motives are necessarily egoistic, and not all autonomous motives
are altruistic. For example, volunteering to help because one
wishes to gain particular skills or a set of experiences could be an
autonomous, yet egoistic, motive. Despite this, there may be
considerable overlap between egoism and control (e.g., self-esteem
motives reflect both) and altruism and autonomy (e.g., when both
reflect a sense of caring for the recipient). Thus, the dynamics of
autonomy/control may be relevant for understanding the impact of
altruism versus egoism on the quality of helping behavior and the
well-being effects derived from it.
Volition in Volunteers
Thoits (1994) proposed that volition or agency must be involved
for volunteer work to occur. In this way, she is consistent with
SDT in highlighting the importance of agency. An important
component of Thoits’s theory is the expectation that people with
greater personal well-being may have more volition to volunteer,
such that those with positive personality attributes (happiness,
self-esteem, low depression) and social resources are more likely
to volunteer, and that this work in turn promotes further well-
being. In this way, Thoits focused on helpers before engaging in
volitional prosocial acts, whereas the present article tests a model
in which volitional prosocial acts enhance needs after action. By
examining need satisfactions after helping, the present work may
speak to expectations that helping provides a sense of purpose and
can facilitate well-being (Thoits, 1992).
In sum, the theories discussed above imply that the source of
helpers’ motivations is important, although none specifically or
directly focuses on the relative volition of helpers’ motivation.
Other approaches not presently discussed, such as the functional
approach to prosocial behaviors (Clary et al., 1998), examine the
content of reasons provided by helpers for their actions. These
content reasons can also vary in their relative autonomy or degree
of controlled motivations. The present article thus extends previ-
ous research by testing whether the distinction between autonomy
and control is important for prosocial acts and whether it holds
discriminative power in predicting outcomes of helping for both
helpers and recipients.
The Present Research
The present research investigated a number of primary hypoth-
eses. First, we examined the hypothesis that prosocial acts foster
well-being in the actor when these acts are autonomous, but not
when they are controlled. Second, we hypothesized that these
positive effects of autonomous versus controlled helping on well-
being would be mediated by the satisfaction of basic psychological
needs. Finally, we examined the effects of the helper’s motivation
on the recipient; we expected that recipients would benefit most
when helpers are motivated autonomously rather than by control.
To test these hypotheses we present four studies. Study 1 ex-
amines the effects of prosocial behavior engagement, motivation to
help, and well-being consequences in naturalistic settings using a
diary methodology. To date, experience sampling approaches have
not been used widely in prosocial research, yet they offer the
opportunity to examine varied acts of helping over time at both
between- and within-persons levels of analysis. Studies 2 and 3 use
experimental designs to investigate well-being benefits to the
helper and to better establish a causal model for the effects of
variations in autonomy in helper motives. Study 3 also explores the
effects of autonomous motivation on recipient outcomes in a
dyadic helping task. Finally, Study 4 manipulates motivation and
presence of helping to compare effects on helper and recipient.
Study 1
In Study 1 we used a daily diary method to examine the effects
of helping (i.e., instances of prosocial behavior) on daily subjective
well-being (SWB), vitality, and self-esteem. Helping in the present
study was defined as engagement in activities that involved help-
ing someone else or doing something for a good cause. We used
measures that allowed us to examine the effects of prosocial
behavioral engagement on physical as well as psychological well-
being. We also used a diary approach that allowed us to assess the
occurrences, antecedents, and consequences of different prosocial
behaviors in natural settings, a method that to date has not been
widely applied in the prosocial literature. This design is also better
suited to demonstrate short-term effects of helping on well-being.
We hypothesized that autonomous motivation, as opposed to con-
trol motivation, for a prosocial act (i.e., helping) would predict
well-being resulting from a helping behavior. We also expected
that satisfaction of the basic needs would be gained from autono-
mous helping and that this need satisfaction would in turn mediate
the effects of motivation to help on the subsequent well-being of
the helper.
Method
Participants and procedure. Participants were 138 univer-
sity students (41 men, 97 women) ages 18–24 (M ⫽ 20). Of these,
5% were African American, 4% were Hispanic, 15% were Pacific
Islander or Asian, and 73% were Caucasian. Three percent iden-
tified as another ethnicity. To take part in the present study,
students went to an online registration system and signed up for a
time of their choosing. Although participants received extra credit
for attending, reasonable measures were taken to avoid coercing or
pressuring participants. Attendance was online and voluntary and
participants could select to attend the present study among many
others or to attain extra credit through other means (projects, event
attendance, etc.). A similar recruitment procedure was utilized in
all subsequent studies. Participants attended an introductory ses-
sion held 24–48 hr before beginning the diary portion of the study.
During this session, they completed Big Five traits, trait well-
being, and trait need satisfaction measures and were instructed on
diary procedures. On each of 14 days, participants were asked to
complete a bedtime survey packet that included questions about
helping and well-being throughout the day, as well as filler ques-
tions included to mask hypotheses. At the end of the 14 days,
participants returned the packets and attended a debriefing session.
Person-level measures.
Brief Big Five traits. The brief 10-item measure of the Big
Five traits (Gosling, Rentfrow, & Swann, 2003) asks participants
to use 7-point scales to rate themselves on adjectives reflecting
neuroticism, extraversion, conscientious, openness, and agreeable-
ness. We focused on neuroticism in particular (␣⫽.91) to control
for its known effects on motivation (Deci & Ryan, 1985a) and
well-being (e.g., Costa & McCrae, 1980). In other words, because
226
WEINSTEIN AND RYAN