned that will help develop an understanding of relationships that exist between luminous environments and occupant reactions, using computerized photometry in a new laboratory facility.

This new laboratory facility was designed by a steer- ing committee consisting of prominent designers, ar- chitects, engineers, and academics. The facility itself consists of four rooms furnished to represent "typical" North American, middle-management office stock. A wide range of lighting hardware (eg., lamps, luminaires, baffles, lenses, and diffusers) is available, providing great flexibility to explore the relationships among different lighting design strategies and occupant performance, perception, and subjective

(7)

reactions.

Each of the four offices in the facility is 12 by 15 ft., with an 8 ft concealed spline nonchamfered ceiling. Achromatic shades of gray were used on the interior finishes and furniture to prevent any changes in ap- parent brightness that might occur under light sources of different color.

Experiments are being planned to investigate the effects of the spatial distribution and intensity of light on occupant impressions of room spaciousness and brightness. Preliminary findings indicate, for exam.

ple, that a room illuminated with wall lighting will be judged as brighter and mom pleasant than another providing the same task illuminance from overhead. Although tentative, this finding suggests that it may be possible to achieve impressions of equivalent or greater room brightness with reduced lighting power density, thus identifying a source of potential energy savings. More extensive work along these lines will also deepen our understanding of the psychological aspects of lighting.

Acknowledgements

I would like to thank RG. Davis, M.J. Ouellette, and M.S. Rea for their helpful comments on earlier drafts of the manuscript. This paper is a contribution from the Institute for Research in Construction, National Research Council of Canada.

References

1. NEMAILRI. 1989. Lighting and human pmformance:

A reuiew. Washington, DC: National Electrical Manu-

facturers Association.

2. Bodmann, H.W. 1967. Quality of interior lighting based on luminance. J of the IES 32:22-40.

3. Bodmann, H.W. 1962. Illumination levels and visual performance. International Lighting Reoiezu 13:41. 4. Bodmann, H.W.; Sollner, G; and Voit, E. 1963. Bewertung von beleuchtungsniveaus bei verschiedenen lichtarten. Proc. CZE 15th Session :502-509.

5. Boyce, P.R. 1973. Age, illuminance, visual perfor- mance and preference. Lighting Research 01 Technology

5:125-144.

6. Boyce, P.R. 1981. Human factors in lighting. New York: Macmillan.

7.

Hughes, P.C., and McNelis, J.E 1978. Lighting, pro- ductivity and the work environment. LD + A 832-40.

8 Nemecek, J., and Grandjean, E. 1973. Results of an ergonomic investigation of large-space offices. Humn

Factors 15:lll-124.

9. Saunders, J.E. 1969. The role of the level and diversity of horizontal illumination in an appraisal of a simple office task. Lighting Research €9 Technology 1:37-46.

10. Smith, S.W., and Rea, M.S. 1979. Relationships between office task performance and ratings of feel- ings and task evaluations under different light sources and levels. Proc CZE 19th Session:207-211.

11. Bean, A.R., and Hopkins, A.G. 1980. Task and background lighting. Lighting Research €9 Technology 12:135-139.

12. Rea, M.S.; Ouellette, MJ.; and Tiller, D.K. 1989. The effects of luminous surroundings on visual performance, pupil size and human preference. J of the

ZES lg(no.2).

13. Roll, K.E, and Hentschel, H.J. 1988. Fulfillment of modern lighting requirements and stable perception. Paper presented at the 1988 IESNA Annual Con- ference, 6-12 August, Minneapolis, MN.

14. Roll, K.E, and Hentschel, H.J. 1987. Luminance patterns in interiors and balanced perception. Proc

CZE 21st Session:194-196.

15. van Ooyen, M.H.E; van de Weijgert; J.A.C. and Begemann, S.H.A. 1987. Preferred luminances in offlces. J of the IES 2:152-156.

16. van Ooyen, M.H.E; van de Weijgert, J.A.C.; and Begemann, S.H.A. 1986. Luminance distribution as a basis for office lighting design. National Lighting Con-

ference: Nottingham, U K : pp. 103-106.

17. Touw, L.M.C. 1951. Preferred brightness ratio of task and its immediate surroundings. Proc CZE 12th

Session.

18. Tregenza, f.R; Romaya,

S.M.;

Dawe, S.P.; Heap,

L.J.;

and Tuck, B. 1974. Consistency and variation in preferences for office lighting. Lighting ReJe~rch IY Techmhgy 6:205-212.

19. Cuttle, C., and Boyce, P.R 1988. Kruithof revisited: A study of people's responses to illuminance and color temperature of lighting. Lkhting in Australia 12:17-28.

20. Davis, R.G., and Ginthner, D. 1989. Light source color, light level, and subjective response: EvaIuating the Kruithof curve] of the

IES

1927-58.

21. Kanaya, S.; Hashimoto, K.; and Kichize, E. 19'79. Subjective balance between general color rendering index, color temperature, and illuminance of interior lighting. Pmc CZE 19th Session:274-278.

22. Aston, S.M., and Bellchambers, H.E. 1969. I1- lumination, color rendering and visual clarity.

Lighting Research & Technology 1:259-261.

23. Bellchambers, H.E., and Godby, A.C. 1972. 11- lumination, color rendering and visual clarity.

Lighting Research €9 Technology 4:104-106.

24. Boyce, P.R. 1977. Investigation of the subjective balance between illuminance and lamp color proper- ties. Lighting Research €9 Technology 9:ll-24.

25. Rowlands, E.; Loe, D.L.; McIntosh, R.M.; and Mansfield, K.P. 1985. Lighting adequacy and quality in office interiors by consideration of subjective assess-

(8)

ment and physical measurement CIE Journal 4:23-37. 26. Osgood, C.E.; Suci, G.J.; and Tannenbaum, P.H. 1957. The measurment of mean%. Urbana: University of Illinois Press

27. Coombs, C. 1964. A thomy ofdata. New York: John Wiley and Sons.

28. Schiiman, S.S.; Reynolds, M.L.; and Young,

EW.

1981. Inhodwtiun to multrdimmsional scaling: Themy, m e t m and applications. New York: Academic Press.

29. Kim, J.O., and Meuller, C.W. 1978 Introdwtiun to factor analysis: What it 8 and how to do it. Beverley Hills:

Sage.

30. Hays, W.L. 1988. Statistics, Fourth Edition. New

York: Holt, Rinehart & Winston.

31. Flynn, J.E. 197211973. The psychology of light Elac- trical Consultant 88:12-897 (series of eight articles).

32. Flynn, J.E. 1975. A study of lighting as a system of spatial cues EDRA.6 workshop on "The Psycho- logical Potential of Illumination? University of Kan- sas, April.

33. Flynn, J.E. 1977. A study of subjective responses to low energy and non~niform lighting systems.

W ) + A 7: 167-179.

34. Flynn, J.E., and Spencer, T.J. 1979. The effects of light source color on user impression and satisfaction. J of the IES 6267-179.

35. Flynn, J.E., and Subisak, G.J. 1978. A procedure for qualitative study of light level variations and system performance. J of the IES 8:28-35.

36. Flynn, J.E.; Segil, A.W.; and Steffy, G.R. 1988. Ar-

chztectural interior systemc Lighting, air conditiuntng, acoustics, Second Edition. New York: Van Nostrand Reinhold.

37. Flynn, J.E.; Hendrick, C.; Spencer,

T.J.;

and Mar-

tyniuk, 0. 1979. A guide to the methodology pro-

cedures for measuring subjective impressions in

lighting. J of the IES 995-110.

38. Flynn, J.E.; Spencer, T.J.; Martyniuk, 0.; and Hen.

drick, C. 1973. Interim study of procedures for in. vestigating the effect of light on impressions and behavior. J of the IES 387-94.

39. Flynn, J.E., Spencer, T.J., Martyniuk, Q, and Hen.

drick, C. 1975. The influence of spatial light on human judgement. Proc CIE 18th Session:39-46.

40. Flynn, J.E., Spencer, T.J., Martyniuk, O., and Hen.

drick, C. 1975. Interim report: The effect of light on

human judgement and behavior. IERI Project 9 2 (Un- published monograph).

41. StefFy, G.R. 1988. Office Lighting Seminar. Ot-

tawa, Ontario, Canada: IESNA National Capital Sec- tion Education Committee.

42. Kaufman, J.; and Christensn, J.; eds. 1981. IES lighting handbook: 1981 Applicatum uolzlme. New York: I1- luminating Engineering Society of North America.

43. Hawkes, R.J.; Loe, D.L.; and Rowlands,

E.

1979. A

note toward the understanding of lighting quality. J of the IES 8:111-120.

44. Kimmel, P.S. and Blasdell, HG. 1973. Multidi- mensional scaling of the luminous environment. J of the IES 3:113-120.

45. Stone,

m.;

Parsons, KC.; and Harker, S.D.P. 1975.

Subjective judgements of lighting in lecture rooms. Lighting Research and Technology 7:259-260.

46. Farber-Fleming, P. 1988. Designer melds the sacred and profane: Two sets of design criteria meet

LD + A 18:12-53.

47. Poulton, EC. 1975. Observer bias. Applaed Elgo- nomics 6: 3-8.

48. Poulton, EC. 1979. Models for biases in judging sensory magnitude Psychological Bulletin 86377-803.

49. Rea, M.S. 1982. Calibration of subjective scaling responses. Ltghting Research & Bchnology 14121-129.

50. Tiller, D.K., and Rea., M. 1989. Prospects for semantic differential scaling in lighting research. Presented at the CIBSE National Lighting Conference

1990, Cambridge, UK.

51. Garner, W.R.; Hake, H.W.; and Erikson, C.W. 1956. Operationism and the concept of perception. The Psychological Revinv _62149-159.

52. Coleman, R.M. 1986.

Wide

awoke at 3:00 A.M.: By

choice m by chance? New York: W.H. Freeman and Company.

53. LaGuisa,

RE,

and Perney, L.R. 1974. Further

studies on the effect of brightness variations on atten- tion span in a learning environment. J of the IES 4249-252.

54. Taylor, L.H., and Socw, E.W. 1974. The move. ment of people towards lights. J of the IES 4237-241.

55. Yorkq P., and Ginthner, D. 1987. Wall lighting placement: Effect on behavior in the work environ.

ment. L D + A 17:30-39.

56. Delay, E.R., and Richardson, M.A. 1981. Time estimation in humans: Effects of ambient illumina- tion and sex. Pmceptual and Motor Skills 53747-750.

57. Kallman, W.M. and Isaac, W. 1977. Altering arousal in humans by varying ambient sensory condi- tions. Perceptual and Motor Skills 4419-22.

58. Sanders, M.; Gustanski, J.; and Lawton,

M.

1974.

Effect of ambient illumination on noise level of groups Journal of Applied Psychology 59527-528.

59. Alexander, M., and Isaac, W. 1965. Effect of illumination and d.amphetamine on the activity of the rhesus macaque Psychological l?e@rts 16311-313.

60. Isaac, W. 1969. The influence of illumination upon the temporal patterning of responses in squirrel and owl monkeys P s y h o m i c S c i m e 14:243-244.

61. Isaac, W., and Reed, WG. 1961. The effect of sen. sory stimulation on the activity of cats. J n l of Com. paratzue and Physwlogrcal Psychology 54677-678.

(9)

the effect of environmental stressors upon perfor. mance Psychological Bulletan 72260-272.

63. Shaw, D.R. 1955. Variation in inter-rcale cmrelath

on the semantic dzfferential as a function of the concept

judged. Unpublished masters thesis, University of 11. linois at Urhana.

64. Campbell, N.R. 1938 Symposium: Measurement and its importance for philosophy. Aristotelian Society

Supplement 17 London: Harrison and Sons.

65. Stevens, S.S. 1946. On the theory of scales of measurement. Science 103:677-680.

66. Gescheider, GA. 1976. Psychophysicst method and

theory. Hillsdale, N.J.: Lawrence Erlbaum & Associates. 67. Gescheider, G.A. 1988 Psychophysical scaling.

Annual Review of Psychology 39:169-200.

6 8 Stevens, S.S. 1958. Problems and methods of psychophysics. Psychologrcal Bulletin 55:177-197.

69. Stevens, S.S. 1975. Psychophysics: Introduction to zts

perceptual, neural, and social pospects. New York: John

Wiley & Sons

70. Rea, M.S. and Jeffrey, 1.J. 1990. A new luminance

and image analysis system for lighting and vision: I.

equipment and calibration. J of the IES 19:64-72.

71. Kambicb, DG., and Rea, M.S. 1987. New Cana- dian lighting analysis system. Lighting 1:16-17.

72. Chang, J.J., and Carroll, J.D. 1968. How to usePRO-

FI?; a computer program for property fitting by optimising

nonlinear or l m a r currelatiom. Unpublished

manuscript, Bell Laboratories, Murray Hill, NJ. 73. Levine, M.S. 1977. Canonical analysis and factor com-

parrra. Beverly Hills: Sage.

74. Thompson, B. 1984. Canonical correlation analysis. Beverly Hills: Sage.