the s c i en c e of health promotion methods, issues, and

THE S C I EN C E OF HEALTH PROMOTION

Methods, Issues, and Results in Evaluation and Research

The Conceptualization and Measurement ofPerceived Wellness: Integrating BalanceAcross and Within DimensionsTroy Adams, Janet Bezner, Mary Steinhardt

Abstract

Purpose. The impact of individual perceptions on health is well-established. Howev~no valid and reliable measure of individual wellness perceptions exists. Therefore, the pur-pose was to introduce a measure called the Perceived Wellness Survey (PWS).

Design. Convenience sampling facilitated recruitment of a sample large enough to per-form factor analysis with adequate power (. 85). The appropriateness of factor analysis supported by Bartlett’s test (X2 = 711 O, p <- . 01) and the Kaiser-Meyer-Olkin measure ofsampling adequacy (. 91).

Setting. The sample (n = 558) was composed of 3M Inc. employees from multiple sites Austin, Texas (n = 393); employees from MuRata Electronics, Inc., College Station, Penn-sylvania (n = 53); and students enrolled at the University of Texas at Austin (n = 112).

Subjects. Racial, gend~ and age distribution was, respectively, 6.3% AJi~can-Ameri-can (n = 35), 8.2% Asian (n = 46), 73.3% Caucasian (n = 409), 9.5% Hispanic(n --53), and 2.7% other (n -- 15); 47.8% male (n = 267), and 52.2%female 291); and 36.8 years.

Measures. Measures included the Perceived Wellness Survey, and two additional ver-sions of the Perceived Wellness Survey designed to measure both discriminant and face va-lidity. Perceived Wellness Survey subscales include physical, spiritual, intellectual, psycho-logical, social, and emotional dimensions.

Results. All subscales were correlated (p <- . 05) with the Perceived Wellness Surveycomposite and zoith each oth~ Evidence of internal consistency (o~ = . 88 to . 93), anddiscriminant, face, and factorial validity was provided. Finally, the Perceived WellnessSurvey appears to be a unidimensional scale.

Conclusion. The unidimensional nature of the Perceived Wellness Survey suggests thatperceptions of wellness in various dimensions are intertwined by their affective nature. ThePerceived Wellness Survey appears to be reasonably valid and reliable; hozoev~ further re-search is needed. (Am J Health Promot 1997;1113]:208-218.)

Key Words: Wellness, Wellness Measurement, Wellness Theory, Wellness Mod-els, Perceptions, Mind/Body

Troy Adams, PhD, is an Assistant Professor at Oklahoma State University. Janet Bezn~PhD, UF is an Assistant Professor at Southwest Texas State University. Mary A. Steinhardt,EdD, is an Associate Professor at the University of Texas, Austin. Many of the concepts inthis paper were formulated while the first and second author were students at the Universityof Texas, Austin.

Send reprint requests to Troy Adams, PhD, Oklahoma State University, School of Health, PhysicalEducation, and Leisure, Coh,in Center 102, Stillwater, OK 74078.

This ma~useHpt was submitted for publicatio±~ October 1 O, 1995," re~dsio~s ~aet~ t~quested December 27, 1995; themanuscript was accepted for publication May 14, 1996.

A m J Health Promot 1997;11(3):208-218.Copyright © 1997 by American Journal of Health Promotion, Inc.0890-1171/97/$5.00 + 0

We see the worm not as it is, butas we are.

--H.M. Tomlinson

INTRODUCTION

Unexplainable phenomena suchas the placebo effect and diseasesthat spontaneously go into remissionsupport the notion that many factorswhich influence health are simplyunknown. And, it is increasingly evi-dent that we do not yet possess all ofthe tools to fully describe and predicthuman health--in particular, positivehealth or wellness. Many modelshave been developed in an effort tobetter understand the naturally oc-curring variability in health. Most re-cently, Wilson and Cleary~ integratedseveral components including biolog-ical and physiological variables, symp-tom status, functional status, andgeneral health perceptions, amongothers. Regarding perceptions ofhealth they stated, "They representan integration of... health concepts.They are among the best predictorsof the use of general medical andmental health services" (62). Thus,how we "see the world" appears topowerfully impact our health andwellness.

The presumed power of percep-tions raises a challenge for research-ers and practitioners alike. How canwe sanctimoniously apply absolutestandards of wellness in a relativeworld? Dunnz constrained his origi-nal definition of wellness to the indi-vidual environment, suggesting thatthe experience of wellness is uniqueto each individual. -~.4 The reality isthat individuals process and interpret

208 American Journal of Health Promotion

information from internal and exter-nal sources in highly variable ways.This variation can either be viewed asuncontrollable, residual error due toindividual differences, or as a richsource of information about influ-ences upon health and wellnesswhich remain unexplained by otherindicators. Thus, the study of howwellness perceptions fit into an over-all model of health would be a posi-tive contribution.

The study of perceptions is empir-ically well-supported by other bodiesof research. Social support research-ers have suggested that perceivedsupport has a powerful influenceupon health. 5-9 Stress researchershave indicated that a tension-produc-ing stimulus elicits the stress re-sponse only if it is perceived as threat-ening.1°-13 Finally, epidemiological re-searchers have concluded that self-rated perceptions of health areamong the most powerful predictorsof subsequent health outcomes.~-v~

In addition, health perceptions havebeen identified as one of the strong-est predictors of physical and mentalhealth care utilization. 2° Thus, whilethe influence of standard risk factorscannot be ignored, individual per-ceptions are also important becausethey may actually precede overt man-ifestation of illness or wellness andmay therefore be fertile ground forearly intervention or enduring cele-bration, respectively.

The purpose of this paper isthreefold. First, to construct a philo-sophical foundation which provides acontext in which the conceptualiza-tion and measurement of perceivedwellness can be discussed. Second, tointegrate empirically supportable di-mensions of wellness into a wellnessframework. Third and primarily, tointroduce a conceptually congruent,empirically sound wellness measurewhich rests upon the philosophicalfoundation and which is supportedby the theoretical framework.

PHILOSOPHICAL FOUNDATION

IntroductionSince Dunn, many have conceptu-

alized wellness.21-97 All of these con-ceptualizations contain convergentand repeated wellness themes that

will be discussed later. In addition,much philosophical and theoreticalsupport for the overall wellness con-struct has been derived from relatedtheories.

Systems TheoryAccording to systems theory, each

part of a system is both an essentialsubelement of a larger system and anindependent system with its own su-belements.2s--~° Elements are recipro-cally interrelated such that disrup-tion of homeostasis at any level re-quires adaptation of the entire sys-tem.28,3° Dunn stated that individualwellness involves "an integrated .method of functioning" suggestingreciprocal integration. 2 At the indi-vidual level, this implies simultaneousfunction in multiple dimensions andat various levels within these dimen-sions including the physical, spiritual,psychological, social, emotional, andintellectual. 3,4,29--~1 The multidimen-sionality of wellness is supported byseveral authors.3,4,21,22,24,25,33-35

To best describe and predict indi-vidual wellness, models should in-clude several dimensions which areoperationalized and interpreted con-sistent with the systems approach.Specifically, the wellness magnitudewithin each dimension and the bal-ance among them should be simulta-neously considered) In addition, valid wellness model should either in-clude cultural, organizational and en-vironmental factors, or be connecta-ble to models that include these fac-tors.28.30

Salutogenic OrientationAntonovsky contributed to the

widespread use of the term "saluto-genie," which simply means "healthcausing."a°.~l Salutogenesis is alsosuggested in the World Health Orga-nization (WHO) definition of healthas: "a complete physical, mental, andsocial well-being and not merely the ab-sence of disease..."-~ ( I, italics add-ed) and Dunn’s definition of well-ness as being "oriented toward maxi-mizing the potential of which the indi-vidual is capable’’2 (4-5, italicsadded). It is evident that wellness iswidely recognized as the conceptualanchor of a salutogenic orienta-tion. 4’2w~,2~ Yet a heavy emphasis is

placed on the detection, treatment,and prevention of disease in what isoften called "wellness practice." Thisis probably attributable to the avail-able selection of measurement tools(e.g., skinfold calipers, blood pres-sure equipment), which are only ca-pable of detecting disease risk factorsor the lack thereof. In addition, re-search that is limited to a pathogenicperspective due to the use of suchmeasures is sometimes mislabeledhealth promotion instead of diseaseprevention research.

By measuring wellness perceptionswhich typically precede observablesymptomology, practitioners and re-searchers could focus on the saluto-genie pole of each dimension repre-sented by the perimeter of Figure 1.This model is similar to those pre-sented elsewhere4,2~,~,~6 except that itexplicitly incorporates vertical andhorizontal directions. Vertical move-ment occurs between the illness andwellness poles, whereas horizontalmovement is the dynamic, balance-seeking force along each dimensionof wellness. In summary, salutogenicpractice should emphasize optimalhealth and balance. In addition, thefocus of salutogenic research shouldbe to determine what factors are as-sociated with being well.

Framework SummaryPerceived wellness is a multidimen-

sional, salutogenic construct, whichshould be conceptualized, measured,and interpreted consistent with anintegrated systems view. What types ofwellness indicators and measures areavailable and how well are they sup-ported by this foundation?

Measurement of Wellness. Indicators ofhealth and wellness have includedclinical variables such as blood pres-sure and cholesterol; physiologicalvariables such as VO2 max and mus-cular strength; and behavioral vari-ables such as smoking and dietaryhabits. Practitioners have reliedheavily on clinical, physiological, andbehavioral measures to plan individu-al and community interventions, andto predict various health outcomes?Although these types of variables arevaluable indicators of bodily wellness,they provide little information about

January/February 1997, Vol. 11, No. 3 209

Figure 1

The Wellness Model

Wellness

Illness

The top of the model represents wellnessbecause it is expanded to the fullest possibleextent, whereas the tightly constricted bottomrepresents illness. In between areinnumerable combinations of wellness inseveral dimensions and the various states ofbalance among them. The lines which extendfrom the inner to the outer circle indicate thepossibility of bidirectional movement alongeach continuum. Movement in everydimension influences and is influenced bymovement in all other dimensions. Forexample, in extreme wellness conditions, oneor more dimensions expand and place an"outward wellness force" on each of the otherdimensions. In contrast, in extreme illnessconditions, one or more dimensions contractand cause either compensatory orconcomitant change in each of the otherdimensions,

the wellness of the mind. Perceptualmeasures have been used to predicteffectively a variety of health out-comes.16,xs,’~°,37 Hence, valid perceptu-al measures could complement body-centered indicators of wellness andprovide both researchers and practi-tioners with important informationabout the persons whom they serve?

Wellness-related perceptual con-structs include the following: psycho-logical well-being) ~4° mental well-be-ing, 41,42 subjective well-being,4~-45 gen-eral well-being, 46,47 morale,4s~9 happi-ness,5°,51 life satisfaction,52-54 andhardiness.1"-’,55 Many of these have

been used as wellness measures inspite of the fact that none were origi-nally designed as such. Moreover, ex-tant wellness scalesm,~4,26 were primar-ily constructed as teaching tools andwere not originally validated, al-though t~ro 24’26 have since demon-strated evidence of reasonable inter-nal consistency)6 In addition, allthree2L24,26 contain a mixture of per-ceptual and behavioral items; thus,the unique contribution of percep-tions to wellness cannot be deter-mined from these scales.

Anecdotal evidence supportingthe validity of the perceptual ap-proach to wellness assessment isabundant in the lay literature,~7-’~9

but has not received extensive empir-ical attention .because a theoretically-based, empirically sound measure ofperceived wellness is lacking. Theprimary purpose is therefore to de-velop such a measure called the Per-ceived Wellness Survey.

REVIEW OF LITERATURE

IntroductionMany different wellness dimen-

sions have been identified includingthe physical, spiritual, and intellectu-al;4,2~-26,~ L~4 ¯ ~lpsychological;- socialand eIllotional; 4,23-26,31 occupation-al, 24’26 and community or environ-mental.2~,24,-~1 Six dimensions were se-lected for inclusion in the PerceivedWellness Survey based on thestrength of theoretical support andthe quality of empirical evidence sup-porting each. All six dimensions aredefined and supported below.

Physical WellnessPhysical wellness is defined as a

positive perception and expectationof physical health. Stewart and othersstated that measuring physical healthperceptions is important because itintegrates available health informa-tion by accounting for differences inhealth preferences, values, needs,and attitudes. ")° Mossey and Shapirofurther suggested that measurementof perceived health can be combinedwith objective ratings to provide amore accurate interpretation of pa-tient conditions. Is In their seminalstudy, subjects with poor perceivedhealth had a risk of mortality three

times greater than subjects with goodperceived health. Furthermore, sub-jects with good/excellent objectivehealth but poor perceived health hada greater risk of death than subjectswith poor/fair objective health butexcellent perceived health. These re-suits have been supported else-where.~,~7,19 Good perceived healthhas also been positively associatedwith higher levels of physical activityand negatively associated with muscu-loskeletal symptoms and diseases,and psychosocial problems.14

Spiritual WellnessSpiritual wellness has been de-

fined as a belief in a unifying force,4

an integrative force between themind and body,6° or as a positive per-ception of meaning and purpose inlife. -%2~,6~ Of these, the latter is themost empirically supported and hasbeen associated with positive healthoutcomes and well-being.62m4

Paloutzian and Ellison developeda measure to tap the life purposeconstruct called the existential well-being scale.6"~ In limited researchsince its development, the life pur-pose construct as measured by theexistential well-being scale has dem-onstrated negative associations withloneliness 66 and depression67 andpositive associations with self-esteem;family togetherness; and socialskills, 6° coping beliefs, 67 and connect-edness.68

Psychological WellnessPsychological wellness is defined

here as a general perception thatone will experience positive out-comes to the events and circumstanc-es of life. This definition refers to apsychic resource called dispositionaloptimism.69 Aaa individual who is dis-positionally optimistic believes thatevery situation and circumstance willultimately produce positive out-comes. Optimism, as measured bythe Life Orientation Test, has beenpositively correlated with hardiness,general well-being, 7° happiness,71 like-lihood of completing a program foralcoholics,72 post-surgical recovery,7~

and quality of life appraisal in survi-vors of coronary bypass surgery. TM Inaddition, optimism has been nega-tively associated with anxiety,75 escape


avoidant coping,7~ and various mea-sures of distress.76

Social WellnessSocial wellness is defined as the

perception of having support avail-able from family or friends in timesof need and the perception of beinga valued support provider. Social sup-port has been the dominant themein social wellness research.6,s In sever-al prospective studies of the associa-tions between social support and ei-ther morbidity or mortality, men withhigh levels of support had fewer riskfactors and symptoms of cardiovascu-lar disease and had lower mortalityrates after other risk factors wereheld constant. 77-sl Some have sug-gested that social support is also pro-tective for women,~z,~ but othershave disagreed.7s,s4 Social supporthas been positively correlated withphysical and psychological well-be-ing~5 and overall life satisfaction ss butnegatively correlated with distresssymptoms and psychopathology.7

Researchers have identified fourkey associations between social sup-port and health. First, the perceptionof available support is the most im-portant health protecting feature.5,6,9

Second, the quality of available sup-port is more important than thequantity. 6,s6,87 Third, support fromfamily and friends varies in impor-tance depending on situational sup-port needs,s Fourth, the support rela-tionship is healthiest when it is recip-rocal.9,77

Emotional WellnessEmotional wellness is defined as

possession of a secure self-identityand a positive sense of self-regard,both of which are facets of self-es-teem. Self-esteem is a major compo-nent of emotional wellness44 and isone of the strongest predictors ofgeneral well-being,s8-9° Self-identity isconceptualized as one’s internal im-age of self. The value placed on self-identity is called self-regard and hasbeen defined as "the extent to whichone prizes, values, approves, or likesoneself’’01 (115). An individual with focused sense of self-identity tends topossess a higher self-regard4’~ and ismore capable of meaningfully inter-preting discrepant information.1°

Likewise, a person with a high self-re-gard interprets situations and eventsin ways that preserve and reinforceself-identity. Researchers have indicat-ed that self-esteem is positively associ-ated with principle-centeredness, in-ternal wellness orientation, physicalself-esteem, and physical activity,9"

and negatively related with body dis-satisfaction and restrained eating.9s

Intellectual WellnessIntellectual wellness is defined as

the perception of being internallyenergized by an optimal amount ofintellectually stimulating activity. Re-searchers have suggested that intel-lectual overload and underload canadversely affect health.1°,94,95 Thus,moderate amounts of intellectuallyenriching activity are optimal.95-97

In a related vein, Langer and Ro-din conducted a study to assess theimpact of involving nursing homeresidents in decisions regarding theirown care. 98 This simple interventionwas reinforced by providing the ex-perimental group with a plant totend. Cross-sectionally, subjects in theexperimental group felt more in con-trol and were more alert and active.~8

In a longitudinal follow-up, residentsin the experimental group were"judged to be significantly more ac-tively interested in their environ-ment, more sociable and self-initiat-ing, and more vigorous than resi-dents in the control group’’9~ (899).In addition, experimental group resi-dents were more active, healthier,and had a lower mortality rate.These findings are supported bythree additional studies of elderlypopulations. 100-102

SummaryEmpirical support for each of the

dimensions included in the wellnessmodel (Figure 1) has been provided.Efforts to validate the internal struc-ture of the wellness model and thecorresponding scale, the PerceivedWellness Survey, are described below.

METHODS

DesignA convenience sampling method

was used to recruit participants. Thismethod facilitated recruitment of the

sample size necessary to perform afactor analysis with adequate power.The computed power was .85. Theappropriateness of factor analysiswith this sample is highly indicatedby Bartlett’s test of sphericity (X2 =7110, p --< .01) and the Kaiser-Meyer-Olkin measure of sampling adequacy(.91).1°~ The primary variable of in-terest was the Perceived Wellness Sur-vey. Other measures included two ad-ditional versions of the PerceivedWellness Survey designed to assess,respectively, discriminant and face va-lidity of the Perceived Wellness Sur-vey. All three of these measures aredescribed in greater detail in theMeasures section.

SampleParticipants included employees of

two 3M Inc. sites in Austin, Texas (n= 295, administrative center; n = 98manufacturing plant); employeesfrom MuRata Electronics, Inc., Col-lege Station, Pennsylvania (n = 53);and students from a health educa-tion class at the University of Texasat Austin (n = 112). The racial mixof the sample was, for the combined3M samples, 7.4% African-American(n = 29), 11.2% Asian (n = 44),68.4% Caucasian (n = 269), 10.7%Hispanic (n = 42), and 2.3% other(n = 9); and for the MuRata sample,2% Asian (n = 1), 87% Caucasian = 46), and 11% other (n = 6). Racewas not included as a variable in thestudent sample, but based on esti-mates from the course instructor, themix for the student sample was ap-proximately 5% African-American,1% Asian, 84% Caucasian, and 10%Hispanic. The samples were dividedwith respect to gender as follows:3M--61% male (n = 240), 39% fe-male (n = 153); MuRata--11% male(n = 6), 89% female (n = 47); students--19% male (n = 21), 81%female (n = 91). The mean ages forthe samples were 3M--41.1 -+ 10.8,MuRata--34.34 _+ 10.36, and stu-dents-23.15 -+ 5.4. All data weregathered between August 1994 andMarch 1995.

3M employees were provided theopportunity to participate in thestudy during an annual healthscreening. The overall health screen-ing response rate in the 3M Austin


population (n = 1800) was 28%. those who attended the healthscreening (n = 503), 78% (n = 393)both agreed to participate in thestudy and provided useable data.Those included in the sample werenot significantly different (p -< .05)from the health screening attendeesomitted from the sample (n = 110)with respect to age or gender. Non-participating health screening atten-dees were more likely to be Asianand less likely to be Caucasian (p -<.05). Demographic data were unavail-able for employees who did not par-ticipate in the health screening.

Health screenings are available toMuRata employees throughout theyear on a rotating schedule. Becausethe Perceived Wellness Survey wasadministered on a pilot basis, onlythose MuRata employees who com-pleted a health screening in Marchof 1995 were given the opportunityto participate (n = 90). Thus, the re-sponse rate of the MuRata sample (n= 53) was 59%. The nonrespondentswere not significantly different fromrespondents with respect to age,race, or gender. Demographics onthe total MuRata population duringthe screening period were unavail-able.

All but a few of the students chose

to participate in the study. Nonparti-cipants were more likely to be malebut were not significantly different inany other way.

Prior to participation, the 3M andstudent participants had the opportu-nity to read and sign an informedconsent document approved by aninstitutional review board for re-search with human subjects. MuRataparticipants similarly provided indi-vidual informed consent according tothe guidelines established by an in-ternal review committee of the com-pany responsible for data collection,LifeQuest, Inc., of Memphis, Tennes-see. All participants were assured thattheir decision whether or not to par-ticipate would have no effect on fu-ture relations with the researchersnor with the respective organizations.Students were assured that their par-ticipation status would not influencetheir final grade in the class.

MeasuresCompletion of psychometric in-

struments, which included the Per-ceived Wellness Survey in all foursamples, required 20 to 40 minutesand, in every case, took place in aquiet setting. Additionally, fourhealth professionals employed by oneof the companies completed a dis-criminant validity version of the Per-ceived Wellness Survey and studentscompleted a form of the PerceivedWellness Survey designed to assessface validity.

Perceived Wellness Survey. The Per-ceived Wellness Survey is a salutogen-ically-oriented, multidimensionalmeasure of perceived wellness per-ceptions in the physical, spiritual,psychological, social, emotional, andintellectual dimensions. Sample itemsfront each dimension are, respective-ly, "I expect to always be physicallyhealthy, .... I believe there is a realpurpose for my life," "In the past, Ihave expected the best," "My friendswill be there for me when I needhelp," "In general, I feel confidentabout my abilities," and "In the past,I have generally found intellectualchallenges to be vital to my overallwell-being." Each dimension is repre-sented by six items which are scoredfrom 1, "Very strongly disagree" to6, "Very strongly agree." The dimen-sional scores are integrated by com-bining the magnitude or mean ofeach dimension with the balance orthe standard deviation among dimen-sions into a wellness composite score.In mathematical terms, the wellnesscomposite score equals the sum ofthe subscale means divided by thesum of the standard deviation amongthe subscale means and 1.25. The ad-dition of 1.25 to the denominatorprevents a rare but statistically possi-ble deviation of "0" from nullifyingthe wellness composite score. Copiesof the scale and detailed scoring pro-cedures are included in the appen-dix. The Perceived Wellness Surveyconstruction process is briefly dis-cussed below.

Initially, a total of 69 content-relat-ed items from six separate scales (seeTable 1) were combined to form thePerceived Wellness Survey, which waspiloted several times. Included in this

pool were items which tapped per-ceptions of physical health, sense ofmeaning and purpose in life, posidveexpectancies, self-identity and self-rergard, and social support both re-ceived and provided. The two socialsupport scales were consolidated intoone, reducing the number of originalscales to five, but an additional scalecreated by the authors was later add-ed; thus, the final number of subs-cales was six. Three item reductionsschemes were employed. First, anitem correlation matrix was exam-ined to determine whether any setsof items were redundant (r -> .70).Redundant sets or pairs of itemswere reduced to the single best item.Second, the magnitude of the item-to-total-scale correlation was consid-ered. Items with coefficients smallerthan .40 on the total scale were ex-cluded. Third, all items were re-viewed to determine the degree ofcontent match between the itemsand subscale definitions. After the sixbest items were selected to rewesentthe physical, spiritual, psychological,emotional, and social dimensions, sixitems written by the authors wereadded to represent the intellectualdimension. In addition, a few of theitems were revised to add clarity andconsistency to the subscales. Ulti-.mately, six items for each of the sixdimensions were included, giving thePerceived Wellness Survey-a total of36 items.

In an attempt to minimize itemorder effects, the dimension orderwas randomly shuffled, creating sixblocks. The items were then placedinto each block so that each dimen-sion was represented by every sixthitem and so that the 21 positive and15 negative items were spread evenlythroughout.

To demonstrate the general psy-chometric soundness of the scalesfrom which Perceived Wellness Sur-vey items were derived, reliability andvalidity coefficients for each parentscale are reported in Table 1. Con-structing a new scale by combiningitems from several existing scales isan acceptable practice; ~°4 however, byremoving items from the originalscale, the psychometric context is al-tered. Thus, pilot research was con-ducted to determine the psychomet-


Table 1

Sources of Items for the Perceived Wellness Survey

Num-berof

ItemsUsed/Num-ber

Internal Type of ofConsistency, Validity/ ItemsReliability, Coefficient, in

Dimension Scale Reference Reference Scale Emphasis

Physical MOS-36 (~ = 0.89-0.9111° Convergent r 0.9611o

Spiritual Existential (x = 0.786s Convergent r =well-being 0.31-0.686~

Psychological Life Orienta- (x = 0.76s9 Convergent r =tion Test 0.34-0.5569

Social Perceived so- e = 0.84-0.917,~ Convergent r =cial support 0.25-0.72efriends andfamily

Social Perceived so- (~ = 0.85-0.917~ Divergent X2 = 4/12 Perception of beingcial support 4.65, p < a valued support

0.05~7 providerEmotional Sate self-es- e = 0.72-0.88TM Convergent r = 6/20 Secure self-identity

teem scale 0.84~1~ and a sense ofself-regard

6/36 Positive physicalhealth percep-tions and expec-tations

6/10 Sense of meaningand purpose inlife

6/12 Positve expectan-cies

2/40 Perception of so-cial supportavailable

The social scales were consolidated. Intellectual items written by the authors were added afterthe initial round of pilot testing.

ric properties of the new scale. Infour pilot studies, the Perceived Well-ness Survey demonstrated evidenceof convergent validity (r = .37 to .56)and internal consistency (a = .89 to.91). In the current study, the totalscale internal consistency for thecombined sample (n = 558) was a .91. In the samples considered inde-pendently, total scale internal consis-tency ranged from a = .88 to .93.The internal validity of the total scaleis demonstrated by a high percentageof items (90%), with an item to totalscale correlation greater than .30 inthe four samples considered inde-pendently.

The scale was designed so thateach of the subscale scores couldalso be used independently to assesswellness in each dimension. Practi-tioners may find that using the Per-ceived Wellness Survey in this man-

ner is an effective way of gaining ad-ditional perceptual data to morethoroughly diagnose before prescrib-ing. The internal consistency esti-mates for each of the subscales inthis sample were, physical (a = .81),spiritual (a = .77), psychological, (c~= .71), social (c~ = .64), emotional(a = .74), and intellectual (a = .64).Nunnally1°5 has suggested that an al-pha coefficient of .70 is the mini-mum acceptable value for internalconsistency reliability. However, theinternal consistency coefficient is di-rectly dependent on the number ofitems in a given scale. Thus, to assessthe degree to which the coefficientswere a function of subscale length,the split-half reliability of each subs-cale was assessed. Correspondingly,the split-half correlation coefficientswere, physical (r = .71), spiritual = .68), psychological, (r = .62),

cial (r = .52), emotional (r = .61),and intellectual (r = .53). Implica-tions of these findings are outlinedin the discussion.

Perceived Wellness Survey for Discrimi-nant Validity. The Perceived WellnessSurvey discriminant validity versionconsists of two sets of six statementsderived from the Perceived WellnessSurvey subscale definitions. One setof statements describes a well personand the other set describes an unwellperson. Sample statements describinga well person are, "Identify five em-ployees who seem to always be physi-cally healthy," and "Identify five em-ployees who seem to expect that pos-itive things will result no matter whatthe circumstances." Sample state-ments describing an unwell personare, "Identify five employees whoseem to always be physically un-healthy," and "Identify five employ-ees who seem to be insecure withwho they are." Five out of five ex-perts familiar with the theoreticalfoundations of the Perceived Well-ness Survey agreed that the Per-ceived Wellness Survey discriminantvalidity version statements accuratelyrepresented the content of the Per-ceived Wellness Survey subscales. Dis-criminant validity was assessed by ask-ing four nursing/wellness/EAP pro-fessionals at one of the companies toidentify, from a list of employees whohad completed the Perceived Well-ness Survey, the five employees whobest exemplified the set of state-ments describing a well person andthe set of statements describing anunwell person. The method of esti-mating discriminant validity is de-scribed in the Analysis section.

Perceived Wellness Survey for Face Valid-ity. Face validity is the "extent towhich an instrument looks like itmeasures what it is intended to mea-sure’’1°5 (345). Face validity was as-sessed by administering a modifiedversion of the Perceived WellnessSurvey to the student sample. ThePerceived Wellness Survey face validi-ty version and Perceived WellnessSurvey items were identical, but theitem order of the Perceived WellnessSurvey face validity version was com-pletely randomized. In addition, the


Perceived Wellness Survey face validi-ty version contained a description ofeach dimension of wellness. Face va-lidity was estimated by the degree towhich students were able to identifycorrectly which dimension of well-ness was reflected by each PerceivedWellness Survey item.

AnalysisThe Perceived Wellness Survey

model has six dimensions, all ofwhich have proven to be significantlyand positively correlated (p -< .05)with the Perceived Wellness Surveycomposite score and with each otherin pilot research. In the currentstudy, the data were first analyzed byestimating the extent of partial inter-correlations between the PerceivedWellness Survey and subscales aftercontrolling for age and gender. Posi-tive relationships were expected.

Next, discriminant validity was esti-mated. Four health-related profes-sionals at one of the companies wereeach asked to identify five employeeswho best exemplified the set of state-ments describing both a well and anunwell person. The discriminant va-lidity of the Perceived Wellness Sur-vey was assessed by a t-test compari-son of the Perceived Wellness Surveywellness composite score means be-tween the well and unwell groups.

Face validity was estimated by stu-dents (n = 36) who were asked match each of the items from thePerceived Wellness Survey face validi-ty version to the appropriate dimen-sion. The criterion for face validitywas set a pr/0r/at r >- .80 between thestudent matches and the correctmatches.

Finally, the Perceived WellnessSurvey items were factor analyzed us-ing principal axis factoring with vari-ous solutions ranging from one toseven. The sub~ject-to-variable ratioeasily exceeded the minimum of 5 to1 recommended by Gorsuch. ~°6 Twocriteria were used to determinewhich solution best explained thedata. First, the factor loadings wereexamined to see if the items clus-tered into intuitively meaningfulgroups. Second, a scree plot was ex-amined to determine the most ap-propriate number of factors.

Because moderate intercorrela-

Table 2

Partial Correlation Coefficients of the Perceived Wellness Survey Compositewith the Perceived Wellness Survey Subscales Controlling

for Age and Gender (n = 537)*

Variable 1 2 3 4 5 6

1. Wellness composite2. Physical wellness 0.583. Spiritual wellness 0.66 0.454. Psychological wellness 0.69 0.44 0.695. Social wellness 0.57 0.30 0.506. Emotional wellness 0.66 0.45 0.707. Intellectual wellness 0.67 0.40 0.59

0.510.65 0.460.55 0.40 0.54

* The sample size is smaller than the overall sample size due to missing data. All values aresignificant at p -< 0.01.

dons among the subscales existed inpreliminary research, an oblique ro-tation was used for all of these analy-ses except for the one factor solutionwhich could not be rotated, j°7 Theintercorrelations among the subscalesalso indicated that the underlyingcontent was more alike than differ-ent. Thus, although the PerceivedWellness Survey model incorporatessix dimensions of wellness, the resultsof the factor analysis were not ex-pected to yield six separate factors.In fact, this finding would be unusu-al given the perceptual nature of theitems.

RESULTS

As expected, each subscale was sig-nificantly correlated (p <- .05) withthe composite and with each other(Table 2). The positive partial corre-lation between the composite andeach of the subscales was not surpris-ing since the subscale scores wereused to compute the composite.However, it was important to exam-ine the intercorrelation matrix to de-termine whether the subscales werepositively or negatively correlatedwith the wellness composite. The pat-tern of positive correlations providespreliminary support for the model.Perceived Wellness Survey means andstandard deviations for all fourgroups are displayed in Table 3.

Health professionals who were fa-miliar with the participants in one ofthe corporate samples were able toidentify from a list of health screen-

Table 3

Perceived Wellness Survey SampleMeans and Standard Deviations

Group Mean SD

3M 1 16.51 3.543M 2 15.31 3.34MuRata 15.35 4.04Students 16.49 3.12

3M 1 = Administrative center (n = 295);3M 2 = manufacturing plant (n = 98); Mu-Rata (n = 53); Students (n = 112).

ing participants, respectively, thosewith high and low perceived wellness(t = 5.46, p -< .05, df = 38). Thus,the Perceived Wellness Survey dem-onstrated preliminary evidence ofdiscriminant validity. In addition, stu-dents were able to match accuratelythe Perceived Wellness Survey itemsand dimensions. The correlation be-tween student matches of item-to-di-mension and the intended match wasvery high (r --- .98, p -< .05), indicat-ing that when provided with wellnessdefinitions, lay individuals were suc-cessful in discerning the content ofPerceived Wellness Survey items.

Because all items clustered on onefactor and because the first and sec-ond factors were widely separated onthe scree plot, a one-factor solutionaccounting for 24% of the commonvariance seemed to be the most par-simonious interpretation of the data.However, some have suggested thatfactor analysis of agree-disagree scales


such as the Perceived Wellness Sur-vey sometimes reveal unwanted popu-larity factors which have no relevancein terms of scale content. ~°8 To checkwhether the one-factor solution wasindeed the best explanation of thedata, two methods were used. First, amatrix of intra-item correlation coef-ficients was principal axis factoredwith a one-, two-, and three-factor so-lution. Of these, a one-factor solutionwas clearly the most meaningful andparsimonious. Second, in order todetermine whether there were any la-tent factors, a matrix of subscale in-tercorrelations was principal axis fac-tored with a one-, two-, and three-fac-tor solution. Again, a one-factor solu-tion provided the best explanation ofthe data.

In summary, all of the items load-ed on a single factor which was la-beled perceived wellness (Table 4).All but two items loaded above .30,indicating that the Perceived Well-ness Survey possesses reasonable fac-torial validity.

DISCUSSION

Health professionals were able todiscriminate between those with highand low levels of perceived wellness,and students were able to match Per-ceived Wellness Survey items andsubscale definitions. The high alphacoefficient for the total sample (~ .91) and for the samples consideredindependently (c~ = .88 to .93) pro-vides strong support for the internalconsistency of the overall scale. Col-lectively, these findings provide firmsupport for the content validity andreliability of the Perceived WellnessSurvey.

As expected, the results of the fac-tor analysis suggest that the Per-ceived Wellness Survey is a unidimen-sional scale. Because the content ofthe items is so related and becauseall of the items are perceptual, this isnot surprising. This does not suggestthat wellness is a unidimensionalphenomenon, only that perceptionsof wellness in hypothetical dimen-sions are more related by their per-ceptual nature than they are differ-entiated by their content. Interesting-ly, the strongest loading items (--->.50), share only three common

Table 4

Factor Loadings for the PerceivedWellness Survey (n = 556)

Factor IPerceived

Items Wellness

Emotional 1 0.38Emotional 2 0.71Emotional 3 0.56Emotional 4 0.45Emotional 5 0.64Emotional 6 0.50Spiritual 1 0.60Spiritual 2 0.55Spiritual 3 0.48Spiritual 4 0.62Spiritual 5 0.58Spiritual 6 0.70Social 1 0.36Social 2 0.27*Social 3 0.42Social 4 0.37Social 5 0.35Social 6 0.49Psychological 1 0.55Psychological 2 0.45Psychological 3 0.63Psychological 4 0.55Psychological 5 0.48Psychological 6 0.44Intellectual 1 0.54Intellectual 2 0.25*Intellectual 3 0.58Intellectual 4 0.41Intellectual 5 0.44Intellectual 6 0.38Physical 1 0.38Physical 2 0.44Physical 3 0.53Physical 4 0.51Physical 5 0.52Physical 6 0.42

* Indicated Ioadings did not meet the mini-mum criteria of 0.30.

themes: (1) purpose in life, (2) opti-mism, and (3) self-esteem, stronglysuggesting that the Perceived Well-ness Survey is an affective construct.This may also explain why the socialdimension was not any stronger thanit was. Despite strong associations be-tween social support measures andhealth in other studies,5-9 the socialsubscale is unique because it assessesperceptions of available external re-sources, whereas the other PerceivedWellness Survey subscales assess per-

ceptions of available internal re-sources.

Each hypothetical wellness dimen-sion is supported by a separate bodyof empirical inquiry and the contentof each is conceptually robust. In ad-dition, intervention programs basedon each dimension may have a slight-ly different look and feel, and mayappeal to different population seg-ments even though the interventionoutcomes may be highly similar. Inthis light, the wellness model and def-initions remain useful as conceptualguidelines. In addition, practitionersmay choose to use the subscalescores to assess perceived wellness ineach dimension. In this regard, fourof the six Perceived Wellness Surveysubscales possessed acceptable esti-mates of alpha internal consistency.Based on the split-half correlations,the remaining two (social and intel-lectual) are also adequately consis-tent. Finally, the Perceived WellnessSurvey as currently operationalizedpossesses evidence of discriminantand face validity.

This study has a few potential limi-tations. First, these data are subjectto limitations commonly associatedwith self-report measures.~°5 Typically,the error associated with self-reportmeasures is viewed as a random errorwhich must be controlled. However,this "error" may be valuable clientinformation that has been previouslyignored. As suggested earlier, healthperceptions, which are truly uniqueto each individual, have demonstrat-ed value as a source of predictive in-formation. ~6.~8.~0

Second, a self-selection bias mayhave existed because participantsfrom both of the corporate sampleswere recruited during general healthappraisal programs. Employees whovoluntarily participate in healthscreenings have been shown to behealthier than the normal popula-tion. ]°9 The health-related variablesin all four samples were indeedskewed in a healthful direction sup-porting this assertion. Random sam-ple selection will improve the popula-tion distribution and may enhancethe generalizability of future results.Interestingly, because the data wereskewed toward positive health, theability of the factor analyses to de-


scribe the structure of perceived well-ness was limited. In spite of this, theresults of the factor analysis support-ed the underlying perceptual natureof the Perceived Wellness Survey aspreviously suggested.

Further, wellness is probably bestexplained when accounting for cul-tural and environmental factors.2~,28

Hence, researchers interested in ap-plying the model or using the Per-ceived Wellness Survey are encour-aged to consider wellness perceptionswithin a broader systems frame-work.28,s0

Overall, the findings were promis-ing. Practitioners may find the Per-ceived Wellness Survey subscale andcomposite scores to be an additionalsource of useful information for indi-vidual level programming. For exam-ple, based on typical health assess-ment measures, an obese man wouldprobably receive a recommendationthat he modify his diet and com-mence a cardiovascular exercise pro-gram. If the Perceived Wellness Sur-vey were used in addition to thecommon health assessment measures,it might become evident that theman’s spiritual health is very low. Hishealth prescription might then in-clude a referral to an EAP counseloror to a course based on principles ofspiritual health in addition to theprevious recommendations.

Researchers may find the wellnesscomposite score to be a parsimoni-ous index of perceived wellness inseveral dimensions and a balanceamong dimensions. Additionally, useof the subscales as research tools mayalso be fruitful. The physical, spiritu-al, and intellectual subscales may beparticularly useful given the lack ofquality perceptual scales in these di-mensions. The emotional, psychologi-cal, or social subscales may also beuseful for brief assessments; however,an abundance of longer, more specif-ic scales exist in these three dimen-sions.

Given the potential impact of thePerceived Wellness Survey for bothpractitioners and researchers, furtherresearch is warranted. Research ef-forts should be focused on establish-ing the construct and concurrent va-lidity of the Perceived Wellness Sur-vey. In addition, psychometric prop-

erties such as test-retest reliabilityneed to be addressed. Finally, the val-ue of the Perceived Wellness Surveyas a perceptual measure of wellnesscould best be established by employ-ing it concurrently with widely-usedclinical, physiological, and behavioralmeasures of health to determinewhether it indeed would provide ad-ditional information.

SO WHAT? Implications forHealth Promotion Researchersand Practitioners

This study seems to providesupport for the use of the Per-ceived Wellness Survey as a validmeasure of wellness perceptions.Based on these preliminary re-sults, health promotion practition-ers may find wellness perceptionsto be a fruitful addition to clientprofiles, which have been typicallylimited to clinical, physiological,and/or behavioral data. For exam-ple, the Perceived Wellness Surveysubscale scores may assist practi-tioners in their efforts to moreacutely focus programs on individ-ual needs.

Acknowledgments

We wish to acknowledge our colleagues who ~rovidedsupport in the completion of this p~roject: Alice Buttino,Robert Dalrymple, Bill McLellan, Pat Ruff, Jenni Kirk,Tom Hallberg, Cris Champlain, Kendra Fa~ias, andothers far too numerous to mention at 334, in Austin,Texas;Jerry Ward, Rich Luscomb, and Keith Fred atLifeQuest, Inc., in Memphis, Tennessee; and Fred Peter-son, Bob Zambarano, and Tom Bohman at the Univer-sity of Texas, Austin.

References

1. Wilson I, Cleary P. Linking clinical variableswith health-related quality of life. JAVA1995;273 ( 1 ) :59-65.

2. Dunn H. High Level Wellness. Washington,DC: Mr. Vernon, 1961.

3. Crose R, Nicholas D, Gobble D, et al. Gen-der and wellness: a multidimensional systemsmodel for counseling.J Couns Dev 1992;71:149-56.

4. GreenbergJ. Health and wellness: a concep-tual differentiation. J Sch Health1985;55 (10) :403-6.

5. Cohen S, Sherrod D, Clark M. Social skillsand the stress-protective role of social sup-port. J Pers Soc Psychol 1988;50:963-73.

6. Cohen S, Wills TA. Stress, social support,and the buffering hypothesis. Psychol Bull1985;98:310-57.

7. Procidano ME, Heller tC Measures of per-ceived support from friends and from fami-ly: three validation studies. Am J Comm Psy-chol 1983; 11 ( 1 ) : 1-24.

8. Procidano M. The nature of perceived socialsupport: findings of meta-analytic studies. In:

Spielberger, Butcher, editors. Advances inPersonality Assessment. Vol. 9. Hillsdale,New Jersey: Lawrence Erlbaum 1992;1-26.

9. Wethington E, Kessler R. Perceived support,received support, and adjustment to stressfullife events. J Health Soc Behav 1986;27:78-89.

10. Antonovsky A. Unraveling the Mystery ofHealth: How People Manage Stress and StayWell. San Francisco: Josey-Bass, 1988.

11. Antonovsky A. The sense of coherence as adeterminant of health. In: Matazarro, et aL,editors. Behavioral Health: A Handbook ofHealth Enhancement and Disease Preven-tion. New York: John Wiley, 1984;114-29.

12. Kobasa S. Stressful life events, personality,and health: an inquiry into hardiness. J PersSoc Psychol 1979;37(1):1-11.

13. Selye H. Stress without Distress. New York:LP Lippincott, 1974.

14. Fylkesnes K, Forde O. The Tromso study:predictors of self-evaluated health--has soci-ety adopted the expanded health concept?Soc Sci Med 1991;32(2):141-6.

15. Idler E, Angel R. Self-rated health and mor-tality in the NHANES1 epidemiologic follow--up study. AmJ Pub Health 1990;80(4):446-52.

16. Idler E, Kasl S. Health perceptions and sur-vival, do global evaluations of health statusreally predict mortality? J Gerontol1991 ;46(2) :$55-65.

17. Kaplan G, Camacho T. Perceived health andmortality: a nine-year follow-up of the hu-man population laboratory cohort. Am J Ep-idemiol 1983;117(3):292-304.

18. MosseyJ, Shapiro E. Self-rated health: a pre-dictor of mortality among the elderly. Am JPub Health 1982;72(8) :800-8.

19. Singer E, Garfinkel R, Cohen S, et al. Mor-tality and mental health: evidence from Mid-town Manhattan restudy. Soc Sci Med1976;10:517-25.

20. Stewart A, Hays R, Ware J. Health percep-tions, energy/fatigue, and health distressmeasures. Measuring Functioning and Well-Being: The Medical Outcomes Study Ap-proach. Durham, North Carolina: Duke Uni-versity, 1992,

21. Ardell D. High Level Wellness. Berkeley, Cal-ifornia: Ten Speed Press, 1986.

22. Ardell D. High level wellness strategies.Health Educ 1977;8(4):2-4.

23. Depken D. Wellness through the lens ofgender: a paradigm shift. Wellness Perspect1994;10(2) :54-69.

24. Hettler W. Wellness: encouraging a lifetimepursuit of excellence. Health Values1984;8 (4): 1 3-7.

25. LaffertyJ. A credo for wellness. Health Educ1979;10(5):10-1.

26. TravisJ, Ryan R. Wellness Workbook. 2nded. Berkeley, California: Ten Speed Press,1988.

27. Wu R. Behavior and Illness. New Jersey:Prentice-Hall, 1986;77---85.

28. Jasnoski M, Schwartz G. A synchronous sys-tems model for health. Am Behav Sci1985;28(4):468-85.

29. Nicholas D, Gobble D, Crose R, et al. A sys-tems view of health, wellness and gender:implications for mental health counseling.JMent Health Couns 1992;14(1):8-19.

30. Seeman J. Toward a model of positivehealth. Am Psychol 1989;44(8):1099-109.

31. Whitmer J, Sweeney T. A holistic model forwellness prevention over the life span. JCouns Dev 1992;71,140-8.

33. Egbert E. Concept of wellness. J PsychiatrNurs Ment Health 1980;18(1):9-12.

34. Morrill S. Holistic stress management. In:Rothman, Levine, editors. Prevention Prac-


tice: Strategies for Physical Therapy and Oc-cupational Therapy. Philadelphia: Saunders,1992;334-44.

35. Walker S, Sechrist K, Pender N. The health-promoting lifestyle profile: Development andpsychometric characteristics. Nurs Res1987;36(2) :76-81.

36. World Health Organization. Basic Docu-ments. 15th ed. Geneva, Switzerland, 1964.

37. Eysenck H. Prediction of cancer and coro-nary heart disease mortality by means of apersonality inventory, results of a 15-year fol-low-up study. Psychol Rep 1993;72(2):499-516.

38. Bradburn N. The Structure of PsychologicalWell-being. Chicago: Aldine, 1969.

39. Reker G, Wong E Psychological and physicalwell-being in the elderly: the perceived well-being scale. J Aging 1984;3(1):23-32.

40. Ryff C. Happiness is everything, or is it? Ex-ploration on the meaning of psychologicalwell-being. J Pers Soc Psychol 1989;57 (6) 1069-81.

41. Jahoda M. Current Concepts of PositiveMental Health. New York: Basis Books, 1958.

42. Taylor S, Brown J. Illusion and well-being: asocial psychological perspective on mentalhealth. Psychol Bull 1988;103(2):193-210.

43. Andrews E Robinson J. Measures of subjec-tive well-being. In: Robinson, Shaver, andWrightsman, editors. Measures of Personalityand Social Psychological Attitudes. Vol. 1.San Diego: Academic Press, 1991;61-114.

44. Diener E. Subjective well-being. Psychol Bull1984;95(3) :542-75.

45. Kammann R, Flett R. Affectometer 2: a scaleto measure current level of general happi-ness. AustJ Psychol 1983;35(2):259-65.

46. Campbell A, Converse P, Rodgers W. TheQuality of American Life. New York: RussellSage Foundation, 1976.

47. Fazio A. A concurrent validational study ofthe NCHS General Well-being Schedule.DHEW Publication Number (HRA)1977;2(73) :78-1347.

48. Lawton M. The Philadelphia geriatric centermorale scale: a revision. J Gerontol1975;30(1) :85-9.

49. Morris J, Sherwood S. A retesting and modi-fication of the PGC morale scale. J Gerontol1975;30(1):77-84.

50. Fordyce M. The PSYCHAP inventory: a mul-ti-scale to measure happiness and its con-comitants. Soc Ind Res 1986;18,1-33.

51. Kozma A, Stones M. The measurement ofhappiness: development of the MemorialUniversity of Newfoundland scale of happi-ness (MUNSH).J Gerontol 1980;35(6):906-12.

52. Diener E, Emmons R, Larsen R, et al. Thesatisfaction with life scale. J Pers Assess1984;49 ( 1 ) :71-5.

53. Neugarten B, Havighurst R, Tobin S. Themeasurement of life satisfaction. J Gerontol1961;16:134-43.

54. Wood V, Wylie M, Sheafor B. An analysis ofa short self-report measure of life satisfac-tion: correlation with rater judgments. J Ger-ontol 1969;24(4):465-9.

55. Williams P, Wiebe D, Smith T. Coping pro-cesses as mediators of the relationship be-tween hardiness and health. J Behav Med1992;15(3):237-55.

56. Palombi B. Psychometric properties of well-ness instruments. J Couns Dev 1992;71:221-5.

57. BorysenkoJ. Minding the Body, Mendingthe Mind. Toronto: Bantam, 1987.

58. Cousins N. Anatomy of an Illness. New York:W.W. Norton, 1979.

59. Moyers B. Healing and the Mind. New York:Doubleday, 1993.

60. Ellison C. Spiritual well-being: conceptualiza-tion and measurement. J Psychol Theol1983; 11 (4) :330-40.

61. Chapman L. Spiritual health: a componentmissing from health promotion. Am JHealth Promot 1986;1(1):38-41.

62. Reker G, Peacock E. The life attitude profile(LAP): a multidimensional scale for assess-ing attitudes toward life. Can J Behav Sci1981;13(3):264-73.

63. Reker G, Peacock E, Wong P. Meaning andpurpose in life and well-being: a life-spanperspective. J Gerontol 1987;42 ( 1 ) :44-9.

64. Zika S, Chamberlain K. On the relation be-tween meaning in life and psychologicalwell-being. BrJ Psychol 1992;83(Part 1):133-45.

65. Paloutzian R, Ellison C. Loneliness, spiritualwell-being, and the quality of life. In: Pe-plau, Perlman, editors. Loneliness: A SourceBook of Current Theory, Research andTherapy. New York: John Wiley, 1982;224-37.

66. Ellis J, Smith P. Spiritual well-being, socialdesirability and reasons for living: is there aconnection? IntJ Soc Psychol 1991;37(1):57-63.

67. Fehring R, Brennan P, Keller, M. Psychologi-cal and spiritual well-being in college stu-dents. Res Nurs Health 1987;10:391-8.

68. Bailey K, Nava G. Psychological kinship,love, and liking: preliminary validity data. JClin Psychol 1989;45(4):587-94.

69. Scheier M, Carver C. Optimism, coping andhealth: assessment and implications of gen-eralized outcome expectancies. Health Psy-chol 1985;4(3):219-47.

70. Sweetman M, Munz D, Wheeler R. Opti-mism, hardiness, and explanatory style aspredictors of general well-being among at-torneys. Soc Ind Res 1993;29:153-61.

71. BroadJ. Coping with the mental illness ofone’s spouse (spousal coping). DAI-B1993;53(12):6218.

72. Strack S, Carver C, Blaney, P. Predicting suc-cessful completion of an aftercare programfollowing treatment for alcoholism: the roleof dispositional optimism. J Pers Soc Psychol1987;53:579-84.

73. Scheier M, Matthews K, Owens J, et al. Dis-positional optimism and recovery from coro-nary artery bypass surgery: The beneficial ef-fects on physical and psychological well-be-ing. J Pers Soc Psychol 1989;57:1024-40.

74. Fitzgerald E, Tennen H, Affleck G, et al.The relative importance of optimism andcontrol appraisals in quality of life after cor-onary artery bypass surgery. J Behav Med1993;16(1 ) :25-53.

75. Lee C, Ashford S, Jamieson L. The effects oftype A behavior dimensions and optimismon coping strategy, health and performance.J Org Behav 1993;14:143-57.

76. Carver C, Pozo C, Harris S, et al. How cop-ing mediates the effect of optimism on dis-tress: A study of women with early stagebreast cancer. J Pers Soc Psychol 1993;65 (2) 375-90.

77. BlumenthalJ, Burg M, Barefoot J, et al. So-cial support, type A behavior, and coronaryartery disease. Psychosom Med 1987;49:325-40.

78. House J, Robbins C, Metzner H. The associ-ation of social relationships and activitieswith mortality: prospective evidence fromthe Tecumseh community health study. Am JEpidemiol 1982;116(1):123-40.

79. Kaplan G, Salonen J, Cohen R, et al. Socialconnections and mortality from all causesand from cardiovascular disease: prospectiveevidence from eastern Finland. Am J Epide-miol 1988;128(2):370-80.

80. Reed D, McGee D, Yano K, et al. Social net-works and coronary heart disease amongJapanese men in Hawaii. Am J Epidemiol1983;117 (4) :384-96.

81. Ruberman W, Weinblatt E, GoldbergJ, et al.PsychoSoc influences on mortality after myo-cardial infarction. N Engl J Med1984;311 (9) :552-9.

82. Berkman L, Syme S. Social networks, hostresistance, and mortality: a nine-year follow--up study of Alameda county residents. AmJEpidemiol 1979;109(2):186-204.

83. Klein H, Tatone C, Lindsay N. Correlates oflife satisfaction among military wives. J Psy-chol 1989;123(5):465-75.

84. Shumaker S, Hill D. Gender differences insocial support and physical health. HealthPsychol 1991;10(2):102-11.

85. Manning F & Fullerton T. Health and well-being in highly cohesive units of the USarmy.J Appl Soc Psychol 1988;18(6):503-19.

86. Levitt M, Clark M, Rotton J, et al. Social sup-port, perceived control, and well-being: astudy of an environmentally stressed popula-tion. IntJ Aging Hum Dev 1987;25(4):247-58.

87. Seeman T, Syme S. Social networks and cor-onary artery disease: a comparison of thestructure and function of social relations aspredictors of disease. Psychosom Med1987;49:341-54.

88. Burckhardt C. The impact of arthritis onquality of life. Nurs Res 1985;34(10):11-8.

89. Dirksen S. Perceived well-being in malignantmelanoma survivors. Oncol Nurs Forum1989;16(3):353-8.

90. Kozma A, Stones M. Some research issuesand findings in the study of psychologicalwell-being in the aged. Can Psychol Rev1978;19(3):241-9.

91. Blascovich J, Tomaka J. Measures of self-es-teem. In: Measures of Personality and SocialPsychological Attitudes. Vol. 1. Robinson,Shaver, and Wrightsman, editors. San Diego:Academic Press, 1991;115.

92. Adams T, Bezner J, Steinhardt M. Principle-centeredness: a values clarification approachto wellness. Meas Eval Couns Dev1995;28 (3) : 139-47.

93. Bezner J, Adams T, Steinhardt M. The rela-tionship of body dissatisfaction to physicalhealth and wellness. In press.

94. Aronsson G. Dimensions of control as relat-ed to work organization, stress, and health.IntJ Health Serv 1989;19(3):459-68.

95. Suedfeld P. Stressful levels of environmentalstimulation. In: Sarason & Spielberger, edi-tors. Stress and Anxiety. Vol. 6. New York:John Wiley, 1979.

96. Cohen S. Cognitive processes as determi-nants of environmental stress. Issues MentHealth Nurs 1985;7(104):65-81.

97. Lawton M. Residential environment and self-directedness among older people. Am Psy-chol 1990;45(5):638-40.

98. Langer E, Rodin J. The effects of choice andenhanced personal responsibility for theaged: a field experiment in an institutionalsetting. J Pers Soc Psychol 1976;34(2):191-8.

99. Rodin J, Langer E. Long-term effects of acontrol-relevant intervention with the institu-tionalized aged. J Pers Soc Psychol1977;35 (12) :897-902.Gotterer S. Storytelling: a valuable supple-ment to poetry writing with the elderly. ArtsPsychother 1989;16(2):127-31.Kolanowski A, Gunter L. What are thehealth practices of retired career women? JGerontol Nurs 1985;11 (12) :22-30.Stevens E Ten-Have M. Activity level, bore-dom, health and participative climate in

100.

101.

102.


nursing homes. Tijdschr Gerontol Geriatr1993;24(1) :18-23.

103. SPSS Reference Guide. Chicago: SPSS Inc.,1990.

104.Sudman S, Bradburn N. Asking Questions: APractical Guide to Questionnaire Design.San Francisco:Josey-Bass, 1991.

105. NunnallyJ. Psychometric Theory. New York:McGraw-Hill, 1978.

106. Gorsuch R. Factor Analysis. Hillsdale, NewJersey: Lawrence Erlbaum, 1983.

107. McDonald R. Factor Analysis and RelatedMethods. Hillsdale, New Jersey: LawrenceErlbaum, 1985.

108.Thorndike R. Applied Multivariate Data.New York: Harcourt Brace Jovanovich, 1982.

109.Golaszewski T, Lynch W, Clearie A, et al.The relationship between retrospectivehealth insurance claims and a health risk ap-praisal-generated measure of health status. JOccup Med 1989;31 (3) :262-4.

110.Ware J, Sherbourne C. The MOS 36-itemshort-form health survey (SF-36). Med Care1992;30(6) :473-83.

111.Heatheron T, Polivy J. Development and val-idation of a scale for measuring state self-es-teem. J Pets Soc Psychol 1991;60:895-910.

Appendix

Psychological Items1. I am always optimistic about my fu-

7. I rarely count on good things happen-ing to me.*

13. I always look on the bright side ofthings.

19. In the past, I have expected the best.25. In the past, I hardly ever expected

things to go my way.*31. Things will not work out the way I

want them to in the future.*

Emotional Items2. There have been times when I felt in-

ferior to most of the people I knew.*8. In general, I fecl confident about my

abilities.14. I sometimes think I am a worthless in-

dividual.*20. I am uncertain about my ability to do

things well in the future.*

26. I will always be secure with who I am.32. In the past, I have felt sure of myself

among strangers.

Social Items3. Members of my family come to me

for support.9. Sometimes I wonder if my family will

really be there for me when I am inneed.*

15. My friends know they can always con-fide in me and ask me for advice.

21. My family has been available to sup-port me in the past.

27. In the past, I have not always hadfriends with whom I could share myjoys and sorrows.*

33. My friends will be there for me whenI need help.

Physical Items4. My physical health has restricted me

in the past.*10. My body seems to resist physical ill-

ness very well.16. My physical health is excellent.22. Compared to people I know, my past

physical health has been excellent.28. I expect to always be physically

healthy.34. I expect my physical health to get

worse.*

Spiritual Items5. I believe that there is a real purpose

for my life.l 1. Life does not hold much future

promise for me.*17. Sometimes I don’t understand what

life is all about.*23. I feel a sense of mission about my fu-

ture.29. I have felt in the past that my life was

meaningless.*35. It seems that my life has always had

purpose.

Intellectual Items6. I will always seek out activities that

challenge me to think and reason.12. I avoid activities which require me to

concentrate.*18. Generally, I feel pleased with the

amount of intellectual stimulation Ireceive in my daily life.

24. The amount of information that Iprocess in a typical day is just aboutright for me (i.e., not [too much, nottoo little] ).

30. In the past, I have generally found in-tellectual challenges to be vital to myoverall well-being.My life has often seemed devoid ofpositive mental stimulation.*

36.

The Perceived Wellness SurveyScoring Methods1. Score each item from 1, "Very strong-

ly disagree" to 6, "Very stronglyagree." Items with * are reversescored.

2. Sum all of the subscale means. Theresult is the Wellness Magnitude.

3. Divide Wellness Magnitude by 6. Theresult is called "xbar."

4. For each subscale, compute the fol-lowing: (subscale mean - xbar) 2. Theresult is called subscale deviation.

5. Sum all of the subscale deviations,then divide the total by 5 (n - 1).The result is called the variance.Compute the Wellness Balance withthe following formula ([square rootof the variance] + 1.25). The 1.25 isadded to the denominator to preventa Wellness Balance of 0 from creatingan invalid Wellness Composite.

6. Compute the Wellness Compositewith the following formula: WellnessMagnitude + Wellness Balance.

7. For a copy of the SPSS program usedto score the Perceived Wellness Sur-vey, please contact the author.


the s c i en c e of health promotion methods, issues, and

Documents