PERSONALITY PROCESSES AND INDIVIDUAL DIFFERENCES

Understanding and Assessing the Motivations of Volunteers:A Functional Approach

E. Gil ClaryCollege of St. Catherine

Mark SnyderUniversity of Minnesota, Twin Cities Campus

Robert D. RidgeBrigham Young University

John CopelandSt. James Associates

Arthur A. StukasUniversity of Pittsburgh

Julie Haugen and Peter MieneWinona State University

The authors applied functionalist theory to the question of the motivations underlying volunteerism,hypothesized 6 functions potentially served by volunteerism, and designed an instrument to assessthese functions (Volunteer Functions Inventory; VFI). Exploratory and confirmatory factor analyseson diverse samples yielded factor solutions consistent with functionalist theorizing; each VFI motiva-tion, loaded on a single factor, possessed substantial internal consistency and temporal stability andcorrelated only modestly with other VFI motivations (Studies 1, 2, and 3). Evidence for predictivevalidity is provided by a laboratory study in which VFI motivations predicted the persuasive appealof messages better when message and motivation were matched than mismatched (Study 4), and byfield studies in which the extent to which volunteers' experiences matched their motivations predictedsatisfaction (Study 5) and future intentions (Study 6). Theoretical and practical implications arediscussed.

Every year, millions of people devote substantial amounts oftheir time and energy to helping others. One important manifes-tation of human helpfulness is volunteerism, whereby peopleprovide, among other services, companionship to the lonely,tutoring to the illiterate, counseling to the troubled, and healthcare to the sick, and do so on a regular, ongoing, voluntary

E. Gil Clary, Department of Psychology, College of St. Catherine;Mark Snyder, Department of Psychology, University of Minnesota,Twin Cities Campus; Robert D. Ridge, Department of Psychology, Brig-ham Young University; John Copeland, St. James Associates, North-brook, Illinois; Arthur A. Stukas, Department of Epidemiology, Univer-sity of Pittsburgh; Julie Haugen and Peter Miene, Department of Psy-chology, Winona State University.

This research was supported by grants from the Gannett Foundation,the Aspen Institute's Nonprofit Sector Research Fund, the National Sci-ence Foundation, and the National Institute of Mental Health. We thankSimone French for her contributions to Study 1 and Steve Asche forhis assistance with the confirmatory factor analyses.

Correspondence concerning this article should be addressed to E. GilClary, Department of Psychology, College of St. Catherine, St. Paul,Minnesota 55105, or Mark Snyder, Department of Psychology, Uni-versity of Minnesota, Twin Cities Campus, 75 East River Road, Minne-apolis, Minnesota 55455. Electronic mail may be sent to gclary@stkate.edu or msnyder@tc.umn.edu.

basis, with their voluntary helping often extending over longperiods of time. According to one estimate, 89.2 million Ameri-can adults engaged in some form of volunteerism in 1993, with23.6 million of them giving 5 or more hours per week to theirvolunteer service (Independent Sector, 1994). Moreover, volun-teerism is not simply an American phenomenon but rather isan activity that can be found in many other parts of the world(Curtis, Grabb, & Baer, 1992).

The idea that an individual would make significant personalsacrifices for another person, particularly when that person is astranger, has long fascinated students of social behavior (e.g.,Batson, 1991; Eisenberg, 1986; Latane & Darley, 1970; Piliavin,Dovidio, Gaertner, & Clark, 1981; Schroeder, Penner, Dovi-dio, & Piliavin, 1995; Staub, 1978). Although studies of helpinghave long been mainstays of psychological inquiry, the existingliterature speaks largely to varieties of helping somewhat differ-ent from volunteerism, focusing on helping in contexts wherea potential helper is faced with an unexpected need for help,calling for an immediate decision to act and an opportunity toprovide one and only one relatively brief act of help (Bar-Tal,1984; Benson et al., 1980; Piliavin & Charng, 1990).

To be sure, factors uncovered by research on the helpingthat occurs in these kinds of contexts, sometimes referred to asspontaneous helping, may be important influences in volun-

Joumal of Personality and Social Psychology. 1998, Vol. 74, No. 6, 1516-1530Copyright 1998 by the American Psychological Association, Inc. 0022-3514/98/J3.00

1516

FUNCTIONS OF VOLUNTEERING 1517

teerism as well. Yet, volunteerism appears to be exemplary ofa rather different kind of helping, a kind that is prototypic ofplanned helping, which often ' 'calls for considerably more plan-ning, sorting out of priorities, and matching of personal capabili-ties and interests with type of intervention" (Benson et al.,1980, p. 89). Thus, volunteers (a) often actively seek out oppor-tunities to help others; (b) may deliberate for considerableamounts of time about whether to volunteer, the extent of theirinvolvement, and the degree to which particular activities fitwith their own personal needs; and (c) may make a commitmentto an ongoing helping relationship that may extend over a con-siderable period of time and that may entail considerable per-sonal costs of time, energy, and opportunity. Why, then, do peo-ple volunteer and what sustains voluntary helping?

The defining and characteristic features of volunteerism asvoluntary, sustained, and ongoing helpfulness suggest that itmay be productive to adopt a motivational perspective and toinquire about the motivations that may dispose individuals toseek out volunteer opportunities, to commit themselves to vol-untary helping, and to sustain their involvement in volunteerismover extended periods of time. After all, the fundamental con-cerns of motivational inquiry with understanding the processesthat move people to actionthe processes that initiate, direct,and sustain actionare precisely the concerns engaged by thequestions "why do people volunteer?" and "what sustains vol-untary helping?" In addressing these questions, we haveadopted the strategy of functional analysis, an approach that isexplicitly concerned with the reasons and the purposes, the plansand the goals, that underlie and generate psychological phenom-enathat is, the personal and social functions being served byan individual's thoughts, feelings, and actions (Snyder, 1993).

The Functional Approach to Motivation

In psychology, the themes of functionalism have a long anddistinguished tradition and are reflected in diverse perspectivesthat emphasize the adaptive and purposeful strivings of individu-als toward personal and social goals (Cantor, 1994; Snyder,1993). A central tenet of functionalist theorizing is that peoplecan and do perform the same actions in the service of differentpsychological functions. Arguably, the most familiar examplesof functional theorizing in personality and social psychologyare the classic accounts of attitudes and persuasion articulatedby Smith, Bruner, and White (1956) and by Katz (1960). Thesetheorists proposed that the same attitudes could serve differentfunctions for different people and that attempts to change atti-tudes would succeed to the extent that they addressed the func-tions served by those attitudes. More recently, however, there hasbeen a broadening of the scope of the application of functionalisttheorizing. Functions with at least a family resemblance to thoseproposed for attitudes appear with some regularity in motiva-tionally oriented analyses of diverse cognitive, affective, behav-ioral, and interpersonal phenomena (e.g., Cantor, 1994; Snyder,1992, 1993).

In the tradition of such theorizing, we suggest that the keythemes of functional analyses that have contributed to the under-standing of phenomena and processes in the realms of attitudesand persuasion, social cognition, social relationships, and per-sonality also hold the promise for unraveling the complex moti-

vational foundations of volunteer activity (see also Clary &Snyder, 1991; Snyder & Omoto, 1992). The core propositionsof a functional analysis of volunteerism are that acts of volun-teerism that appear to be quite similar on the surface may reflectmarkedly different underlying motivational processes and thatthe functions served by volunteerism manifest themselves inthe unfolding dynamics of this form of helpfulness, influencingcritical events associated with the initiation and maintenance ofvoluntary helping behavior.

Functions Served by Volunteerism

What, then, are the functions served by volunteerism? Toanswer this question, we took previous functional theorizing,specifically the classic theories of attitudes offered by Katz(1960) and Smith et al. (1956), as a heuristic point of departurefor considering the motivational foundations of volunteerism.Although the labels vary, several functions are common to bothKatz's and Smith et al.'s taxonomies (in the following, Katz'slabels are used and Smith et al.'s labels are in parentheses).Some attitudes are thought to serve a knowledge (object ap-praisal) function, bringing a sense of understanding to the world;other attitudes serve a value expressive (quality of expressive-ness) function, helping people express deeply held values, dis-positions, and convictions; and still other attitudes serve an egodefensive (externalization) function, buffering people againstundesirable or threatening truths about the self. In addition tosharing these functions in common, Katz proposed a utilitarianfunction by which attitudes reflect experiences with rewardingand punishing events, and Smith et al. proposed a social ad-justive function served when attitudes help people fit in withimportant reference groups.

Part of the appeal of these earlier functional theories was thediversity of motivations that they could embrace, a diversityreflecting to some extent the functional theorists' (e.g., Katz,1960) invoking of the themes of the grand psychological theoriesof human nature in the functions they proposed (e.g., the defen-sive function captures elements of psychodynamic theory, theknowledge function is reminiscent of Gestalt psychology, theexpressive function has the flavor of self-psychology about it,and the utilitarian and adjustive functions are reminiscent of thebehaviorist perspective on human nature). In this same spiritof inclusive theorizing about the motivational foundations ofaction, we have proposed that the diverse functions identified insuch functional theorizing have their counterparts in volunteers'motivations. However, further refinement and articulation of therole of ego-related functions have suggested an important dis-tinction between eliminating negative aspects associated withthe ego and promoting positive strivings associated with the ego.The result was a set of six motivational functions served byvolunteerism. Also, in fact, previous research is readily inter-pretable within this functional framework.

Values

One function that may be served by involvement in volunteerservice centers on the opportunities that volunteerism providesfor individuals to express values related to altruistic and humani-tarian concerns for others. Related to Katz's (1960) value ex-

1518 CLARY ET AL.

pressive functions and Smith et al.'s (1956) quality of expres-siveness functions, concern for others is often characteristic ofthose who volunteer (Anderson & Moore, 1978), distinguishesvolunteers from nonvolunteers (Allen & Rushton, 1983), andpredicts whether volunteers complete their expected period ofservice (Clary & Miller, 1986; Clary & Orenstein, 1991).

UnderstandingA second function potentially served by volunteering involves

the opportunity for volunteerism to permit new learning experi-ences and the chance to exercise knowledge, skills, and abilitiesthat might otherwise go unpracticed. Related to the knowledgeand object appraisal functions in theories of attitudes and persua-sion, this understanding function is exemplified by the largenumber of Gidron's (1978) volunteers in health and mentalhealth institutions who expected to receive benefits related toself-development, learning, and variety in life through their vol-unteer service.

SocialA third function that may be served by volunteering reflects

motivations concerning relationships with others. Volunteeringmay offer opportunities to be with one's friends or to engagein an activity viewed favorably by important others. This socialfunction is clearly related to Smith et al.'s (1956) social ad-justive function and has figured prominently in several accountsof helpfulness, including Rosenhan's (1970) portrait of the par-tially committed civil rights activists whose helpfulness wasguided by concerns over social rewards and punishments.

Career

A fourth function that may be served by volunteering is con-cerned with career-related benefits that may be obtained fromparticipation in volunteer work. Related to the utilitarian func-tion described by Katz (1960), this career function is exempli-fied by the Junior League volunteers studied by Jenner (1982),15% of whom perceived volunteering to be a means of preparingfor a new career or of maintaining career-relevant skills.

Protective

A fifth function traces its roots to functional theorizing'straditional concerns with motivations involving processes asso-ciated with the functioning of the ego. Related to ego defensive(Katz, 1960) or externalization (Smith et al., 1956) concerns,such motivations center on protecting the ego from negativefeatures of the self and, in the case of volunteerism, may serveto reduce guilt over being more fortunate than others and toaddress one's own personal problems. This protective functionoffers an interpretation of Frisch and Gerard's (1981) findingthat some Red Cross volunteers reported that they volunteer toescape from negative feelings.

EnhancementFinally, a sixth proposed function of volunteering derives

from indications that there may be more to the ego, and espe-

cially the ego's relation to affect, than protective processes.First, recent research on mood suggests that negative affect andpositive affect are separate dimensions rather than endpoints ona bipolar scale (e.g., Watson, Clark, Mclntyre, & Hamaker,1992; Watson, Clark, & Tellegen, 1988). Second, research onmood and helping points to different mechanisms by whichpositive and negative moods influence helpfulness (Carlson,Charlin, & Miller, 1988; Carlson & Miller, 1987; Cunningham,Steinberg, & Grev, 1980); in the case of positive mood, peopleuse helping as a means of maintaining or enhancing positiveaffect. Finally, research on volunteerism has found evidence ofpositive strivings, as when some respondents report that theyvolunteer for reasons of personal development (Anderson &Moore, 1978) or to obtain satisfactions related to personalgrowth and self-esteem (Jenner, 1982). Thus, in contrast to theprotective function's concern with eliminating negative aspectssurrounding the ego, the enhancement function involves a moti-vational process that centers on the ego's growth and develop-ment and involves positive strivings of the ego.

Previous research on volunteers' motivations, then, is clearlycompatible with the motivations suggested by functional theo-rizing, even though none of the studies that we have reviewedwas designed specifically to investigate the functional approachto motivation. We should note that, in proposing six functionsfor volunteering, we were well aware that the question ofwhether six is the optimal number of functions would quicklyspring to readers' minds. We recognized that the theories ofKatz (1960) and Smith et al. (1956), which served as launchingpads for our analysis, each used only four functions, and wewere mindful of the fact that our own preliminary considerationsof the functional bases of volunteerism began with examinationsof but four functions (Clary & Snyder, 1991). However, as wehave attempted to convey here, our attempts to build on previoustheorizing (our own included), to integrate the shared and dis-tinct functions proposed in previous theorizing, and to furtherarticulate such theorizing to incorporate important distinctionsin self- and ego-related functioning clearly led to a conceptual-ization involving six motivational functions potentially servedby volunteerism.

Whatever the final word on the number of motivations forvolunteerism (a matter that is discussed further in the GeneralDiscussion section), the essential message of the functionalperspective is that it encourages us to consider a wide range ofpersonal and social motivations that promote this form of sus-tained helping behavior. Moreover, in doing so, the functionalapproach advances an interactionist position, as it argues thatimportant consequences follow from matching the motivationscharacteristic of individuals to the opportunities afforded bytheir environments. For, as we examined in our empirical investi-gations, it follows from the functional account of volunteerismthat people can be recruited into volunteer work by appealingto their own psychological functions, that they will come to besatisfied volunteers to the extent that they engage in volunteerwork that serves their own psychological functions, and thatthey will plan to continue to serve as volunteers to the extent thattheir psychological functions are being served by their service.

In the present article, we report on the development of aninstrument designed to measure the functions served by volun-teerism, the Volunteer Functions Inventory (VFI). During the

FUNCTIONS OF VOLUNTEERING 1519

course of its development, we sampled from diverse populationsof volunteers, as well as from the population of nonvolunteers,using diverse methodological tools to demonstrate the reliabilityand validity of the VFI and using both field and laboratorymethods of investigation. More specifically, our first set of threestudies is concerned with the construction of the VFI, investiga-tions into its factor structure, and assessment of its reliability.The second set of three studies is devoted to validation, usingthe VFI to test functionally derived hypotheses in the context ofeach stage of the volunteer processrecruitment of volunteers,satisfaction with the experience of volunteering, and commit-ment to the role of volunteer.

Development of an Inventory of Volunteers' Motivations

Study 1: The VFI

The functional approach to volunteerism is predicated on theassumption that the motivations underlying volunteer activitycan be identified and measured with some degree of precision.In this regard, it is important to emphasize that the need toidentify and measure psychological functions underlying beliefsand behaviors has been a critical task in functional analyses inother domains, with the lack of measuring devices often imped-ing theoretical and empirical progress (Kiesler, Collins, &Miller, 1969; Snyder & DeBono, 1987). Moreover, several au-thors have noted that studies of volunteers' motivations haveoften been conducted so in ways that raise conceptual and meth-odological concerns, including using instruments without a con-ceptual foundation and of unknown reliability and validity (forelaboration, see Clary & Snyder, 1991; Cnaan & Goldberg-Glen, 1991). Finally, although there exist reliable and validmeasures of motivations for specific kinds of volunteer service(e.g., the measure of motivations for AIDS volunteerism devel-oped by Omoto & Snyder, 1995), there remains a need for aninventory that reliably and validly taps a set of motivations ofgeneric relevance to volunteerism.

To develop such an inventory, we generated a set of items toreflect the psychological and social functions of volunteerismidentified by our conceptual analysis. In accord with an investi-gative strategy for personality and social behavior outlined bySnyder and Ickes (1985), we sought a population for whommotivations for volunteering would be salient, accessible, andmeaningful. Thus, we administered the inventory to currentlyactive volunteers engaged in diverse forms of volunteer service,individuals who, by virtue of having involved themselves involunteer activity, could be presumed to possess motivationsrelevant to volunteerism.

Method

Participants. Participants were 321 female and 144 male (and 2 ofunspecified gender) volunteers from five organizations in the Minneapo-lis and St. Paul metropolitan areas that used volunteers to provide awide range of services to children, families of cancer patients, socialservice and public health clients, and the physically handicapped, aswell as blood services and disaster relief. The mean age of these volun-teers was 40.9 years (SD = 13.38); their mean length of volunteerservice was 68.2 months (SD = 87.08); and 89% reported educational

experiences beyond high school, with 60% reporting at least an under-graduate degree.

Procedure. The items of the VFI were rationally derived from con-ceptualizations of the six proposed psychological and social functionsserved by involvement in volunteer work (values, understanding, career,social, protective, and enhancement). Development of the VFI was alsoinformed by previous research on volunteerism, which has used bothquantitative and qualitative means for identifying motivations (seeClary & Snyder, 1991, for a fuller review of that literature). Preliminarystudies with volunteers and nonvolunteers permitted us to identify andeliminate unreliable and ambiguous items, resulting in an instrumentconsisting of 30 items, with 5 items assessing each of the six functions.(The items of the VFI can be found in Table 1.)

Each participating organization's director of volunteer services ad-ministered the VFI to its volunteers. Respondents were asked to indicate"how important or accurate each of the 30 possible reasons for volun-teering were for you in doing volunteer work,'' using a response scaleranging from 1 (not at all important/accurate) to 7 (extremely im-portant/accurate). Scale scores resulted from averaging scores on thefive items, such that individuals' scores on each scale could range from1 to 7; the higher the score, the greater the importance of the motivation.

Results and Discussion

A primary reason for this first investigation was to examinethe structure of volunteers' motivations for volunteering and toevaluate the psychometric properties of the VFI as a measureof these motivations. To do so, we conducted a factor analysis ofparticipants' responses to the VFI. Beginning with a principal-components analysis, we identified 6 components with eigenval-ues greater than 1.0, suggesting 6 factors underlying responsesto the VFI (Kim & Mueller, 1988). For the first 12 factorsresulting from this analysis, the eigenvalues (with percent vari-ance accounted for presented in parentheses) were 8.26 (28%),3.29 (39%), 2.47 (47%), 2.14 (54%), 1.40 (59%), 1.20(63%), 0.88 (65%), 0.84 (68%), 0.77 (71%), 0.70 (73%),0.65 (75%), and 0.58 (77%). In addition, the resulting screeplot of the eigenvalues revealed that the leveling off to a straighthorizontal line occurred after the sixth eigenvalue, further sug-gesting 6 factors (Cattell, 1966). With this evidence indicatinga 6-factor solution, we then performed a principal-axis factoranalysis with oblique rotation to a preselected 6-factor solution.

The factors that emerged from this analysis clearly reflecteach of the functions that we proposed were served by volun-teering. Almost without exception, items from each scale loadedon their intended factor and did not load with items from differ-ent scales. (The only exception is one item from the enhance-ment scale, Item 29, which loaded with the understanding itemson the fifth factor; generally, however, the enhancement itemsloaded together on one factor.) Thus, this factor analysis resultedin a nearly perfectly clean structure, providing evidence that ourtheoretically derived motivations were distinct and evident inthe responses of actual volunteers.

Although these exploratory analyses clearly point to a six-factor solution, the question of whether there are more or fewerfactors did arise. To answer this question, we conducted twoadditional principal-axis factor analyses with oblique rotations,one to a preselected five-factor solution and one to a preselectedseven-factor solution. For the five-factor solution, the pattern ofloadings for the Values, Career, Social, and Understanding itemswas consistent with the results of the six-factor solution, and

1520 CLARY ET AL.

Table 1Volunteer Sample Factor Pattern Matrix (Principal-Axis Factor Analysis, Oblique Rotation,Six Factors Specified) for VFI Items, Study 1

Factor

VFI scale and items

Protective7. No matter how bad I've been feeling, volunteering helps me to forget about it.9. By volunteering I feel less lonely.

11. Doing volunteer work relieves me of some of the guilt over being morefortunate than others.

20. Volunteering helps me work through by own personal problems.24. Volunteering is a good escape from my own troubles.

Values3. I am concerned about those less fortunate than myself.8. I am genuinely concerned about the particular group I am serving.

16. I feel compassion toward people in need.19. I feel it is important to help others.22. I can do something for a cause that is important to me.

Career1. Volunteering can help me to get my foot in the door at a place where I would

like to work.10. I can make new contacts that might help my business or career.15. Volunteering allows me to explore different career options.21. Volunteering will help me to succeed in my chosen profession.28. Volunteering experience will look good on my resume.

Social2. My friends volunteer.4. People I'm close to want me to volunteer.6. People I know share an interest in community service.

17. Others with whom I am close place a high value on community service.23. Volunteering is an important activity to the people I know best.

Understanding12. I can learn more about the cause for which I am working.14. Volunteering allows me to gain a new perspective on things.18. Volunteering lets me learn things through direct, hands on experience.25. I can learn how to deal with a variety of people.30. I can explore my own strengths.

Enhancement5. Volunteering makes me feel important.

13. Volunteering increases my self-esteem.26. Volunteering makes me feel needed.27. Volunteering makes me feel better about myself.29. Volunteering is a way to make new friends.

.53

.63

.43

.72

.78

.63

.64

.72

.70

.62

.83

.85

.68

.73

.68

.58

.59

.70

.79

.80

- .43-.56- .64- .65- .82

-.42

-.62- .75- .64-.77

Note. Only factor loadings greater than .30 are shown, n = 427. VFI = Volunteer Functions Inventory.

the Protective and Enhancement items loaded together on onefactor (with Item 29 again loading with Understanding). Forthe seven-factor solution, the pattern of loadings was fully identi-cal to that for the six-factor solution, as no items loaded onFactor 5. We also examined the number of factors issue withconfirmatory factor analytic techniques, using LISREL to testeach of the five-, six-, and seven-factor solutions with an obliquemodel using item pair indicators and a pattern of loadings speci-fied such that items were constrained to load on their specifiedfactor from the exploratory analyses. The test of the six-factorsolution provided confirmatory support for the six-factor model;LISREL's goodness-of-fit index (GFI) was .91, the normed fitindex (NFI; Bentler & Bonett, 1980) was .90, and the root meansquared residual index (RMSres) of goodness of fit was .057.For the five-factor solution, the GFI was .88, NFI was .87, andRMSres was .064. For the seven-factor solution, GFI was .91,NFI was .90, and RMSres was .054. A similar pattern of support

for the six-factor model is provided by considering chi-squares:For the five-factor solution, * 2(125, N = 434) = 519.19; forthe six-factor solution, x2(120, N = 434) = 412.69; and forthe seven-factor solution, x 2 ( H 5 , N = 434) = 399.49, ps