Content uploaded by Christopher Carpenter
Author content
All content in this area was uploaded by Christopher Carpenter
Content may be subject to copyright.
The Legitimization of Paltry Favors
Effect: A Review and Meta-Analysis
Kyle R. Andrews, Christopher J. Carpenter,
Allison S. Shaw, & Franklin J. Boster
A meta-analysis was conducted on the legitimization of paltry favors (LPF) effect
(Cialdini & Schroeder, 1976). A total of 19 studies met the inclusion criteria, with a
combined sample of 2,730 subjects. Excluding studies in which the LPF request was deliv-
ered via mail and those studies that accepted pledges as the dependent variable resulted
in a homogeneous set of effect sizes (r ¼.18, OR ¼2.41). While the data provide clues as
to possible mediating mechanisms, the cause of the effect is still not clear. Directions for
future research are suggested.
Keywords: Compliance-Gaining; Even a Penny Will Help; Legitimization of Paltry
Favors
Legitimizing paltry favors (LPF) as a means of gaining compliance was first studied
by Cialdini and Schroeder in 1976. Confederates dressed as American Cancer Society
solicitors added the phrase ‘‘even a penny will help’’ to a direct request, and com-
pliance nearly doubled. Since then, many studies have been conducted on the LPF
as a compliance gaining tactic, some designed to replicate the findings, some to test
boundary conditions, some to examine potential moderators, and some to investi-
gate possible cognitive mediators. This meta-analysis synthesizes this sometimes
paradoxical body of literature.
The critical feature of an LPF request is a message validating small contributions,
typically with the phrase ‘‘even a penny will help.’’ Researchers have tinkered with
this basic technique in a number of ways. In the original LPF study, Cialdini and
Schroeder (1976) also used the phrases ‘‘even a dollar will help’’ and ‘‘we’ve already
received some contributions, ranging from a penny on up.’’ The effect has been
Correspondence to: Kyle R. Andrews, Michigan State University, 473 Communication Arts and Sciences
Building, East Lansing, MI 48824, USA. E-mail: andre170@msu.edu
Communication Reports
Vol. 21, No. 2, July–December 2008, pp. 59–69
ISSN 0893-4215 (print)/ISSN 1745-1043 (online) #2008 Western States Communication Association
DOI: 10.1080/08934210802305028
found with zlotys, the Polish currency most equivalent to a penny. In addition to
changing the wording and monetary amount, the effect has been found with other
target requests, such as ‘‘even a flyer will help’’ and ‘‘even a few minutes will help.’’
The organizations and causes for which confederates collected money also varied and
included the American Cancer Society (Cialdini & Schroeder, 1976; Weyant & Smith,
1987), Urszula Jaworska Foundation (Dolinski, Grzyb, Olejnik, Prusakowski, &
Urban, 2005), Vermont Heart Fund (DeJong & Oopik, 1992), American Heart
Association (Reingen, 1978), Humane Society (Reeves & Saucer, 1993), Afghan refu-
gees and World Expo (Dolinski et al., 2005), The Good Friends Shelter Group
(Reeves, Macolini, & Martin, 1987), and university orientations for international
students (Takada & Levine, 2007).
Despite the success of variants on the basic LPF technique, the results have not
always been uniform. Some studies have found no or little effect, and some have even
found a negative effect (DeJong & Oopik, 1992; Reeves & Saucer, 1993, Study 1, no
commitment condition). The reason for the discrepant findings is unclear, but there
are a number of moderators that could be the cause. For instance, the studies vary on
the physical immediacy of the request (face-to-face requests v. mailed solicitations),
the target behavior requested (money, pledges, or time commitments), demographic
characteristics of the sample (e.g., age, sex of requester, sex of subject), wording of the
request, and where the request was made.
In addition to the apparent variance in effects across studies, there is no consensus
on the cognitive mechanisms that cause the effect. With a few exceptions, subsequent
studies have focused on replicating the results and boundary testing, and not on
investigating mediating mechanisms. Cialdini and Schroeder (1976) eliminated per-
ceived need as an explanation, and proposed that the results were potentially caused
by self-presentation goals, barrier removal (taking away the ability to say no), guilt, or
sympathy. DeJong and Oopik (1992) investigated the LPF in a mailed request and did
not find effects, potentially evidence for the self-presentation explanation. Dolinski
et al. (2005) implied that the technique may work because in many of the LPF experi-
ments, the requestor engaged in dialogue with the subject before making the request;
eliminating dialogue, however, did not eliminate the LPF effect.
Due to the discrepant results and the multiple potential moderators, a meta-
analysis was conducted. Because to our knowledge no such analysis exists, it
has the potential to provide answers not otherwise discernable and to set a future
LPF research agendum.
Method
Meta-Analytic Technique
The Hunter and Schmidt (1990) variance-centered method of meta-analysis was
employed to examine these data.
1
The first step in performing this meta-analysis
was to locate pertinent studies. LPF studies were found through a number of
methods. First, searches were conducted on PsychInfo, PubMed, Communication
& Mass Media Complete, and ProQuest Dissertations & Theses using the terms
60 K. R. Andrews et al.
‘‘legitimization of paltry favors,’’ ‘‘even a penny will help,’’ and ‘‘even a dollar will
help.’’ The literature reviews and reference sections of these articles were then
searched to see if any studies were missed.
The second step was to develop inclusion criteria. For a study to be included in
this meta-analysis, it had to fulfill four conditions. First, the study had to have a con-
dition that legitimized a truly minimal contribution (e.g., Brockner, Guzzi, Kane,
Levine, & Shaplen, 1984, was excluded because it used an ‘‘even a dollar will help’’
message induction). Second, it had to have a control condition that used a (relatively)
direct request (e.g., Mark & Shotland, 1983, was eliminated because the control
condition included pregiving). Third, the unit of analysis had to be an individual
decision to comply or not to comply (e.g., Perrine & Heather, 2000, was eliminated
because the unit of analysis was the donation receptacle). Fourth, the experiment
had to provide sufficient information for an effect size to be computed. A total of
19 studies from 11 articles were found that fulfilled these selection criteria, with year
of publication ranging from 1976 to 2007. All located studies were published in
English. All of the experiments included in the final analyses were performed outside
of laboratories (i.e., in natural settings). Some of the participants in these studies may
have been students, but certainly not all of them.
The third step was to transform quantitative information in each pertinent report
to a common metric. In this meta-analysis we chose the familiar metric r, the Pearson
Product Moment Correlation Coefficient. Fourth, the effects were weighted by
sample size and averaged (mean). Fifth, the variability in effect sizes across studies
(i.e., homogeneity) was assessed with the Hunter and Schmidt (1990) chi square
test. Sixth, a sampling error analysis was performed to ascertain if the results were
homogeneous across studies or if there was substantial heterogeneity. The latter case
would indicate the presence of differential methodological artifacts, moderator
variables, or both. The last step involved attempting to isolate important moderators
if they existed.
Instrumentation
Examining this corpus for differences in study execution led to the measurement of a
number of potential moderators (see Table 1). This set of measures included presen-
tation mode (face-to-face v. other), sex of requestor, sex of subject, age of subject
(adult v. other), phrasing of the request (e.g., ‘‘even a penny will help,’’ ‘‘taking
contributions from a penny on up’’), length of the message, type of request (prosocial
or not), requesting organization (e.g., American Cancer Society, Humane Society),
the number of confederates, whether or not the confederates were blind to condition
and hypothesis, year of publication, and country in which the experiments were con-
ducted (Poland vs. United States). Four researchers agreed on all coding decisions.
Results
There were 19 studies with a combined sample size of 2,730 for which a compliance
effect size could be computed (see Table 2). The weighted correlation for this set of
Communication Reports 61
Table 1 Descriptive Information for Studies Meeting Inclusion Criteria
Study Mode
1
DV
2
Sex of
requestor
3
Number of
requestors Place
4
Blind to
hypotheses
5
Country
LPF
message
6
Control
message
7
Cialdini & Schroeder (1976), Study 1 FTF I B 2 H NA USA 1 1
Cialdini & Schroeder (1976), Study 2 FTF I B 2 H Y USA 2 1, 2
DeJong & Oopik (1992) M I I NA H NA USA 1, 3 1, 3
Dolinski et al. (2005), Study 1 FTF I B 1 P Y Poland 1, 4 1, 4
Dolinski et al. (2005), Study 2 FTF I B 1 P Y Poland 1, 4 1, 4
Dolinski et al. (2005), Study 3 FTF P M 1 P Y Poland 1, 4 1, 4
Fraser & Hite (1989) FTF I B 1 H Y USA 1 1
Fraser, Hite, & Sauer (1988) FTF I B 1 H Y USA 1 1
Reeves, Macolini, & Martin (1987) FTF I B 2 H NA USA 1, 5 1, 2
Reeves & Saucer (1993), Study 1,
face-to-face condition
FTF I M 2 P N USA 1, 5 1, 2
Reeves & Saucer (1993), Study 1,
commitment condition
FTF P M 2 P N USA 1, 5 1, 2
Reeves & Saucer (1993), Study 1,
no commitment condition
FTF P M 2 P N USA 1, 5 1, 2
62
Reeves & Saucer (1993), Study 2,
face-to-face condition
FTF I M 2 P Y USA 1, 5 1, 2
Reeves & Saucer (1993), Study 2,
commitment condition
FTF P M 2 P Y USA 1, 5 1, 2
Reingen (1978), Study 1 FTF I B 2 P Y USA 1 1
Reingen (1978), Study 2 FTF I M 1 P Y USA 1 1
Takada & Levine (2007) FTF I F 1 P Y USA 1 1
Weyant (1984) FTF I B 2 H NA USA 1, 3 1, 3
Weyant & Smith (1987), Study 1 FTF I B 2 H Y USA 1, 3 1, 3
1
Presentation mode: FTF ¼face to face, M ¼mailed.
2
Dependent variable: I ¼immediate, P ¼pledge.
3
Sex of requestor: M ¼male, F ¼female, B ¼both male and female, I ¼impersonal request.
4
Place request made: H ¼subject’s home, P ¼public.
5
Whether confederate was blind to condition.
6
LPF Message: 1 ¼variant of ‘‘even a penny will help,’’ 2 ¼‘‘contributions from a penny on up,’’ 3 ¼included pregiving, 4 ¼included dialogue induction, 5 ¼included
normative information.
7
Control message: 1 ¼simple direct request, 2 ¼included normative information, 3 ¼included pregiving, 4 ¼included dialogue induction.
63
studies was .11 (unweighted mean ¼.18) with a weighted variance of .019
(unweighted variance ¼.026). The weighted odds ratio was 1.92 (unweighted
OR ¼2.71). The variance that would be expected from sampling error alone, given
a set of 19 studies with 2,730 participants, is .007, a figure that is significantly less
than the obtained variance (v
2
(18, N¼2730) ¼53.15, p<.001).
2
Table 2 Compliance Effects for Studies Meeting Initial Inclusion Criteria by Moderator
Control LPF
Compliance
effect
Study %
Complaint
subjects=n%
Compliant
subjects=nr OR
Face to face, dependent variable collected immediately
Cialdini & Schroeder (1976), Study 1 28.6 12=42 50.0 21=42 0.22 2.50
Cialdini & Schroeder (1976), Study 2 32.3 10=31 58.1 18=31 0.29 3.37
Dolinski et al. (2005), Study 1 50.0 30=60 68.3 41=60 0.19 2.16
Dolinski et al. (2005), Study 2 31.7 38=120 41.7 50=120 0.10 1.54
Fraser & Hite (1989) 23.8 19=80 30.0 24=80 0.07 1.38
Fraser, Hite, & Sauer (1988) 15.0 24=160 34.0 54=160 0.22 2.89
Reeves, Macolini, & Martin (1987) 30.0 9=30 56.7 17=30 0.27 3.05
Reeves & Saucer (1993), Study 1,
face-to-face condition
40.0 6=15 66.7 10=15 0.27 3.00
Reeves & Saucer (1993), Study 2,
face-to-face condition
34.4 11=32 66.7 20=30 0.32 3.82
Reingen (1978), Study 1 19.0 6=32 47.0 15=32 0.30 3.82
Reingen (1978), Study 2 11.0 3=28 39.0 11=28 0.33 5.39
Takada & Levine (2007) 13.8 4=29 23.1 6=26 0.12 1.88
Weyant (1984) 39.0 23=59 57.0 30=53 0.18 2.04
Weyant & Smith (1987), Study 1 36.0 34=94 39.0 37=95 0.03 1.13
Combined (weighted by sample size
where applicable)
28.2 229=812 44.2 354=802 0.18 2.41
Mailed request DeJong & Oopik (1992) 7.0 24=340 4.0 14=348 0.07 0.55
Pledge as dependent variable
Dolinski et al. (2005), Study 3 30.8 37=120 43.3 52=120 0.13 1.72
Reeves & Saucer (1993), Study 1,
commitment condition
33.3 5=15 80.0 12=15 0.47 8.00
Reeves & Saucer (1993), Study 1,
no commitment condition
86.7 13=15 66.7 10=15 0.24 0.31
Reeves & Saucer (1993), Study 2,
commitment condition
35.5 11=31 61.1 22=36 0.26 2.86
Combined (weighted by sample
size where applicable)
36.5 66=181 51.6 96=186 0.11 2.27
64 K. R. Andrews et al.
A search for variables that could account for the dispersion yielded two potential
moderators. First, the vast majority of the experiments involved face-to-face
solicitation (M¼.17, OR ¼2.38, K¼18, N¼2,042). The one experiment that
employed a mail solicitation produced an effect size substantially variant from the
face-to-face solicitations (M¼.07, OR ¼.55, N¼688). Because in only one
experiment was the solicitation not made face-to-face, it is possible that this study
was aberrant in some other way. Thus, firm conclusions concerning the moderating
effect of presentation mode must await subsequent research. Second, most experi-
ments involved collecting money directly from the participants, but five experiments
employed pledges as the dependent variable. The latter produced larger, but not sig-
nificantly different, effects than did the former (pledge, M¼.15, OR ¼2.27, N¼422
v. money, M¼.11, OR ¼1.85, N¼2,308; z¼0.75, ns;r¼.09), although the relation-
ship is reversed (and still not significant) when the lone mail solicitation is not
included in the analysis (see below).
Eliminating the experiment in which the solicitation was made by mail and the
five experiments that employed pledges as the dependent variable yielded a set of
13 experiments (N¼1,620). The weighted mean effect size in these experiments
was r¼.18 (OR ¼2.41, weighted variance ¼.009). The variance expected from
sampling error alone is .008. Testing the null hypothesis that the weighted variance
did not differ substantially from zero indicated that the null hypothesis could
not be rejected (v
2
(12, N¼1,620) ¼15.57, p¼.21). Thus, the data are consistent
with the hypothesis that this set of 13 experiments produces a set of homogeneous
effect sizes. The null hypothesis that the weighted variance did not differ substantially
from zero for the group of five pledge experiments could also not be rejected
(v
2
(4, N¼422) ¼8.88, p ¼.06).
A number of additional moderators were examined, but none accounted for a
statistically significant amount of variance. These included presentation mode,
sex of requestor, sex of subject, age of subject, phrasing of the request, length of
the message, type of request, requesting organization, the number of confederates,
whether or not the confederates were blind to condition and hypothesis, year of
publication, and country in which the experiment(s) was conducted.
Finally, a file drawer analysis was conducted. The results indicated that 34 studies
of N¼125 (the mean sample size of this database), each producing an effect size of
zero, would be necessary to reduce the mean effect size to .05. A total of 221 studies of
the same mean sample size, each producing an effect size of zero, would be necessary
to reduce the mean effect size to .01.
Discussion
Summary of the LPF Studies
Despite the fact that only 19 studies met the inclusion criteria, there is still sufficient
evidence from the literature to draw conclusions regarding the legitimization of
paltry favors technique. The analysis found that adding the LPF phrase produced
more compliance than the direct request control. Two moderators were found.
Communication Reports 65
One moderator was the use of a face-to-face solicitation. When the compliance-
gaining attempt was not face-to-face, the technique was not successful. But, one must
be cautious in drawing firm conclusions; in only one experiment was the solicitation
not made face-to-face. The other moderator was whether pledges or immediate
donations were requested. After eliminating the study that was not a face-to-face
solicitation, pledges produced smaller compliance effect sizes than immediate
monetary donations. When the studies that contained these two moderators were
eliminated, a homogeneous effect of r¼.18 was found for LPF on compliance. An
effect size of this magnitude is comparable to effect sizes reported for other compliance
gaining techniques, such as the foot-in-the-door (FITD) and the door-in-the-face
(DITF; Burger, 1999; Dillard, Hunter, & Burgoon, 1984; O’Keefe & Hale, 1998).
Two studies found interaction effects that cannot be examined using meta-analytic
techniques without additional studies. Takada and Levine (2007) found that people
who were high in perspective-taking were more likely than those low in perspective-
taking to comply with an LPF request but less likely to comply with the control request.
Fraser, Hite, and Sauer (1988) found that when a large donation suggestion was men-
tioned before the LPF request, the LPF technique did not increase compliance rates
above the control rate. No other studies that examined potential moderators found sig-
nificant evidence for them. For example, Fraser and Hite (1989) found that adding a
statement that there was an organization that would make a matching offer did not
affect compliance rates, while Reingen (1978) found that combining the LPF with
the DITF or FITD technique did not produce more compliance than the LPF alone.
Two studies examined possible mediators for the LPF effect. Cialdini and
Schroeder (1976) proposed that the targets of an LPF request would conclude that
if a charity was asking for pennies, they must have a high need for donations. Cialdini
and Schroeder’s subsequent experiments eliminated this explanation. Dolinski et al.
(2005) proposed that the LPF increased compliance by creating a dialogue with the
target. Here also, the results of the subsequent experiments produced no evidence
consistent with this hypothesis; therefore, at this juncture in the development of
the LPF literature, there is little evidence to support specific mediational mechanisms
for the observed effect. A number of possible explanatory mechanisms remain to be
considered in future studies. Perhaps the LPF message removes a reason to say no to
the request or that self-presentation goals lead targets to want to avoid being
perceived negatively by those soliciting their compliance or those observing the
compliance gaining interaction; the extent to which these reasons are actually differ-
ent is also undetermined. It is also possible that the LPF leads to targets anticipating
negative affect if they do not contribute. Finally, it is also possible that the most
obvious explanation is true, namely that the LPF condition legitimizes the contri-
bution of small sums, making subjects feel comfortable donating what they can afford
rather than what they think the charity might expect. While it was not possible to
calculate the mean effect size for average donation, evidence from individual studies
indicates that the average donation does not differ substantially based on condition,
making the legitimization argument less likely. Nevertheless, more research is needed
before hard conclusions can be drawn.
66 K. R. Andrews et al.
Limitations and a Comment on Conducting Compliance Gaining Experiments
A limitation of this body of research stems from the nature of the control group
employed typically in LPF experiments. As Mark and Shotland (1983) pointed out,
the generalization of the results of the typical LPF experiment to a phenomenon such
as donating to charities is marred by the fact that charities rarely employ only a direct
request in their campaigns. Instead, their typical message is likely much more
persuasive than the direct request control message, and, consequently, the effects
of the extant LPF literature may overestimate the benefit of the technique in applied
contexts. Of course, this limitation can be addressed by subsequent research.
Moreover, it is important to note that the LPF technique may have limited appli-
cations. In the set of studies reviewed here, the topic was prosocial in each case. It is
unclear how the technique could be adopted to compliance gaining situations in
which the request is not clearly prosocial or is decidedly antisocial.
Finally, it is worth noting that, although meta-analysis can be used to examine
moderator effects, and may uncover evidence of interesting, important, and non-
obvious moderators, it may also fail to discover other, equally interesting, important,
and nonobvious moderators. For instance, many individual differences (e.g., self-
esteem, dogmatism, and need for cognition) cannot yet be examined via meta-
analysis, as they are not measured in the primary research literature. If such unknown
individual differences do moderate the LPF-compliance relationship, they can
contribute to the heterogeneity of the effect estimates observed across studies. Unfortu-
nately, finding heterogeneous effects does not provide a clear clue as to which individ-
ual differences might serve as moderators. Thus, the investigation of such moderators
in primary studies, as in Takada and Levine (2007), has the potential to make impor-
tant contributions to our understanding of the LPF-compliance relationship.
Conclusion
This meta-analysis demonstrated that when solicitations were made face-to-face and
donations were collected at the time of the request, a mean effect size was obtained
that is comparable to other techniques known to be effective in gaining compliance
(e.g., door-in-the-face and foot-in-the-door). Moreover, the LPF technique is easier
to implement than other, multistage strategies. Additionally, two research opportu-
nities arise from this review. First, there is a lack of definitive research locating
mediators of the effect, making this area ripe for additional exploration. Second,
the technique has not been implemented with antisocial or self-interested requests.
Subsequent research investigating these issues has the potential to circumscribe the
boundary conditions surrounding this phenomenon.
Notes
[1] Clearly, there are alternative methods of performing meta-analysis (e.g., Glass, McGaw, &
Smith, 1981; Hedges & Olkin, 1985; Rosenthal, 1991). Hunter’s variance-centered method
has been used extensively and effectively in communication research (e.g., Allen et al.,
Communication Reports 67
2007; Boster & Mongeau, 1984; Dillard et al., 1984; Hullett, 2005). Moreover, it has been our
observation that alternative methods rarely yield estimates that differ substantially when the
methods are executed competently (e.g., Schmidt & Hunter, 1999). In any event, Tables 1
and 2 present the raw data that allow our conclusions to be examined with alternative
methods.
[2] Another dependent variable of interest is the difference in mean donation in the
experimental and control groups. In some studies (e.g., DeJong & Oopik, 1992; Dolinski
et al., 2005; Fraser & Hite, 1989), this value was reported for all experimental and control
group subjects. In other studies (e.g., Fraser et al., 1988; Reeves & Saucer, 1993; Takada &
Levine, 2007), this value was reported only for those who donated. This fact made the
calculation of the within cell variances impossible, and thus, made the calculation of an effect
size impossible for this dependent measure.
References
Allen, M., Emmers-Sommer, T. M., D’Alessio, D., Timmerman, L., Hanzal, A., & Korus, J. (2007).
The connection between the physiological and psychological reactions to sexually explicit
materials: A literature summary using meta-analysis. Communication Monographs,74,
541–560.
Boster, F. J., & Mongeau, P. (1984). Fear-arousing persuasive messages. In R. N. Bostrom &
B. H. Westley (Eds.), Communication yearbook (Vol. 8, pp. 330–375). Newbury Park,
CA: Sage.
Brockner, J., Guzzi, B., Kane, J., Levine, E., & Shaplen, K. (1984). Organizational fundraising:
Further evidence on the effects of legitimizing small donations. Journal of Consumer Research,
11, 611–614.
Burger, J. M. (1999). The foot-in-the-door compliance procedure: A multiple-process analysis and
review. Personality and Social Psychology Review,3, 303–325.
Cialdini, R. B., & Schroeder, D. A. (1976). Increasing compliance by legitimizing paltry contribu-
tions: When even a penny helps. Journal of Personality and Social Psychology,34, 599–604.
DeJong, W., & Oopik, A. J. (1992). Effect of legitimizing small contributions and labeling potential
donors as ‘‘helpers’’ on responses to a direct mail solicitation for charity. Psychological
Reports,71, 923–928.
Dillard, J. P., Hunter, J. E., & Burgoon, M. (1984). Sequential-request persuasive strategies:
Meta-analysis of foot-in-the-door and door-in-the-face. Human Communication Research,
10, 461–488.
Dolinski, D., Grzyb, T., Olejnik, J., Prusakowski, S., & Urban, K. (2005). Let’s dialogue about
penny: Effectiveness of dialogue involvement and legitimizing paltry contribution techni-
ques. Journal of Applied Social Psychology,35, 1150–1170.
Fraser, C., & Hite, R. E. (1989). The effect of matching contribution offers and legitimization of
paltry contributions on compliance. Journal of Applied Social Psychology,19, 1010–1018.
Fraser, C., Hite, R. E., & Sauer, P. L. (1988). Increasing contributions in solicitation campaigns: The
use of large and small anchorpoints. The Journal of Consumer Research,15, 284–287.
Glass, G. V., McGaw, B., & Smith, M. L. (1981). Meta-analysis and social research. Beverly Hills, CA:
Sage.
Hedges, L. V., & Olkin, I. (1985). Statistical methods for meta-analysis. Orlando, FL: Academic
Press.
Hullett, C. R. (2005). The impact of mood on persuasion: A meta-analysis. Communication
Research,12, 423–442.
Hunter, J. E., & Schmidt, F. L. (1990). Methods of meta-analysis: Correcting error and bias in research
findings. Newbury Park, CA: Sage.
68 K. R. Andrews et al.
Mark, M. M., & Shotland, R. L. (1983). Increasing charitable contributions: An experimental
evaluation of the American Cancer Society’s recommended solicitation procedures. Journal
of Voluntary Action Research,12, 8–21.
O’Keefe, D. J., & Hale, S. L. (1998). The door-in-the-face influence strategy: A random-effects
meta-analytic review. In M. E. Roloff (Ed.), Communication Yearbook (Vol. 21, pp. 1–33).
Thousand Oaks, CA: Sage.
Perrine, R. M., & Heather, S. (2000). Effects of picture and even-a-penny-will-help appeals on
anonymous donations to charity. Psychological Reports,86, 551–559.
Reeves, R. A., Macolini, R. M., & Martin, R. C. (1987). Legitimizing paltry contributions:
On-the-spot vs. mail-in requests. Journal of Applied Social Psychology,17, 731–738.
Reeves, R. A., & Saucer, P. R. (1993). A test of commitment in legitimizing paltry contributions.
Journal of Social Behavior and Personality,8, 537–544.
Reingen, P. H. (1978). On inducing compliance with requests. The Journal of Consumer Research,5,
96–102.
Rosenthal, R. (1991). Meta-analytic procedures for social research. Newbury Park, CA: Sage.
Schmidt, F. L., & Hunter, J. E. (1999). Comparison of three meta-analysis methods revisited: An
analysis of Johnson, Mullen, and Salas (1995). Journal of Applied Psychology,84, 144–148.
Takada, J., & Levine, T. R. (2007). The effects the even-a-few-minutes-would-help strategy,
perspective taking, and empathic concern on the successful recruiting of volunteers on
campus. Communication Research Reports,24, 177–184.
Weyant, J. M. (1984). Applying social psychology to induce charitable donations. Journal of Applied
Social Psychology,14, 441–447.
Weyant, J. M., & Smith, S. L. (1987). Getting more by asking for less: The effects of request size on
donations of charity. Journal of Applied Social Psychology,17, 392–400.
Communication Reports 69
